Docstoc

HIFI calibration use cases

Document Sample
HIFI calibration use cases Powered By Docstoc
					         HIFI calibration use cases
Version DRAFT 1.0 of 20/10/2003, by Frank Helmich, Peter Roelfsema and David Teyssier




  Abstract
  This document gives the use cases that have been defined for HIFI.

                                                        Document: ICC/2001-005
                                                        DRAFT 1.0 of 20/10/03
                                                        137 pages
                    HIFI calibration use cases            Doc.: ICC/2001-005
                                                          Date: 20/10/03
                                                          Issue: Draft 1.0




Document approval

  Prepared by:               Frank Helmich, Peter 20 October, 2003
                             Roelfsema and David
                             Teyssier
  Checked by:
  Authorized by:




Distribution

   ESA:
       P. Estaria                    ESTEC

   HIFI steering committee
        Th. de Graauw                SRON


   HIFI project:
        H. Aarts                     SRON
        C. Wafelbakker               SRON




                                      2
                    HIFI calibration use cases                Doc.: ICC/2001-005
                                                              Date: 20/10/03
                                                              Issue: Draft 1.0




Revision history

  Version      Date         Changes                          Author
  Draft 0.1    26/9/2001    First version assembled from     P. Roelfsema
                            inputs by the HIFI calibration
                            group
  Draft 0.2    20/02/2002   More equations added             D. Teyssier
  Draft 0.3    01/07/2002   Update of most use cases to      Frank Helmich
                            reflect increased knowledge
  Draft 0.31   01/04/2003   Cleaned up all formatting        P. Roelfsema
  Draft 0.4    27/08/2003   Update of most use cases and     D. Teyssier
                            completion of missing items      F. Helmich
                            with inputs by NW and DB
                            Addition of some observing
                            mode verification UC’s
  Draft 1.0    20/10/2003   Last updates for 1.0 release     D. Teyssier
                                                             F. Helmich




                                       3
                      HIFI calibration use cases               Doc.: ICC/2001-005
                                                               Date: 20/10/03
                                                               Issue: Draft 1.0


Applicable documents

  Doc. ref.   Title
  AD1         HERSCHEL Science Management Plan, ESA/SPC(97)22
  AD2         HERSCHEL Science Implementation Requirements Document -
              SIRD (PT-03646)
  AD3         HERSCHEL Operations Interface Requirements Document - FOIRD
              (FP-ESC-RS-0001)
  AD4         HERSCHEL/PLANCK Ground Segment Interface Document - GSID
              (PT-04829)
  AD5         Guide to applying the ESA Software Engineering Standards (PSS-
              05-0) to small Software Projects (BSSC-96-2)
  AD6         Software Engineering & PA Standards for the HCSS
              (FIRST/FSC/DOC/0127)
  AD7         HCSS Software PA plan (FIRST/FSC/DOC/0161)
  AD8         HCSS Software Configuration Management Plan
              (FIRST/FSC/DOC/0166)
  AD9         Science Implementation Plan, ICC/1998-001
  AD10        Herschel/Planck Instrument Interface Document, SCI-PT-IIDA-
              04624




Reference Documents

  Doc. ref.   Title
  RD1         HERSCHEL Ground Segment Operations Concept and Ground
              Segment Document (PT-03056)
  RD2         HERSCHEL Operations Scenario Document
              (FIRST/FSC/DOC/0114)
  RD3         HERSCHEL Ground Segment Design Description
              (FIRST/FSC/DOC/0146)
  RD4         HERSCHEL Ground Segment Interface Requirements Document
              (FIRST/FSC/DOC/0117)
  RD5         List of Interface Control Documents (FIRST/FSC/DOC/0150)
  RD6         HCSS User Requirements Document (FIRST/FSC/DOC/0115)
  RD7         HCSS Actor Descriptions (FIRST/FSC/DOC/0157)
  RD8         HCSS Use-Cases (FIRST/FSC/DOC/0158)
  RD9         HCSS Supplementary Specifications (FIRST/FSC/DOC/0159)
  RD11        HIFI ICC Actor Descriptions (HIFI-ICC-2001-nnn)
  RD12        HIFI ICC Use-Cases (HIFI-ICC-2001-nnn)
  RD13        HIFI ICC Supplementary Specifications (HIFI-ICC-2001-nnn)
  RD14        HIFI calibration plan (LRM-ENS/HIFI/PL/2000-001)
  RD10        HCSS Software Project Management Plan (FIRST/FSC/DOC/0116)
  RD15        End User Requirements for HIFI Interactive analysis (ICC/2001-004)
  RD16        Chopper Wheel Calibration Method on HIFI (LRM-
              ENS/HIFI/CAL/2000-002)


                                       4
            HIFI calibration use cases                     Doc.: ICC/2001-005
                                                           Date: 20/10/03
                                                           Issue: Draft 1.0


RD17   HIFI Observing Mode Document, ICC/2002-001
RD18   HIFI AIV plan, SRON-G/HIFI/PL/1999-001
RD19   Detailed test plan for ILT, SRON-G/HIFI/PL/2001-001
RD20   The Intensity Calibration for HIFI, ALMA memo 442
RD21   The HIFI Spatial Response, C. Kramer
RD22   The HIFI Frequency calibration framework, L3AB/HIFI/CAL/2003-02
RD23   HIFI most wanted frequencies, SRON/HIFI/TECH/2003-001
RD24   Standing wave analyses: KOSMA data. LRM-ENS/HIFI/TECH/2001-
       001
RD25   The impact of standing wave in the LO path of a heterodyne receiver,
       Siebertz et al. 2003, IEEE Transactions on Microwave Theory and
       Techniques
RD26   Solar Bodies as Calibration Sources (HIFI), L3AB/HIFI/CAL/2003-01
RD27   Quasi-optical verification of HIFI, SRON-G/HIFI/TN/2003-xxx
RD28   QLA for WBS Use Cases, hifi-icc/-00x
RD29   HRS QLA Use Cases for ILT, CESR-HRS-SP-3F11-082
RD30   Technical report: realisation of a gas cell for tests of Herschel/HIFI
       IAS-LERMA/HIFI/AIV/2002-01
RD31   Absolute Hot Black Body design, SRON-G/HIFI/TN/2003-xxx, TBD
RD32   HIFI Observing mode definition, ICC-2003-004
RD33   Note on Allan Variance calculation method, 2000-06-15
RD34   Calibration Source Requirements for Herschel/HIFI, LRM-
       ENS/HIFI/SP/2000-001
RD35   HIFI Observing mode Description, ICC/2003-008
RD36   HIFI Observing mode Implementation, ICC/2003-009
RD37   HIFI Observing mode Calibration, ICC/2003-010




                                 5
                                HIFI calibration use cases                                  Doc.: ICC/2001-005
                                                                                            Date: 20/10/03
                                                                                            Issue: Draft 1.0




                                                   Contents

HIFI calibration use cases ......................... 1
Contents.................................................................................................. 6
1 Introduction ...................................................................................... 7
    1.1      Summary of actor descriptions................................................................. 7
2      The HIFI calibration use cases........................................................ 8
    2.1    UC-1 Create a calibration measurement .................................................. 8
    2.2    UC-1.1 Sensitivity properties ...................................................................11
      2.2.1   UC-1.1.1 Measure Instrument Sensitivity .........................................14
      2.2.2   UC-1.1.2 Determine Instrument Response Times ...........................17
      2.2.3   UC-1.1.3 Determine Instrument Intensity Stability ..........................29
      2.2.4   UC-1.1.4 Make Long Duration Integrations ......................................34
      2.2.5   UC-1.1.5 Measure platforming...........................................................40
      2.2.6   UC-1.1.6 Measure Baseline ripple.....................................................42
      2.2.7   UC-1.1.7 Measure Continuum emission ...........................................46
    2.3    UC 1.2 Beam Properties............................................................................49
      2.3.1   UC 1.2.1 HIFI Focal-Plane Geometry ................................................52
      2.3.2   UC-1.2.2 Beam Patterns.....................................................................60
      Goldsmith P., 1998, ”Quasi-optical Systems”, IEEE Press UC-1.2.3 Pointing
      performances....................................................................................................63
      UC-1.2.3 Pointing performances .....................................................................64
    2.4    UC-1.3 Spectral properties .......................................................................67
      2.4.1   UC 1.3.1 Frequency calibration.........................................................69
      2.4.2   UC-1.3.2 Measure Instrument Line Profile .......................................75
      2.4.3   UC-1.3.3 Measure Spectral Purity .....................................................85
      2.4.4   UC 1.3.4 Measure Side Band Ratio ...................................................92
      2.4.5   UC 1.3.5 Measure diplexer performance ..........................................97
    2.5    UC-1.4 Intensity properties.......................................................................99
      2.5.1   UC-1.4.1 Measure internal calibrator radiometric properties .......102
      2.5.2   UC-1.4.2 Measure internal calibrator coupling ..............................105
      2.5.3   UC 1.4.3 Measure Telescope Aperture Efficiency .........................109
      2.5.4   UC-1.4.4 Measure telescope beam efficiency ................................112
      2.5.5   UC-1.4.5 Intensity calibration ..........................................................115
      2.5.6   UC-1.4.6 Measure dynamic range ...................................................118
      2.5.7   UC-1.4.7 Measure non-linearity.......................................................120
    2.6    UC-1.5 Validation of observing modes..................................................122
      2.6.1   UC 1.5.1 Validation Double Beam Switching Mode.......................124
      2.6.2   UC-1.5.2 Validation On-the-fly Mapping .........................................127
      2.6.3   UC 1.5.3 Validation load-chop with OFF calibration .....................130
      2.6.4   UC 1.5.4 Validation Frequency survey with chopper ....................134




                                                           6
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0




1 Introduction

While the general calibration structure is described in RD 14 (the calibration plan),
there is a need to spell out the HIFI calibration requirements in a separate document.
Because the whole Herschel ground segment uses Use Cases to describe the
requirements, the HIFI ICC has chosen the same approach to describe the HIFI
calibration requirements. The advantage of the Use Case approach is that it allows
for a fast update of the requirements whenever necessary; it describes the actors
within the calibration group and even within the Herschel Ground Segment; and it
describes in normal wording the approach the actors have to take. Note that within
one Use Case one person can play several actor roles! For more information on
actors and use cases within the Herschel Ground Segment, and the HIFI ICC in
particular, the reader is referred to RD7, RD8, RD11 and RD 12.
Below we give a summary of the most important actors in the Use Cases of this
document. They are provided to make the document more self contained, however,
the official actor descriptions are found in RD11.

The in-orbit calibration needs to follow the Instrument Level Tests as closely possible
as ILT delivers "calibration" of the instrument without the telescope. While these Use
Cases do not attempt to be complete on the description of the Instrument Level
Tests, reference will be made, whenever possible, to the ILT detailed test plan. The
use cases will try to follow the topical description in the HIFI AIV plan, whenever
possible


1.1   Summary of actor descriptions

AS      Astronomer               The general Herschel astronomer
CS      Calibration Scientist    Scientist or Instrument Scientist involved in the HIFI
                                 calibration program
CO      Calibration Operator
CM      Calibration Manager
IE      Instrument Engineer
MP      Mission Planner




                                           7
                      HIFI calibration use cases               Doc.: ICC/2001-005
                                                               Date: 20/10/03
                                                               Issue: Draft 1.0




2 The HIFI calibration use cases

2.1   UC-1 Create a calibration measurement

Level:                  Summary
Scope:                  HIFI-ICC
Version:                0.1
Status:                 Draft
Author:                 Michel Pérault - 08/31/01

General Grade: High
Grade for ILT: Medium
Grade for implementation: Medium

Brief description:
This use case describes the commonalities in creating a calibration measurement of
presumably all types covered by the calibration requirements.

Phase:
ILT                                                                     Y
IST/EET/GST...                                                          Y
LEOP/Commissioning                                                      Y
Calibration/PV                                                          Y
Science Demonstration                                                   Y
Routine Operations                                                      Y
Post Operations                                                         N

Actors:
CS: Calibration Scientist
CO: Calibration Operator (?)
CM: Calibration Manager
MP: Mission (or measurement) Planner
IE: Instrument (System) Engineer
AS: Astronomer
DP: Data Processing pipeline
GS: Ground Segment

Triggers:
   • Requests from Herschel test manager
   • Calibration plan
   • Observers
   • Failure recovery modes

Preconditions:


                                          8
                       HIFI calibration use cases                   Doc.: ICC/2001-005
                                                                    Date: 20/10/03
                                                                    Issue: Draft 1.0


   •   availability of CUS and (uplink) commanding system
   •   availability of command validation system
   •   availability of calibration source database and visibility tool
   •   availability of measurement optimisation tools (at least the instrument
       handbook, "observation time estimator")
   •   access to the calibration expert system. At least :
              o the calibration cook-book
              o up-to-date reports: progress, status, performance, ...
              o calibration journal and maintenance database
              o calibration processing input database (to file in information for later
                   use by the calibration analyst)
   •   Availability of (downlink) retrieval and data processing systems
   •   Operational instrument (and S/C or Ground Support Equipment).
   •   access to operation planning team

Minimal post conditions:
   • quick return on measurement status (validated, scheduled, performed,
     successfully?, processed, analysed, within-specs?, ...)

Success post conditions:
  • calibration analysis returns result with requested accuracy
  • measurement and results properly filed and input to the calibration expert
     system.

Stakeholders and interests:
   • Calibration Group
   • Instrument System Group
   • Herschel System Group
   • General astronomers

Main success scenario:
  1. CS: edit the appropriate measurement sequence file
  2. CS: fill-in parameters following optimisation process
  3. CO: check validity of sequence file with validation tool
  4. CM: enable calibration measurement
  5. CS: file in calibration measurement to scheduling system
  6. CS: update calibration maintenance database
  7. CS: file in information for later processing and analysis
  8. MP: schedules the measurement and reports (eventual loop back with CS/CM)
  9. GS: perform measurement and archive data
  10. DP: process data and archive
  11. CS: analyse resulting data
  12. CS: update calibration output database
  13. CM/CS: trigger higher level analysis, reporting, calibration updates, etc.

Alternative Success Scenario:


                                           9
                       HIFI calibration use cases                   Doc.: ICC/2001-005
                                                                    Date: 20/10/03
                                                                    Issue: Draft 1.0


   1. SE: include a calibration measurement in astronomical observation template
      (AOT) (with predefined parameters)
   2. AS: insert pre-configured calibration commands in observing sequence
      through means of the AOT.
   3. DP: apply calibration results to the observer's data, and file in calibration result
      into calibration database.

Frequency of occurrence:
As specified in plans or requested by authorized persons (including observers for
their share).

Open issues:
Optimisation of the calibration measurement ensemble

Comments:




                                           10
                        HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                    Date: 20/10/03
                                                                    Issue: Draft 1.0


2.2   UC-1.1 Sensitivity properties

Level:          Summary
Scope:          HIFI-ICC
Version:        0.1
Status:         Draft
Author:         Michel Pérault - 08/31/01
                Michel Pérault - minor update: 09/17/01
                Frank Helmich - minor update: 29-May-02

General Grade: High
Grade for ILT: Medium
Grade for implementation: Medium

Brief description:

This U.C. is a summary use case linking the Calibration general strategy (calibration
plan part 1) with the detailed measurement use cases.

Sensitivity is the most crucial area nowadays in submillimetre instrumentation,
crucially depending on new developments in detector technology, on the overall
quality of the optical and electronic set-up, and, to a lesser extent, of the optimisation
of operational procedures and data processing. This statement of course does not
imply that other areas (spatial and spectral response, intensity calibration) are not
essential or difficult.

Predict or derive sensitivity is thus a permanent concern of the calibration scientist, of
the instrument scientist, and other actors as well. Sensitivity in the first place means:
"faintest (scientific) signal detectable under pre-determined conditions". The nature
of the signal needs to be specified: point/extended source, source environment,
monochromatic/continuum signal, frequency.           The sensitivity depends on the
instantaneous sensitivity, and on the noise model via observation mode and duration
(see the observing modes document: RD17). The proper derivation of sensitivity data
relies on a proper intensity calibration, and overall instrument calibration. Main
components in the HIFI case are the response of the instrument as a whole; and the
noise temperature to study the behaviour of the instrument. The hot, and especially,
cold load in HIFI provide the tools to study the behaviour. The constancy of the loads
will be used to monitor the sensitivity of the instrument from ILT to the end-of-Helium
phase.

Phase: all phases

Actors:
CS: Calibration Scientist
CO: Calibration Operator (?)
CM: Calibration Manager
MP: Mission (or measurement) Planner


                                           11
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


IE: Instrument (System) Engineer
AS: Astronomer
DP: Data Processing pipeline

Triggers:
   • Project Scientist Team
   • P.I.
   • Requests from Herschel test manager
   • Calibration plan
   • Observers
   • Failure recovery modes

Preconditions:
   • Availability of the equipment and analysis methods to achieve the goals of the
     considered child Use Cases

Minimal post conditions:
   • The measurement data required by the considered child Use Case have been
     recorded

Success post conditions:
  • Assessment of the instrument behaviour or of the calibration parameters of
     interest within the accuracy specified by the considered child Use Case

Stakeholders and interests:
   • Calibration Group
   • Instrument System Group
   • Herschel System Group
   • General astronomers
   • Integration and Test Team

Main success scenario:
The following aspects of HIFI sensitivity will be covered under section 2.2:
white noise (high frequency / thermal noise)

   1. measure instrument sensitivity - temporal drifts (low frequency noise) (UC
      1.1.1, section 2.2.1)
   2. measure stability - non-ideal band-pass calibration (spurious effects which limit
      the sensitivity (UC 1.1.3, section 2.2.3)
   3. measure platforming (non-linearity) (UC 1.1.5, section 2.2.5)
   4. measure baseline ripples (standing waves ?) (UC 1.1.6, section 2.2.6)

The above ingredients are then combined into a calibration sensitivity budget which
reads as:
   • make long duration integrations (UC 1.1.4, section 2.2.4)



                                          12
                      HIFI calibration use cases               Doc.: ICC/2001-005
                                                               Date: 20/10/03
                                                               Issue: Draft 1.0


   •  measure instrument observing efficiency (depending on obs.mode) (UC 1.1.2,
      section 2.2.2)
   • estimate sensitivity for continuum emission (UC 1.1.7, section 2.2.7)
(See also Use Case 1.4 -section 2.5- and its child use cases)

Frequency of occurrence:

Open issues:
  • Sensitivity predictions need to be based on a generalized noise model for the
      instrument. The details of this model are TBD. Can NW comment on this?

Comments:




                                        13
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


2.2.1 UC-1.1.1 Measure Instrument Sensitivity

Level:             User
Scope:             HIFI-ICC
Version:           0.1
Status:            Draft
Author:            David Teyssier - 08/31/01
                   David Teyssier - Updates 02/14/02 and 05/05/03

General Grade: High
Grade for ILT: High
Grade for implementation: High

Brief description:
This use case describes the procedure to measure and compute the receiver and
system temperatures defining the instrument sensitivity for short integrations. The
long integration noise model shall be measured in UC-1.1.4.

Phase: all phases except post operations

Actors:
CS: Calibration Scientist

Triggers:
   • Cal. plan
   • Observers
   • Failure notice?

Preconditions:
For both ILT and in orbit:
   • availability of internal or external hot and cold loads with known temperatures
      and coupling to the mixers (see UC-1.4.2)
   • knowledge of parameters required to compute the sensitivity (namely: side-
      band ratio (this can be the correction from SBR=1 from IL tests), forward
      efficiency, sky background temperature and effective temperature of losses.
   • DSB receiver and system temperatures from FP S/S tests and ILT

Minimal post conditions:
   • Get a spectrum of system and receiver temperature in the requested band and
     at the requested frequency

Success post conditions:
  • Estimate of DSB receiver and system temperatures Jrec and Jsys. These should
     be compared with values from FP S/S tests and from ILT




                                          14
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


   •    Accuracy 5% (TBC) Can we give an estimate for each band for an integration
        of 4 seconds on each load? Also give the equation or refer to end-user
        requirement document.
We need to define a criterion which makes a measurement validated or not. This may
be linked to the trend analysis in the sense that reproducibility vs ground-based and
existing in-orbit measurement should be checked. A measurement is also validated
only if the measurement scheme is considered as adapted (may be only assessed in-
orbit).

Stakeholders and interests:
   • Calibration Group
   • General astronomers
   • Instrument System Engineer
   • Integration and Test Team

Main success scenario:

   1. CS: Set hot load to required temperature. We call the load physical
      temperatures Thot and Tcold as given by temperature sensors

   Note: errors involved in the calibration load measurements are minimized with a
   maximum hot load temperature (RD16, RD20). This is always true, independent
   of the source intensity. Highest temperatures also allow to reduced the required
   integration time on the loads. The hot load shall thus always be set to its highest
   value (TBC).

   2. CS: Tune receiver(s)

   Check internal calibrator temperatures (hot and cold) and coupling (cf UC-1.4.2)

   3. Measure zero counts z

   4. Load exposures

          a. CS: Hot load exposure (integration time 0.1-3.5 sec WBS TBC; 0.4-25
             sec. HRS TBC)
          b. CS: Cold load exposure (integration time 0.1-3.5 sec WBS TBC; 0.4-25
             sec. HRS TBC)
          c. CS: OFF sky exposure (integration time TBC)

   Note: the times given here assume that the error on the receiver temperature
   should remain below 1% (see formula in RD20). On the ground, no sky
   measurement is available.

   5. CS: Compute the Y-factor:




                                          15
                       HIFI calibration use cases                        Doc.: ICC/2001-005
                                                                         Date: 20/10/03
                                                                         Issue: Draft 1.0



                                      chot − z
                                Y=
                                      ccold − z
      Note: With 100K Thot and 15K Tcold, Y is expected to be around 2 at low
      frequencies and 1.2 for the highest frequency bands

   6. CS: Compute receiver temperature Jrec and uncertainty. See RD20 for
      formula.

      Note: the corresponding equations are true for each velocity channel. Jrec can
      thus be given on the whole frequency band-pass, or be averaged over TBC
      channels to smooth eventual baseline effects or investigate standing wave
      pattern (cf UC-1.1.6).

   7. CS: Compute system temperature and uncertainty

  For a DSB system, one has:
                                          1 − ηl                  Trec
                    J sys = J bg ,eff +            × J T ,eff +
                                           ηl                     ηl

  where Jbg,eff and JT,eff are the effective cosmic background and ohmic and rearward
  losses (here assimilated to the telescope) radiation temperatures. Jeff refers to
  GssbJssb + (1-Gssb)Jisb and accounts for contribution from both sidebands, with Gssb
  the normalized sideband ratio (see RD20). ηl refers to the forward efficiency.
  The uncertainty on Jsys is presented in RD16.

   8. CS: return receiver and system temperatures with their uncertainties to
      calibration data-base
   9. (CS: File in a calibration report

Second level scenario: validate procedure and analysis with respect to time
   1. CS: Analysis of the measurement set:
      • Trend analysis
      • Health monitoring
      • Optimise procedure (e.g. order of the load measurement, etc..)
      • (if necessary) Update validation criterion

Frequency of occurrence:
   • In PV and every time a tuning and a calibration is done and/or the LO-freq.
     (e.g. for source Vlsr change) is changed

Open issues:
  • Precise design and expected performances of internal calibrators.

Comments:



                                                        16
                       HIFI calibration use cases                Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


2.2.2 UC-1.1.2 Determine Instrument Response Times

Level:                      Summary
Scope:                      HIFI-ICC
Version:                    0.1
Status:                     Draft
Author:                     N. D. Whyborn - 2001-09-20

General Grade: High
Grade for ILT: High
Grade for implementation: Low

Brief description:
Determine various important response times of the instrument through the
procedures described in the child use cases. The results should be included in the
AOT and observing time estimator.

Phase:
ILT                                                                    Y
IST/EET/GST...                                                         Y
LEOP/Commissioning                                                     Y?
Calibration/PV                                                         Y
Science Demonstration                                                  N
Routine Operations                                                     Y
Post Operations                                                        N

Actors:
CS: Calibration Scientist
IE: Instrument Engineer

Triggers:
   • Cal. plan
   • Hardware failure/degradation
   • Changes to OBS (e.g. tuning algorithm)

Preconditions:
   • Functional Common Science System

Minimal post conditions:
   • Required time series obtained and stored for analysis

Success post conditions:
  • determination of instrument response times

Stakeholders and interests:
   • Calibration Group


                                         17
                      HIFI calibration use cases                Doc.: ICC/2001-005
                                                                Date: 20/10/03
                                                                Issue: Draft 1.0


   •   General astronomers (via observing efficiency)
   •   Instrument System Engineer
   •   Integration and Test Team

Main success scenario:
Current belief is that the instrument response times are determined by:
   • UC-1.1.2.1: Chopper response time
   • UC-1.1.2.2: LO settling time for a LO sub-band change
   • UC-1.1.2.3: FPU settling time for a mixer band change (including diplexer
      movements and mixer settling)
   • UC-1.1.2.4: Frequency switch response time
   • UC-1.1.2.5: Hot load thermal response time

These are described in the child use cases. Whenever necessary other settling and
response times can be added as child use cases and included here.

Use case is completed when all child use cases are completed and the results are
inserted in the AOT logic and in the observing time estimator.

Frequency of occurrence:
   • once each during ILT, IST, PV
   • TBD interval during Routine Ops.

Open issues:

Comments:
This use case and its children describe the procedures used to determine various
instrument response times (dead times). These response times will be required in the
observing time estimator and the AOT. The optimum observing strategy is dependent
on these parameters. Periodic monitoring of (some of) these parameters during
routine operations will be desirable.




                                         18
                       HIFI calibration use cases                Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


2.2.2.1 UC-1.1.2.1 Determine Chopper Response Time

Level:                      User
Scope:                      HIFI-ICC
Version:                    0.1
Status:                     Draft
Author:                     N. D. Whyborn - 2001-09-20

General Grade: Medium
Grade for ILT: Medium
Grade for implementation: Medium

Brief description:
Determine response time of the focal plane chopper.

Phase:
ILT                                                                    Y
IST/EET/GST...                                                         Y
LEOP/Commissioning                                                     Y?
Calibration/PV                                                         Y
Science Demonstration                                                  N
Routine Operations                                                     Y
Post Operations                                                        N

Actors:
IE: Instrument Engineer
CS: Calibration Scientist

Triggers:
   • Cal. plan
   • Hardware failure/degradation
   • Changes to OBS (e.g. tuning algorithm)

Preconditions:
   • Working FP sub-system
   • Working TM System
   • On the ground: availability of a simulated point-like emitters rounded by
     absorbing material to chop against
   • In orbit: availability of isolated point-like source (e.g. asteroid)

Minimal post conditions:
   • Required time series obtained and stored for analysis

Success post conditions:
  • determination of chopper response times



                                         19
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


Stakeholders and interests:
   • Calibration Group
   • General astronomers (via observing efficiency)
   • Instrument System Engineer
   • Integration and Test Team

Main success scenario:
  1. IE: Initiate fast monitoring of chopper position via housekeeping.
  2. IE: Set chopper to position A
  3. IE: Wait TBD seconds while acquiring HK
  4. IE: Set chopper to position B
  5. IE: Wait TBD seconds while acquiring HK
  6. IE: Repeat the above TBD times to obtain statistics?
  7. IE: Repeat the above for different chopper positions corresponding to: sky-left,
      sky-right, cold calibration load, hot calibration load
  8. IE: analyse time series

Note: There may be two ways to address the chopper response time.
   • In the first case, one would chop a (point-like) target source (line signal in the
       lab) against an empty sky, or absorber in the lab, with a variable chopping
       delay w.r.t. the readouts time tags.
   • A second approach consists in spatial scanning through the target (in flight
       only) in a DBS-OTF mode. The ideal output is a double beam pattern
       consisting of a positive and a negative profile respectively separated by the
       beam throw angle. A delay in the chopper response would show up as a
       broadening in the profile wings on one side of each profile only, enhancing
       asymmetries (TBC).

Frequency of occurrence:
   • Once each during ILT, IST, PV
   • TBD interval during Routine Ops.

Open issues:

Comments:
The currently anticipated approach for the calibration source is to have a hot and cold
load.




                                           20
                       HIFI calibration use cases            Doc.: ICC/2001-005
                                                             Date: 20/10/03
                                                             Issue: Draft 1.0


2.2.2.2 UC-1.1.2.2 Determine LO Settling Time: LO Sub-band Change

Level:                      User
Scope:                      HIFI-ICC
Version:                    0.1
Status:                     Draft
Author:                     N.D. Whyborn - 2001-09-20

General Grade: High
Grade for ILT: High
Grade for implementation: Medium

Brief description:
Determine thermal response time of the LOU after LO sub-band change. This change
indeed implies switching on and off of sub-bands that result in thermal effects.

Phase:
ILT                                                                 Y
IST/EET/GST...                                                      Y
LEOP/Commissioning                                                  Y?
Calibration/PV                                                      Y
Science Demonstration                                               N
Routine Operations                                                  Y
Post Operations                                                     N

Actors:
IE: Instrument Engineer
CS: Calibration Scientist

Triggers:
   • Cal. plan
   • Changes to LO bias settings?

Preconditions:
   • Working TM System
   • Working LO System

Minimal post conditions:
   • Required time series obtained and stored for analysis

Success post conditions:
  • Determination of LOU thermal response time to sub-band change.

Stakeholders and interests
   • Calibration Group
   • General astronomers (via observing efficiency)


                                       21
                        HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                    Date: 20/10/03
                                                                    Issue: Draft 1.0


   •   Instrument System Engineer
   •   Integration and Test Team

Main success scenario:
  1. IE: Tune instrument to frequency in LO band Xa (X=1,2,3,4,5,6L,6H).
  2. IE: Select cold calibration load
  3. IE: Wait WW minutes for system to stabilise (this may take ~1 hour from cold)
      while making standard 8 sec dumps of spectra (both spectrometers, pols.).
  4. IE: Tune instrument to frequency in LO band Xb (same X as above).
  5. IE: Acquire standard 8 sec dumps of spectra over a period of MM minutes.
  6. IE: Tune instrument to frequency in LO band Xa (same X as above).
  7. IE: Acquire standard 8 sec dumps of spectra over a period of MM minutes.
  8. IE: Repeat the above for different mixer bands X.
  9. IE: Analyse the time sequence of spectra to determine the instrument settling
      time (see comment)

Generally, WW >> MM to make sure that earlier activities do not compromise the
instrument stability.

Alternate success scenario:
As above except that the methods of UC-1.1.3 (Allan variance) are used to determine
the time at which the instrument meets the HIFI stability requirements.

Frequency of occurrence:
   • once each during ILT, IST, PV
   • TBD interval during Routine Ops.

Open issues: What will be the spacing between Xa and Xb, can a table be
generated?

Comments:
Determining the settling time by the main method may involve some subjectivity! The
alternate method may give a more objective result, but perhaps take longer to
execute.

Note, there are two stability metrics in use for HIFI: total power (normal) Allan
variance and spectroscopic Allan variance.




                                           22
                       HIFI calibration use cases              Doc.: ICC/2001-005
                                                               Date: 20/10/03
                                                               Issue: Draft 1.0


2.2.2.3 UC-1.1.2.3 Determine FPU Settling Time - Mixer Band Change (tuning
        settling time)

Level:                      User
Scope:                      HIFI-ICC
Version:                    0.1
Status:                     Draft
Author:                     N.D. Whyborn - 2001-09-20

General Grade: High
Grade for ILT: High
Grade for implementation: Medium

Brief description:
Determine thermal response time of the FPU after mixer band change. This change
indeed implies switching on and off of bands that result in thermal effects.

Phase:
ILT                                                                 Y
IST/EET/GST...                                                      Y
LEOP/Commissioning                                                  Y?
Calibration/PV                                                      Y
Science Demonstration                                               N
Routine Operations                                                  Y
Post Operations                                                     N

Actors:
IE: Instrument Engineer
CS: Calibration Scientist

Triggers:
   • Cal. plan
   • Changes to FPU IF bias settings?

Preconditions:
   • Successful execution of UC-1.1.2.2 (LOU settling time)?
   • Working FP System

Minimal post conditions:
   • Required time series obtained and stored for analysis

Success post conditions:
  • Determination of FPU thermal response time to band change.

Stakeholders and interests:
   • Calibration Group


                                        23
                       HIFI calibration use cases                Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


   •   General astronomers (via observing efficiency)
   •   Instrument System Engineer
   •   Integration and Test Team


Main success scenario:
  1. IE: Tune instrument to frequency in mixer band X (X=1,2,3,4,5,6L,6H).
  2. IE: Select cold calibration load
  3. IE: Wait WW minutes for system to stabilise (this may take ~1 hour from cold)
      while making standard 8 sec dumps of spectra (both spectrometers, polars.).
  4. IE: Tune instrument to frequency in mixer band Y (Y != X).
  5. IE: Acquire standard 8 sec dumps of spectra over a period of MM minutes.
  6. IE: Tune instrument to frequency in mixer band X (same X as above).
  7. IE: Acquire standard 8 sec dumps of spectra over a period of MM minutes.
  8. IE: Repeat the above for different mixer bands X & Y.
  9. IE: Analyse the time sequence of spectra to determine the instrument settling
      time (see also comment).

Generally, WW >> MM to make sure that earlier activities do not compromise the
instrument stability.

Alternate success scenario:
As above except that the methods of UC-1.1.3 are used to determine the time at
which the instrument meets the HIFI stability requirements. This may turn out to be
the preferred scenario and then step 9 becomes: Calculate the total power and
spectroscopic Allan variances (see reference of UC-1.1.3) of the time sequence of
spectra.

Frequency of occurrence:
   • once each during ILT, IST, PV
   • TBD interval during Routine Ops.

Open issues:

Comments:
Determining the settling time by the main method may involve some subjectivity! The
alternate method may give a more objective result, but perhaps take longer to
execute.
Note, there are two stability metrics in use for HIFI: total power (normal) Allan
variance and spectroscopic Allan variance.




                                         24
                       HIFI calibration use cases                Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


2.2.2.4 UC-1.1.2.4 Determine Frequency Switch Response Time

Level:                      User
Scope:                      HIFI-ICC
Version:                    0.1
Status:                     Draft
Author:                     N.D. Whyborn - 2001-09-20

General Grade: High
Grade for ILT: High
Grade for implementation: High

Brief description:
Determine the response time to frequency switching within the same LO band, i.e. by
changing the setting of the LSU.

Phase:
ILT                                                                        Y
IST/EET/GST...                                                             N
LEOP/Commissioning                                                         N
Calibration/PV                                                             ?
Science Demonstration                                                      N
Routine Operations                                                         ?
Post Operations                                                            N

Actors:
IE: Instrument Engineer
CS: Calibration Scientist

Triggers:
   • ILT plan

Preconditions:
   • test signal source or celestial source available. Celestial source should provide
     narrow lines of order 10K (to allow for ~100 msec measurements), e.g. CO in
     PDRs.

Minimal post conditions:
   • Required time series obtained and stored for analysis

Success post conditions:
  • Determination of LO frequency switch response time.

Stakeholders and interests:
   • Calibration Group
   • General astronomers (via observing efficiency)


                                         25
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


   •    Instrument System Engineer
   •    Integration and Test Team

Main success scenario:
It is not clear that this can be measured post ILT since it may require test equipment.

   1.  IE: Set a test signal generator to provide a signal at the desired frequency.
   2.  IE: Tune the instrument to the desired frequency.
   3.  IE: Select a single sub-band of the HRS to minimise data transfer times.
   4.  IE: Wait TBD minutes for system to stabilise.
   5.  IE: Change LO frequency by about 50 MHz (TBC, this is 30 to 8 km/s from
       band 1 to 6H).
   6. IE: Acquire a short (<~100 ms) integration after a delay of τ ms.
   7. IE: Return to the original LO frequency.
   8. IE: Acquire a short (<~100 ms) integration after a delay of τ ms.
   9. IE: Repeat the above four steps for different values of τ ranging from -100 to
       +300 in steps of 10 ms.
   10. IE: Analyse the time sequence of spectra to determine the frequency settling
       time

Note: the analysis consists in checking the line shapes in the two phases and search
for broadening resulting in asymmetries in the two respective profiles, enhanced as
the switching delays increases.

Alternate success scenario:
As above except that an external counter or spectrum analyser is used to measure
the detected IF frequency.

Frequency of occurrence:
   • once each during ILT, IST, PV
   • TBD interval during Routine Ops.

Open issues:

Comments:
This UC may be dropped if the LSU is independently measured. Does the ICU OBS
allow this or is it only feasible using the Test Controller?




                                          26
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


2.2.2.5 UC-1.1.2.5 Determine Hot Load Thermal Response Time

Level:                      User
Scope:                      HIFI-ICC
Version:                    0.1
Status:                     Draft
Author:                     N.D. Whyborn - 2001-09-20

General Grade: Medium
Grade for ILT: Medium
Grade for implementation: Medium

Brief description:
Determine the thermal response time of the hot calibration load.

Phase:
ILT                                                                        Y
IST/EET/GST...                                                             Y
LEOP/Commissioning                                                         ?
Calibration/PV                                                             Y
Science Demonstration                                                      ?
Routine Operations                                                         Y
Post Operations                                                            N

Actors:
CS: Calibration Scientist
IE: Instrument Engineer

Triggers:
   • Cal. plan
   • Hardware failure/degradation?

Preconditions:
   • Working CSA
   • Working chopper

Minimal post conditions:
   • Required time series obtained and stored for analysis

Success post conditions:
  • Determination of hot load thermal response time.

Stakeholders and interests:
   • Calibration Group
   • General astronomers (via observing efficiency)
   • Instrument System Engineer


                                         27
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


   •   Integration and Test Team

Main success scenario:
  1. IE: Tune the instrument to the desired frequency.
  2. IE: Set the hot calibration load heater OFF
  3. IE: Select the hot calibration load
  4. IE: Wait WW minutes for system to stabilise (this may take ~1 hour from cold)
      while making standard 8 sec dumps of spectra (both spectrometers, pols.).
  5. IE: Set the hot calibration load heater ON
  6. IE: Acquire standard 8 sec dumps of spectra over a period of MM minutes.
  7. IE: Set the hot calibration load heater OFF
  8. IE: Acquire standard 8 sec dumps of spectra over a period of MM minutes.
  9. IE: Analyse the time sequence of spectra to determine the hot load thermal
      behaviour versus time.

Generally, WW >> MM to make sure that earlier activities do not compromise the
stability measurement.

Alternate success scenario:

Frequency of occurrence:
   • once each during ILT, IST, PV
   • TBD interval during Routine Ops.

Open issues:

Comments:
This use case could be repeated for the cold load to check thermal effect of the cold
load beam pick-up on the hot load and thus assess influence of the hot load on the
internal cold measurement. Instead of writing a dedicated UC for this purpose, we
include this check in UC-1.4.2.




                                          28
                       HIFI calibration use cases                   Doc.: ICC/2001-005
                                                                    Date: 20/10/03
                                                                    Issue: Draft 1.0


2.2.3 UC-1.1.3 Determine Instrument Intensity Stability

Level:                       User
Scope:                       HIFI-ICC
Version:                     0.1
Status:                      Draft
Author:                      N.D. Whyborn - 2001-10-07

General Grade: High
Grade for ILT: High
Grade for implementation: High

Brief description:
Determine the stability timescale for intensity fluctuations by use of Allan variance
measurements. Both the spectroscopic Allan variance and total power (i.e.
continuum) Allan variance will be measured. From the results it will be possible to
calculate the apparent loss of sensitivity due to instrument instability.

Phase:
ILT                                                                       Y
IST/EET/GST...                                                            Y
LEOP/Commissioning                                                        Y?
Calibration/PV                                                            Y
Science Demonstration                                                     N
Routine Operations                                                        Y
Post Operations                                                           N

Actors:
CS: Calibration Scientist
IE: Instrument Engineer

Triggers:
   • Calibration and AIV plans
   • Hardware failure/degradation
   • Changes to OBS (e.g. tuning algorithm)

Preconditions:
   • Availability of blank sky position (within the chop)
   • Common definition of Allan variance (cf RD33).

Minimal post conditions:
   • Set of spectra recorded

Success post conditions:
  • determination of spectroscopic Allan variance
  • determination of total power Allan variance


                                           29
                        HIFI calibration use cases                   Doc.: ICC/2001-005
                                                                     Date: 20/10/03
                                                                     Issue: Draft 1.0


   •   determination of instrument sensitivity loss due to instability

Stakeholders and interests:
   • Calibration Group
   • General astronomers (via observing efficiency)
   • Instrument System Engineer
   • Integration and Test Team

Main success scenario:
  1. IE: Go to a blank sky position (candidate positions TBD) or to the
      corresponding ground-based equipment (Absolute Hot Black Body, AHBB)
  2. IE: Tune instrument to a set of frequencies (value and number TBD). Then for
      each frequency:
  3. IE: Select both WBS and HRS in low resolution mode
  4. IE: Perform an intensity calibration using the internal calibration source (see
      RD20)
  5. IE: Make repeated 16 second integrations using the normal speed (0.5 Hz)
      focal plane chopping, the total integration time shall be 1 hour (TBC).
  6. IE: Make repeated 16 second integrations using the fast speed (2-5 Hz TBC)
      focal plane chopping, the total integration time shall be 1 hour (TBC).
  7. IE: Make repeated 8 second integrations with the focal plane chopper
      stationary in sky position 1, the total integration time shall be 1 hour (TBC)
  8. IE: Make repeated 8 second integrations with the focal plane chopper
      stationary in the cold calibration load position, the total integration time shall
      be 1 hour (TBC)
  9. IE: Calculate the spectroscopic and total power Allan variances for each case
      (see RD33)
  10. IE: Repeat the above for each frequency and mixer band
  11. IE: Note any systematic baseline offsets (this may require long integration
      times)
  12. IE: Calculate the rms. in the spectrum and compare with the theoretical value
      to determine the loss in sensitivity due to instability

Frequency of occurrence:
   • once each during ILT, IST, PV
   • TBD interval during Routine Ops.

Open issues:
  • identify candidate blank sky positions and optionally perform preparatory
      observations from the ground
  • determine the number of frequencies to be measured per mixer band
  • determine the required integration time taking into account the need to
      compare sub-integrations to detect systematic offsets

Comments:



                                           30
                        HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                    Date: 20/10/03
                                                                    Issue: Draft 1.0


The stability time-scale will be required in the AOT, and the loss of sensitivity due to
instability will be incorporated in the observing time estimator. The optimum
observing strategy is dependent on these parameters. Note that the stability is
expected to be a strong function of mixer band.

Step 5 will set tough requirements on the external calibration load during ILT and
may be impossible to perform on the ground. The AHBB provides a good option.

Periodic monitoring of (some of) these parameters during routine operations will be
desirable as a health check.

Binning of adjacent spectral channels should allow an accurate determination of the
level of instrument gain fluctuations and, therefore, their contribution to the total
power Allan variance.

Reference:




                                           31
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0




2.2.3.1 UC-1.1.3.1 The influence of Bragg cell heating on baseline stability of the
        WBS

Level:                      User
Scope:                      HIFI-ICC
Version:                    0.2
Status:                     Draft
Author:                     Csaba Gal & Carsten Kramer, 17.6.2002

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
This use case describes how to characterize (and monitor) the stability of the WBS
after measuring a strong RF signal. This leads to a transient heating of the Bragg
cell which may influence the baseline stability depending on the receiver
temperature, the observing mode and other factors.

Phase:
ILT                                                                     Y
IST/EET/GST...                                                          N
LEOP/Commissioning                                                      N
Calibration/PV                                                          N
Science Demonstration                                                   Y
Routine Operations                                                      N
Post Operations                                                         N

Actors:
CS: Calibration Scientist
IE: Instrument (System) Engineer

Triggers:
   • Calibration plan
   • Observers

Preconditions:
   • ILT: A white noise-source is connected to the temperature-stabilized
     spectrometer.

Minimal post conditions:
   • A series of spectra is taken in regular time interval of 1-4sec.

Success post conditions:



                                          32
                        HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                    Date: 20/10/03
                                                                    Issue: Draft 1.0


   •   Plot of the Baseline-RMS (Sig-Ref)/(Ref-zero) vs. frequency at different times
       after a large power change.

Stakeholders and interests:
   • Calibration group
   • Instrument System Group

Main success scenario:
  1. CS: Feed white noise (4-8GHz) into the WBS.
  2. CS: Choose the noise power level so high that the spectrometer is optimally
      driven (not saturated). This corresponds to measuring the hot load (SIG).
      Reduce power level by 3dB. This corresponds to the cold load or blank sky
      (REF) using a receiver with low receiver temperature. Take a series of
      measurements in time intervals of 1 to 4sec per spectrum.
      This results in a series of data: SIG, REF1, REF2, REF3, ... REFlast
  3. CS: Construct n spectra: (REFlast-REF1)/(REFlast-Zero), (REFlast-
      REF2)/(REFlast-zero), etc. The first spectra of this series are expected to
      show poor baselines due to heating of the Bragg cell when observing SIG. The
      last spectra of this series are expected to show the normal baseline behaviour.
  4. CS: Calculate the baseline-rms for each spectrum and plot the rms versus
      time elapsed after observing SIG.

Frequency of occurrence:
   • ILT: at least once

Open issues:
Bragg cell heating depends on the quality of the optical equipment. It is expected that
the flight model will have better optics than the DM.

Comments:
After calibration at low freq. bands where the power difference between hot and cold
load is of the order of 3dB, a time interval of typically 20sec (DM) is needed to wait in
order to thermalise the Bragg cell again so that no baseline effects are any longer
visible. The power level difference will be less for the upper HIFI bands, where the
receiver temperatures are higher (0.8dB).
Whether observations should still be taken during this time interval, depends on
which level of baseline distortion is acceptable. It also depends on the observing
mode. OTF-observations are more prone to be influenced than fast double beam
switched observations. To avoid any dead time, calibration on Hot and Cold Loads
may be done while slewing the telescope.

References:




                                           33
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


2.2.4 UC-1.1.4 Make Long Duration Integrations


Level:                  Summary & User
Scope:                  HIFI-ICC
Version:                0.2
Status:                 Draft
Author:                 Frank Helmich
Date:                   24-06-2002

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief Description:
In this use case a long duration observation is done and analysed. It is clear that
when integrating over a long period the radiometric noise decreases with the square
root of the integration time until systematic errors are encountered. Then the noise
levels off and at some point becomes constant.
First goal is to determine the point where noise becomes constant. Second is to
search for systematic patterns like residual standing waves and platforming. Every
observing mode has its own specifics that need to be studied. The residual patterns
will be used to study the effects.

Phase:
ILT                                                                      Y
IST/EET/GST...                                                           Y
LEOP/Commissioning                                                       Y
Calibration/PV                                                           Y
Science Demonstration                                                    Y
Routine Operations                                                       Y
Post Operations                                                          N

Actors:
CS: Calibration Scientist
IE: Instrument (System) Engineer

Triggers:
   • Calibration plan
   • Observers

Preconditions:
   • availability of adequate laboratory (black bodies, gas cell) and celestial
     sources

Minimal post conditions:



                                          34
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


   •   Set of required spectra is measured

Success post conditions:
  • See the noise remain constant as integration time increases.
  • See systematic effects show up (TBD what is expected here)
  • Platforming, standing waves?

Stakeholders and Interests:
   • AIV engineer
   • System engineer
   • Calibration Group
   • HIFi astronomers
This measurement gives the clearest figure for the capabilities of the instrument

Main success scenario:

   1. CS: Choose celestial source or Lab. Source. Note that long integrations pre-
      flight are essential (see Odin experience)!
   2. CS: Choose observing mode (may only be doable in-flight)
   3. CS: Choose critical frequencies (in common with AIV?)
   4. CS: Specify long integration time (maximum TBD) with regular (2 seconds
      TBD) intervals to be sent down.

Question: It is likely that two seconds are the practical minimum. Does it make sense
to go at the highest time-resolution possible, i.e. 1 second?

   5. CS: Collect the data set from the archive.
   6. CS: Do the basic calibration steps appropriate for each observing mode. There
       is a need for more complicated calibration schemes than simple referencing.
       The artefacts showing up in the long duration integrations may serve as a
       guideline on how to proceed with these improved schemes.
   7. CS: Whenever needed subtract a base line and remove bad channels.
   8. CS: There is a need for selecting different set of spectra for individual
       treatment. E.G.: spectrum 1,2, 3 and 4 need to be compared with spectrum
       21, 22, and 23. The CS needs to have the capabilities to select different (sub)-
       sets for comparison.
   9. CS: Sum each set and subtract baseline
   10. CS: Determine the noise (calculate the moments) as a function of integration
       time
   11. CS: Add more sets and repeat the noise determination
   12. CS: Plot noise as a function of Sqrt(integration time on-source)
   13. CS: Determine the spectroscopic and continuum Allan variance times of the
       data set
   14. CS: Plot each CCD line for WBS and correlator bank for HRS separately.
   15. Determine if platforming exists. When more series are available, spread in
       time, determine stability of the platforming.


                                          35
                       HIFI calibration use cases                   Doc.: ICC/2001-005
                                                                    Date: 20/10/03
                                                                    Issue: Draft 1.0


   16. CS: Determine the presence of standing waves. Try to fit sine-functions to
       these standing waves or perform FFT or wavelet analysis. Try to subtract the
       fitted sines. Determine phase, frequency and amplitude. Match frequency with
       optical path-lengths in the telescope. (data-base with all path-lengths must be
       available). See UC-1.1.6 for dedicated use case.
   Note: The scheme on determining the standing waves in the two side-bands is
   described in RD20.
   17. CS: File in calibration report. Critical number is of-course the time where
       further integration is senseless. Other important parameters are stability of
       platforming and standing wave parameters.

Open issues:
  • How many frequencies do we need to do? This will depend on the available
      test time.
  • Which frequencies are crucial for the calibration itself and possibly to the
      science? A list has been compiled in RD23.
  • What is the time frequency?
  • In what respect do we expect different modes to give similar characteristics?
      Or will these mainly be differences due to differing optical paths?

Comments:
It may be useful to use the output of this use-case (a large set of scans) with the use
case for Allan variance (UC-1.1.3).

Specifics per observing mode:

Fast Chop - staring
      The fast chop mode exists because the cold IF-amplifiers have an 1/f noise
      characteristic that prevents good observations for weak sources, like galaxies.
      It remains to be seen if the fast chop mode will really be implemented. Better
      behaviour of the IF-amplifiers may let this mode disappear.
      In this mode the chopper movement is fast, causing the optical paths to be
      slightly different for every timing. Purpose is to study the effect of the changing
      optical paths on the standing wave pattern. It is likely that varying residual
      patterns will be visible in the output spectrum. Also platforming is a likely
      artifact. This is a clear example where the result of varying optical paths will be
      most visible.

       Choose critical frequencies for galaxy lines, like the fine structure line [N II]
       and [C II]. Also the lowest water lines and all the CO lines are good candidate
       frequencies, since these are likely to be targets for this specific observing
       mode. There may be differences between point sources and beam-filling
       sources. The measurements should include both kind of sources. Since this
       mode is geared towards galaxy observations, they are the prime candidates.
       HRS observations for this mode are not necessary.




                                           36
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


Pointed observation slow chop
      The slow chop (0.5 Hz) mode is the normal mode of operation for the chopper.
      While the purpose of the chopper is to enable us to remove the normal
      standing wave patterns, the optical paths for the different chopper positions
      are slightly different. It is expected that these effects show up in a long
      duration integration. It is important to characterize this effect, understand its
      stability and (possibly) remove it automatically. While the fast chop mode may
      suffer from varying residual patterns, it is expected that in the slow chop mode
      the patterns are more stable. Second artifact is platforming. In a long duration
      integration, the amount and stability of the platforming can be studied.
      Moreover, this use case is a likely candidate to separate the standing waves in
      the different side bands.

      The critical frequencies may be chosen the same as for the fast chop mode,
      i.e. the fine-structure lines for galaxies, the lowest energy water-lines and the
      lower-energy CO-lines. There may be differences between point sources and
      beam-filling sources. The measurements should include both kinds of sources.
      Note that settings without lines can also be used. However, quality checking is
      much easier when at least one line is present in the IF-range.

Pointed observation frequency switch
      Observations in frequency switch are either made because no spectral line-
      free off-position is available or to conserve observing time (no integration time
      spend on an off-position). While the optical paths for the two LO-settings are
      the same, the same standing wave pattern is expected for the subreflector-
      mixer optical path. The LO-mixer standing wave is likely to be the most
      important. Artifacts to be expected, are the IF ripple and LO-mixer standing
      wave effects. As usual, platforming is to be expected. The amount of
      platforming and the stability of the platforming have to be established.

      The critical frequencies may be chosen the same as for the fast chop mode,
      i.e. the fine-structure lines for galaxies, the lowest energy water-lines and the
      lower-energy CO-lines. It is especially for these lines that the mode is
      expected to be used. There may be differences between point sources and
      beam-filling sources. The measurements should include both kind of sources,
      but preference is given to beam-filling sources. Note that settings without lines
      can also be used. However, quality checking is much easier when at least one
      line is present in the IF-range.

Pointed observation – Double beam switch
      The double beam switch mode is likely to be the most efficient in removal of
      standing waves and the best to study the radiometric noise equation. Due to
      the fact that there will be two OFF-positions and one ON, the symmetry will be
      used to remove the standing waves. The baseline is therefore as flat as
      possible. Any artifact left will be clear to see. This implies that study of the
      radiometric noise equation will be done best with this mode.



                                          37
                        HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                    Date: 20/10/03
                                                                    Issue: Draft 1.0


       The critical frequencies may be chosen the same as for the slow chop mode,
       i.e. the fine-structure lines for galaxies, the lowest energy water-lines and the
       lower-energy CO-lines. This mode is likely to be used mainly for weak lines,
       which occur at every frequency, so one is not limited to the suggestions
       above. There may be differences between point sources and beam-filling
       sources. The measurements should include both kind of sources, but
       preference is given to point sources,, since most science will be on point-like
       sources and since the maximum chopper throw is only 180 arcseconds. Note
       that settings without lines can also be used. However, quality checking is
       much easier when at least one line is present in the IF-range.

Pointed observation – Position Switch
      The position switched observations will be mainly done in areas on the sky
      where the emission is extended (larger than 180 arcseconds), and thus where
      emission on a OFF-position should be avoided. This mode will be used to
      discover faint emission, so careful understanding of any artifact is needed.

       The critical frequencies may be chosen the same as for the slow chop mode,
       i.e. the fine-structure lines for galaxies, the lowest energy water-lines and the
       lower-energy CO lines. This mode is likely to be used mainly for weak lines,
       which occur at every frequency, so one is not limited to the suggestions
       above. The measurements should include extended emission sources. Point-
       like sources may be used to compare results with the Double Beam Switch
       mode. Note that settings without lines can also be used. However, quality
       checking is much easier when at least one line is present in the IF-range.

Pointed observation – load chop
      This mode is very similar to the chopped mode, only now the chopper moves
      to the cold load for reference. The main difference arise from the lack of
      chopper symmetry between the two phases, which may lead to significant
      standing wave in the baseline if instabilities are of concern.

Mapping
     Because OTF maps and raster scans are mainly supersets (in position) of the
     modes sketched above, there is no need to make separate Long Duration Use
     Cases for them.

Spectral Surveys
      Spectral surveys take a lot of observing time, but this time is distributed over
      so many frequency settings that themselves only take a tiny fraction of the
      time. No long duration is ever expected per frequency setting. The effect of
      varying standing waves per frequency tuning must be resolved in the
      frequency switched case and in the double beam switch case.

Overlap with AIV:
In AIV/ILT there will be stability and linearity measurements. These will serve as
inputs for the long duration use case here. Since no observations can be mimicked in


                                           38
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


AIV/ILT a more detailed study on commonality between in-orbit calibration and ILT
detailed test plan is needed in this respect. One important exception is the load-chop
mode that may be tested on the ground using the external hot black body, although
M2 can not be mimicked.




                                          39
                      HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


2.2.5 UC-1.1.5 Measure platforming

Level:                       User
Scope:                       HIFI-ICC
Version:                     0.1
Status:                      Draft
Author:                      Laurent Ravera/Moncef Belgacem - 17/09/01
                             Updated D. Teyssier – 05/05/03

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
This use case describes the procedure to measure the platforming between WBS
CCD-lines or HRS banks

Phase:
ILT                                                                    Y
IST/EET/GST...                                                         Y
LEOP/Commissioning                                                     Y
Calibration/PV                                                         Y
Science Demonstration                                                  Y
Routine Operations                                                     Y
Post Operations                                                        N

Actors:
   • IE
   • CS

Triggers:
   • Cal. plan

Preconditions:
   • availabity of a noise source in the HIFI frequency range (for ILT and
     IST/EET/GST).

Minimal post conditions:
   • Identify where platforming effects show up for both backends

Success post conditions:
  • Complete platforming characteristics of both backends are assessed

Stakeholders and interests:
   1. Calibration Group
   2. General astronomers


                                         40
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


   3. Instrument System Engineer
   4. Integration and Test Team

Main success scenario:
  1. CS: Set the spectrometers so that respective sub-band jointures do not
      overlap in the IF domain (i.e. several HRS sub-band could fit within one single
      WBS band, while one HRS sub-bands could sample two contiguous WBS sub-
      band)
  2. CS: Do a long integration (20h) on a weak source. This spectra may be
      directly fetched from the long integration use case measurements (UC-1.1.4).
      On the ground, measurements are performed on the noise source or the
      external black body.
  3. CS: Display the contiguous calibrated + added spectra
  4. CS: Analyse platforming characteristics:
              a. Measure the steps (offsets) between two contiguous sub-bands for
                  both backends
              b. Measure the slopes between two contiguous sub-bands for both
                  backends
  5. Repeat the above measurements so that each backend sub-band is checked.

Frequency of occurrence:
   • Once during LEOP/Commissioning
   • TBD during calibration/PV
   • TBD during ILT, likely to be combined with long integration checks

Open issues:

Comments:
When doing this test "on the ground" the temperature has to be monitored (and may
be controlled) to verify that we are compliant with the 3°C/hour slope specification.




                                          41
                      HIFI calibration use cases               Doc.: ICC/2001-005
                                                               Date: 20/10/03
                                                               Issue: Draft 1.0


2.2.6 UC-1.1.6 Measure Baseline ripple

Level:          User
Scope:          HIFI-ICC
Version:        0.1
Status:         Draft
Author:         DT/NW - 08/31/01
                David Teyssier – Updates 02/14/02 and 05/05/03

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
This use case describes the procedure to measure, analyse and eventually cure the
baseline ripples observed/expected to affect the HIFI signals.

Phase:
ILT                                                                  Y
IST/EET/GST...                                                       Y
LEOP/Commissioning                                                   Y
Calibration/PV                                                       Y
Science Demonstration                                                Y
Routine Operations                                                   Y
Post Operations                                                      Y

Actors:
   • CS: Calibration Scientist
   • IE: Instrument (System) Engineer

Triggers:
   • Cal. plan
   • Observers
   • Contingency

Preconditions:
Both pre-launch and in-orbit
   • access to elementary integrations, differential signals on sky and loads in
      various combinations (IF) and eventually as seen before the spectrometers
      (HF), and to the calibrated signals independently.
   • simple formalism for baseline ripple behaviour to serve as a basis for the
      investigation (as e.g. expected frequencies and amplitudes)
   • knowledge of the exact optical design (like e.g. distance between mixer horns
      and reflective surfaces), as well as the location and orientation of any
      absorber/termination in the system.
   • knowledge of spurious signals (cf UC-1.3.3.1)


                                        42
                        HIFI calibration use cases                Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0



Minimal post conditions:
   • identify the sources feeding the ripples, the reflection paths involved and their
     amplitude (first-order approach)

Success post conditions:
  • understanding of the standing wave systems according to the observing
     conditions: where, when, how do they show up (mode and frequency
     dependence) ? What is known from SWAS? See DT report on the Odin
     baseline problems (slides of Calibration Meeting – March 2002, Paris)
  • Measure of ripple frequency(ies), amplitude(s) and phase(s) as a function of
     TBD
  • Data corrected from baseline ripple

Stakeholders and interests:
   • Calibration Group
   • General astronomers
   • Instrument System Engineer
   • System Group
   • Integration and Test Team

Main success scenario:

Pre-flight (preliminary investigation):

   1. CS/IE: Define a simple standing wave formalism for further analysis

       As a simple approach, we propose to consider the ripples as a combination of
       the reflection of given sources (external continuum, internal loads, receiver,
       LO) between TBD surfaces (e.g. sub-reflector, horns, loads). This first-order
       formalism does not take into account a complete Gaussian beam
       representation, and is described in details in RD24. In this approach, the
       standing wave is well constrained by a round-trip distance, reflection
       coefficients and feeding source powers.

       Dedicated modelling studies are also available which describe the behaviour
       of stranding waves in the specific cavity defined by the mixers and the LO
       feeds (RD25).

   2. CS/IE: Define pre-flight methodology for ripple analysis

       As a guideline, one can think of scanning small ranges of parameters such as
       LO frequency, LO power, temperature, mixer tuning parameters (currents,
       etc), mirror positions (chopper, Martin-Pupplet Interferometer). One also has to
       keep in mind the double side-band, as well as the polarized nature of the
       signals (H or V) to design eventual tests done on the diplexers.


                                          43
                        HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0



   3. CS: Perform measurement according to the varying parameters described
      above.

       Any integrated spectrum is in principle a candidate for the ripple analysis, so
       that there are no dedicated standing wave measurements as such. One need
       to get isolated total power signals from e.g. the various so-called "sky" and
       "load" paths. Exposure time shall be chosen to reproduce the expected noise
       rms level to be reached in scientific data.

       A discussion describing how to identify the origin of various baseline ripples in
       the HIFI spectra and their uncertainties is given in RD20. The method makes
       use of a load-calibrated OFF measurements performed on a blank sky position
       and of dedicated analysis for this purpose.

   4. CS: Analyse ripple behaviour with respect to current model and changed
      parameters
            a. Fit models according to measurement parameters
            b. If time series exist, analysis ripple stability with time, temperature
            c. Assess Allan variance time for ripple stability in a given resolution
               element (method TBD)

In-flight (further investigation and monitoring):


   1. CS: Retrieve interesting measurements (signal exhibiting "significant" baseline
      ripple) from the database
      Note: I consider here only obviously affected spectra because a systematic
      baseline correction from the ripple may not be applied at the early stage of the
      mission (unless a satisfying model is quickly or already available).
   2. CS: if available: compute expected ripple models for the type of measurement
      considered (e.g. loads, sky, etc) and fit it to the total power or the calibrated
      data.

   3. CS: Identify standing wave contribution to the signal, e.g.
      • assess plausible frequencies (FFT, model fitting, etc...)
      • measure standing wave amplitude (from e.g. previous fit)
      • assess origin of ripple amplitude and ideally compute coupling efficiencies

   4. CS: correct from standing wave
            • if judged valuable (criterion TBD) correct data baseline from
                modelled ripple (see e.g. "baseline correction" in RD15).
            • Alternative: use dedicated OFF measurements described in RD20 to
                calibrate the standing wave (to first order, fitting of ripples is not
                mandatory here as this step is mainly a cure to the signal).

   5. CS: File in a calibration report


                                            44
                      HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0



Frequency of occurrence:
IN PV and on any affected data

Open issues:
Dedicated and proven methodology for the investigation and analysis.

Comments:
  • The topic discussed in this UC have been considered in two steps: first it
    contains a preliminary investigation approach on an instrumental effect not
    entirely understood at the moment, and second, presents the measurements
    dedicated to monitoring or systematic determination of this effect in order to
    optimally cure it. For both concerns we proposed a draft scenario.
  • It turns out that some of the ripples observed in the Odin baseline had not
    been detected on the ground due to insufficient integration time.
  • Results of this UC will have implications for the band-pass calibration and the
    optimisation of operational procedures (regarding e.g. undesired parameters
    tracking during observing, etc?). Iteration with the description of RD20 is thus
    expected.




                                         45
                        HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                    Date: 20/10/03
                                                                    Issue: Draft 1.0


2.2.7 UC-1.1.7 Measure Continuum emission

Level:                           User
Scope:                           HIFI-ICC
Version:                         0.1
Status:                          Draft
Author:                          Michel Pérault - 1-10-2001

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
This use case describes how to determine the capability of HIFI to estimate the
continuum emission from astronomical sources, and the accuracy of such
determinations.

Phase:
ILT                                                                        Y
IST/EET/GST...                                                             N
LEOP/Commissioning                                                         Y
Calibration/PV                                                             Y
Science Demonstration                                                      Y
Routine Operations                                                         Y
Post Operations                                                            Y

Actors:
   • CS: Calibration Scientist
   • IE: Instrument (System) Engineer

Triggers:
   • Cal. plan
   • Observers

Preconditions:
   • in ILTs: availability of a wide enough source to allow use of the chopper
   • in orbit: availability of double beam switching for point sources.
   • Availability of wide angle position switching (or line scans) for extended
     sources.
   • Access to absolutely calibrated photometry from SPIRE, as well as source
     models. SPIRE has very non-linear detectors and its FTS is a complicated
     instrument part. It may not be possible to use SPIRE data for absolute
     (spectro-)photometry.
   • in orbit: availability of a non reflective cold load, providing the dark reference
     for direct (non chopped) measurements.



                                           46
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


Minimal post conditions:
   • determination of sensitivity to continuum.
   • determination of photometric accuracy for continuum measurements.

Stakeholders and interests:
   • Calibration Group
   • General astronomers

Main success scenario(s):

Pre-flight (preliminary investigation):
   1. CS/IE: try beam switch on a large flat field (Absolute Hot Black Body) source
        input to sky port of HIFI. Do long integrations with chopper rate of 3.5Hz and
        TBD frequency binning (note that 100MHz bandwidth is the max. at this rate).
   2. CS/IE: determine offset and baseline (perform adequate spectral fit).
   3. CS/IE: rotate the source by 180 degrees around the chopping axis of
        symmetry (Note: in the current setup, this is impossible to do without breaking
        the vacuum). Repeat measurement and derivation.
   Note: This is the only use case taking into account polarization behaviour. We
   would have to scan all other use cases for polarization. That is not to say that we
   want polarization observations for the general astronomer, but we need to know
   how much every polarization is adding to the total signal.
   4. CS/IE: integrate as in step 1.
   5. CS/IE: compare both: if not exactly opposite, the sum, normalized by the
        source intensity gives an estimate of the relative error on continuum
        measurements in this optical setup.
   6. CS/IE: repeat for statistics. Vary intensity if possible (likely not !).

Pre-flight (assessment of indirect (beam switch) or direct (load-chop) measurement):
   1. CS/IE: calibrate the internal loads, and coupling efficiencies with an absolute
        external black body (UC's 1.4.1 and 1.4.2 - pre-flight setup).
   2. CS/IE: Use now external Absolute Hot Black Body as a known continuum
        source (expected temperature is 80K).
   3. Calibrate with internal loads.
   Note: the internal loads are by essence preliminary calibrated against the AHBB.
   4. Integrate (3.5Hz chopping rate, binning TBD) in both beam switching and
        load-chop mode (chop against internal cold load)
   5. CS/IE: analyse the data (through baseline fitting and noise assessment)
               a. Beam-switch measurements should give zero continuum (although
                  likely ripple-modulated) if the chopper is symmetric (the Absolute
                  Hot Black Body aperture should be homogeneously emitting to
                  better than the required accuracy)
               b. Load-chop measurements should give a calibrated measurement of
                  the absolute hot black body radiation temperature and thus provide
                  the continuum detection threshold and accuracy. Since the internal
                  cold load is preliminary calibrated against the AHBB, this check
                  actually addressed pre-flight the capacity of the load-chop mode to


                                          47
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


                 measure continuum and its sensitivity to baseline and stability
                 constraints.

Comment: this is very depending on the level, and variation of stray light in the set-
ups considered.

In-flight (further investigation and monitoring):
    1. CS: Perform beam-switched, load-chop, position-switched or line scan
        observations of sources of known continuum intensity. High S/N and
        compactness (for DBS at least) required, so solar system bodies (planet and
        especially asteroids) are good candidates.
    2. CS: Process the data, but do not remove baselines !
    3. CS: Compare output (offset and baseline shape) with expected flux and SED
        (spectral energy distribution). What should be in the data base. How many
        entries, their values and uncertainties ?

In-flight (a posteriori assessment of direct measurement):
    1. CS: Use each phase ON-source of the above data independently against the
        cold internal reference.
    2. CS: Compare the calibrated output to expectation, and to the switched
        measurement.

Frequency of occurrence:
   • ILT: minor issue -- set-up may not allow accurate assessment
   • PV: major issue. But difficult. Any ideas on how to resolve it?
   • Routine: check every month (only commissioned modes) for trends. Could
     provide hints on instrumental degradation (loads, stray light, mid-term stability)

Open issues:
  • Availability of a high quality cold internal reference to serve as a reference for
      direct continuum measurements of extended sources.
  • Capability to use SPIRE calibrated measurements as absolute references.
      Asteroids, AGB stars and normal stars ?

Comments:
Observations triggered by other UC's can be (and should be) used for this purpose.
Processing is different.




                                           48
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


2.3   UC 1.2 Beam Properties

Level:                           Summary
Scope:                           HIFI-ICC
Version:                         0.1
Status:                          Draft
Author:                          Douwe Beintema - 19/09/01


General Grade: High
Grade for ILT: Medium
Grade for Implementation: Medium

Brief description:
Accurate pointing and accurate pointing knowledge are essential for successful HIFI
measurements. HIFI is designed to operate at diffraction-limited resolution and it is
likely that the pointing capabilities of the Herschel spacecraft are marginal with
respect to the HIFI requirements. The calibration steps that the HIFI ICC must take to
assure optimal pointing with the HIFI beams are outlined in UC-1.2.1 and underlying
UC's, which address the location of the different HIFI beams in the Herschel focal
plane, taking into account several chopping positions.

The beam shapes are needed for qualitative interpretation of the measurements.
Mapping of the beam shapes is addressed in UC-1.2.2 and the scientific background
associated to it is described in RD21. UC-1.2.3 addresses pointing performance.

Phase:
ILT                                                                     Y
IST/EET/GST...                                                          N
LEOP/Commissioning                                                      Y
Calibration/PV                                                          Y
Science Demonstration                                                   N?
Routine Operations                                                      Y
Post Operations                                                         N

Actors:
   • CS: Calibration Scientist
   • CO: Calibration Operator (?)
   • CM: Calibration Manager
   • MP: Mission (or measurement) Planner
   • IE: Instrument (System) Engineer
   • AS: Astronomer
   • DP: Data Processing pipeline
   • TE: Test Engineer

Triggers:


                                         49
                          HIFI calibration use cases               Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


   •   Calibration plan
   •   AIV plan

Preconditions:
   • It is essential to have a clear definition of the co-ordinate systems involved.
     The Herschel focal-plane co-ordinates are Y and Z are distances (meters)
     from the telescope boresight. These co-ordinates are part of the satellite co-
     ordinate system. It is usual to translate these distances into celestial angles,
     just by applying the plate scale (dividing by the telescope focal length) and
     leaving the names the same. A confusing situation: a Y shift of +1 mm is the
     same as a Y shift of +7.24 arcsec or a rotation around Z of -7.64 arcsec, a
     positive Z shift will cause a positive Y rotation.

Minimal post conditions:
   • Y and Z positions for all mixer beams in the telescope focal plane with an
     accuracy of less than 1.5 arcsec or 7% of the beam width. What is your
     preference? The smallest of the 2?
   • ditto for at least one mixer in each band, at a different chopper position
   • pointing error statistics per mixer band from HIFI observations
   • most observations free of pointing problems

Success post conditions:
  • Y and Z beam positions to within 1 arcsec, for at least TBD frequencies per
     L.O. sub-band
  • ditto for at least one mixer in each band, for at least one frequency in one of
     the L.O. sub-bands, at a different chopper position
  • pointing error statistics per mixer band from HIFI observations
  • beam width determinations for all mixers, in Y and Z, with an accuracy of 3%,
     for at least one frequency per L.O. sub-band
  • other beam shape parameters TBD
  • no pointing problems at all

Stakeholders and interests:
   • Calibration Group
   • Instrument System Group
   • Herschel System Group
   • General astronomers
   • Integration and Test Team

Main success scenario:
  • Use case 1.2.1 and lower-level UC's will provide Y and Z positions of mixer
      beams at various chopper positions
  • Use case 1.2.2 deals with beam-shape calibrations
  • Use case 1.2.3 should show that we do not have pointing problems

Frequency of occurrence:


                                          50
                     HIFI calibration use cases               Doc.: ICC/2001-005
                                                              Date: 20/10/03
                                                              Issue: Draft 1.0



Open issues:
  • Required number of frequencies for mixer beam positions

Comments:
  • Beam parameters are to be found in RD21




                                      51
                        HIFI calibration use cases               Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


2.3.1 UC 1.2.1 HIFI Focal-Plane Geometry

Level:                           Summary
Scope:                           HIFI-ICC
Version:                         0.1
Status:                          Draft
Author:                          Douwe Beintema - 19/09/01

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
This use case deals with the HIFI Focal-Plane geometry, i.e. the X and Y positions of
the 14 different mixer beams, for at least 2 different chopper positions.

Phase:
ILT                                                                       Y
IST/EET/GST...                                                            N
LEOP/Commissioning                                                        Y
Calibration/PV                                                            Y
Science Demonstration                                                     N
Routine Operations                                                        Y
Post Operations                                                           Y

Actors:
   • CS: Calibration Scientist
   • CO: Calibration Operator (?)
   • CM: Calibration Manager
   • MP: Mission (or measurement) Planner
   • IE: Instrument (System) Engineer
   • AS: Astronomer
   • DP: Data Processing pipeline
   • TE: Test Engineer

Triggers:
   • Calibration plan
   • AIV plan

Preconditions:
   • Control of co-ordinate-systems (see UC-1.2)
   • Chopping mechanism pre-calibrated: angle as a function of read/out (closed-
     loop operation) and of drive current (open loop)

Minimal post conditions:



                                         52
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


   •   Y and Z positions for all mixer beams in the telescope focal plane, with an
       accuracy of less than 1.5 arcsec or 7% of the beam width, Again the smallest?
   •   Y and Z positions for at least one mixer in each band, with an accuracy of less
       than 1.5 arcsec or 7% of the beam width, for a different chopper positions

Success post conditions:
  • Y and Z beam positions to within 1 arcsec, for all mixers and for all at least
     TBD frequencies per L.O. band
  • Y and Z beam positions to within 1 arcsec at a different chopper position, for at
     least 1 mixer in each mixer band


Stakeholders and interests:
   • Calibration Group
   • Instrument System Group
   • Herschel System Group
   • General astronomers
   • Integration and Test Team

Main success scenario:
  • Use case 1.2.1.1 deals with the HIFI focal-plane per se, as determined in
      ground tests
  • Use case 1.2.1.2 deals with the HIFI focal-plane geometry in exploration
      during the LEOP/Commissioning phase
  • Use case 1.2.1.3 deals with the finest determination of the HIFI focal-plane
      geometry during the Calibration/PV phase and eventually routine operation

Frequency of occurrence:

Open issues:
  • Required number of frequencies for mixer beam positions

Comments:




                                          53
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


2.3.1.1 UC-1.2.1.1 HIFI Focal-Plane Geometry, ILT

Level:                            User
Scope:                            HIFI-ICC
Version:                          0.1
Status:                           Draft
Author:                           Douwe Beintema - 20/09/01

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
This use case describes the procedure to measure the beam positions on the ground
to determine the expected location of the HIFI mixer beams in the Herschel focal
plane, for several chopper positions. The geometry is defined in linear focal-plane co-
ordinates Y and Z calculated for the nominal position of HIFI in the Herschel
telescope focal plane.

Phase:
ILT

Actors:
   • CS: Calibration Scientist
   • TE: Test engineer

Triggers:
   • Cal. Plan
   • AIV plan

Preconditions:
   • Availability of beam scanning equipment
   • Detailed plan to characterise beam positions (frequencies, chopper positions)

Minimal post conditions:
   • Beam positions (Y and Z, nominal linear scale) for all mixers, measured at at
     least one frequency per LO band?
   • Beam positions (Y and Z, nominal linear scale) from at least one of the mixers,
     at a second chopper position


Success post conditions:
  • Beam positions (Y and Z, nominal linear scale) for all mixers and L.O. sub-
     bands, for at least TBD frequencies per L.O. sub-band, for at least one
     chopper position



                                          54
                       HIFI calibration use cases                Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


   •   Beam positions (X and Y, nominal linear scale) from at least one
       measurement in each mixer band, at a second chopper position

Stakeholders and interests:
   • Calibration Group

Main success scenario:

   1. TE: Set-up laboratory source (beam scanner)
   2. CS: Select LO sub-bands, mixers, frequency setting, chopper positions (if only
      one, telescope axis is preferred)
   3. TE: Apply settings
   4. Perform measurements:
             a. scan beam profiles
             b. determine beam positions (position at the peak)
   5. CS: Store settings and Y and Z values in calibration data-base
   6. CS: File in a calibration report

Frequency of occurrence:
In ILT.

Open issues:
  • Number of measurements per LO band
  • Optimal sizes for rasters
  • Needed QLA applications ?

Comments:
  • These measurements will be part of the FPU S/S level AIV at room and
    cryogenic temperatures (cf RD18, RD19)




                                         55
                       HIFI calibration use cases                Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


2.3.1.2 UC-1.2.1.2 HIFI Focal-Plane Geometry, Commissioning Phase

Level:                           User
Scope:                           HIFI-ICC
Version:                         0.1
Status:                          Draft
Author:                          Douwe Beintema - 20/09/01

General Grade: High
Grade for ILT: Medium
Grade for Implementation: Medium

Brief description:
This use case describes the procedure to link the HIFI geometry to the
telescope/AOCS configuration. For this purpose one or a few relative large rasters
have to be executed to measure the alignment between at least one HIFI channel
and the telescope bore-sight. At this stage, these measurements will be subject to
more uncertainty than UC-1.2.1.3. These will be the first-light measurements of HIFI.

Phase: LEOP/Commissioning

Actors:
CS: Calibration Scientist

Triggers:
   • Commissioning Plan


Preconditions:
   • UC-1.2.1.1 completed, results converted to angular offsets
   • Availability of strong point continuum sources
   • Detailed plan to characterise beam positions (frequencies, chopper positions)

Minimal post conditions:
   • Required mapped successfully performed

Success post conditions:
  • Preliminary beam positions (Y and Z, angles wrt telescope boresight) for all
     mixer beams at at least two chopper positions.

Stakeholders and interests:
   • Calibration Group
   • Flight Dynamics, Mission Planning

Main success scenario:



                                         56
                      HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


   1. CS: Select source from database. Asteroids are anticipated as best
       candidates as planet may be either too large or not accurately enough known.
       Line sources can also be envisaged (AGB stars, see RD34).
   2. CS: Choose LO sub-band and Mixer, frequency setting, chopper position, no
       beam switching
   3. CS: Select suitable small raster (~ 11 by 11, step-size ~ 1/4 FWHM)
   Note: OTF could also be considered but one should bear in mind that the bore-
   sight definition may not be accurate enough for this mode (at least before this UC
   is fulfilled). Adequacy or not of OTF compared with raster could be addressed by
   UC-1.5.2.
   4. CS: Submit raster programme to Mission Planning, wait until done
   5. CS: Run Interactive Analysis. Very likely identify centre positions of beams
       and assess a first order link with the telescope bore-sight as defined by the
       AOCS system
   6. CS: Store settings and Y and Z values in calibration data-base
   7. CS: Compute and store predicted Y and Z positions for all other mixers and
       chopper/LO/frequency combinations. Preliminary quasi-optical simulations
       may help in that sense (RD27)
   8. CS: File in a calibration report

Frequency of occurrence:
~ 3 measurements in the commissioning phase

Open issues:
  • Number of measurements per LO band
  • Optimal sizes for rasters
  • Needed QLA applications ?




                                         57
                      HIFI calibration use cases                Doc.: ICC/2001-005
                                                                Date: 20/10/03
                                                                Issue: Draft 1.0


2.3.1.3 UC-1.2.1.3 HIFI Focal-Plane Geometry, In-Orbit Calibration

Level:                           User
Scope:                           HIFI-ICC
Version:                         0.1
Status:                          Draft
Author:                          Douwe Beintema - 20/09/01

General Grade: High
Grade for ILT: Low
Grade for Implementation: Medium

Brief description:
This use case describes the procedure to determine the location of the HIFI mixer
beams in the Herschel focal plane, for several chopper positions (3 positions for at
least one mixer band). The geometry is defined in angular focal-plane co-ordinates Y
and Z. Give reference even if it is for the umptieth time

Phase:
ILT                                                                      N
IST/EET/GST...                                                           N
LEOP/Commissioning                                                       N
Calibration/PV                                                           Y
Science Demonstration                                                    N
Routine Operations                                                       Y
Post Operations                                                          N

Actors:
   • CS: Calibration Scientist

Triggers:
   • Cal. Plan
   • AIV plan

Preconditions:
   • UC-1.2.1.1 completed, as a starting point for further refinement
   • UC-1.2.1.2 completed, as a first order link with the telescope bore-sight as
     defined by the AOCS system
   • detailed plan to characterise beam positions (frequencies, chopper positions)

Minimal post conditions:
   • Required mapped successfully performed

Success post conditions:




                                        58
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


   •   Beam positions (Y and Z, angles telescope bore-sight) for all mixers and L.O.
       sub-bands, for at least TBD frequencies per L.O. sub-band, for at least one
       chopper position
   •   Beam positions (Y and Z, nominal linear scale) from at least one measurement
       in each mixer band, at a second chopper position


Stakeholders and interests:
   • Calibration Group

Main success scenario:
  1. CS: Select source from database. Asteroids are anticipated as best
      candidates as planet may be either too large or not accurately enough known.
      Line sources can also be envisaged (AGB stars).
  2. CS: Choose LO sub-band and Mixer, frequency setting, chopper position, no
      beam switching
  3. CS: Select suitable small raster (~ 5 by 5, step size ~ 1/4 FWHM)
  4. CS: Submit raster programme to Mission Planning, wait until done
  Note: OTF could also be considered but one should bear in mind that the bore-
  sight definition may not be accurate enough for this mode (at least before this UC
  is fulfilled). Adequacy or not of OTF compared with raster could be addressed by
  UC-1.5.2.
  5. CS: Run Interactive Analysis. Likely identify centre position of beams and
      relate to AOCS system.
  6. CS: Store settings and Y and Z values in calibration data-base
  7. CS: File in a calibration report

Frequency of occurrence:
In PV phase, likely to be continued during calibration runs in the routine phase.

Open issues:
  • Number of measurements per LO band
  • Optimal sizes for rasters/OTF
  • Needed QLA applications ?




                                          59
                      HIFI calibration use cases                Doc.: ICC/2001-005
                                                                Date: 20/10/03
                                                                Issue: Draft 1.0


2.3.2 UC-1.2.2 Beam Patterns

Level:                     User
Scope:                     HIFI-ICC
Version:                   0.1
Status:                    Draft
Author:                    Frank Helmich/David Teyssier - 31/04/01
                           Douwe Beintema - 20/09/01
                           David Teyssier – 05/05/03

General Grade: High
Grade for ILT: Low
Grade for Implementation: Medium

Brief description:
This use case describes the procedure to perform and analyse the main beam
pattern characteristics (including symmetry and first order side-lobes).

Phase:
ILT                                                                  N
IST/EET/GST...                                                       N
LEOP/Commissioning                                                   N
Calibration/PV                                                       Y
Science Demonstration                                                N?
Routine Operations                                                   Y
Post Operations                                                      N

Actors:
   • CS: Calibration Scientist

Triggers:
   • Cal. Plan

Preconditions:
   • Added by DB: Control of co-ordinate systems (know how to project beam
     asymmetries onto the sky).
   • Knowledge about the AOCS capabilities
   • Availability of source suited for measurement (data-base with sources
     required)
   • Successful execution of UCs 1.2.1.1 and 1.2.1.2

Minimal post conditions:
   • A map of the target with the requested signal-to-noise ratio

Success post conditions:



                                         60
                       HIFI calibration use cases                   Doc.: ICC/2001-005
                                                                    Date: 20/10/03
                                                                    Issue: Draft 1.0


   •   Maps of beam intensities in angular focal plane coordinates, with and without
       beam switching
   •   Accuracy 20 dB (TBD, higher accuracy relative than absolute. This will be a
       mix of noise rms and pointing jitter errors. Note that the required accuracy
       highly depends on where we expect the first order side-lobes)
   •   Exact number for Half Power Beam Width
   •   Measures for deviations from Gaussicity

Stakeholders and interests:
   • Calibration Group
   • General astronomers

Main success scenario:

   1. CS: Select source from database (likely driven by source availability) and
      retrieve required parameters (apparent size θp and eventually the expected
      brightness temperature Tp as seen in the DSB system). For a discussion on
      the anticipated candidates (planets, asteroids) see RD21 and RD26. Note that
      here the absolute intensity is NOT of interest.

   2. CS: Choose LO-band and Mixer(s)

   3. CS: Determine size of maps, based on observing frequency, observing mode,
      as well as their model flux in the image and signal side-bands.

   4. Note: RD26 proposes map sizes depending on the side-lobe level to be
      mapped: 3 arcmin (1 arcmin) and 5 arcmin (3 arcmin) are needed to mapped
      down to 20dB and 30dB resp. at 500 GHz (resp. 1900 GHz).

   5. CS: Perform the observations: the observing mode can be either an OTF
      map or OTF strips across the source (see further discussion in RD21). In order
      to get the best stability required for continuum observation considered here,
      we may want to combine OTF mapping with beam switching modulation.

       Note:
          • we assume that a large enough OTF map will not suffer much from (not
             too significant) pointing errors, so that slight offsets should not preclude
             the measurement success. Also note that a preliminary peak-up (5-pts)
             check would not give more information on the achieved pointing that the
             map itself.
          • beam-switched point rasters or line rasters should have their scanlines
             parallel to the chopping (Z) direction.

       Note (DB): Possible modes:
          • raster scan with fast chopping (to observe differential profile, including
             any smearing through the fast chopping.


                                           61
                     HIFI calibration use cases                   Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


       •   raster scan with slow chopping (clean differential measurement)
       •   line raster with fast chopping (to observe differential profile, including
           any smearing through the fast chopping). This is the option described
           here.
       •   line raster without chopping (OTF)

6. CS: Start Interactive Analysis
7. CS: Create Map (final rebinning to be assessed)
8. CS: Compute 2-D HPBW: measure size of the -3dB level along two orthogonal
   directions or fit a simple Gaussian (good approximation for main beam
   diffracted pattern). One could e.g. compute the HPBW along the telescope
   axis, but also along an eventual elongated feature. If the measured HPBW are
   θx and θy, then the deconvolved HPBW assuming Gaussian beams are:
    HPBWx = θ x2 −θ p
                    2     HPBWy = θ y2 −θ p
                                          2
                      and
9. CS: Assess level of first-order side-lobes: depending on the reached noise
   level and or side-lobe level:
             • Either put an upper limit on side-lobe level (if not visible)
             • Or try a fit of the diffracted main+side-lobe beam invoking
                 standard diffraction theory and/or knowledge of the beam from
                 the ground. See RD21 for theoretical equations.

       Note: We very likely have to add these (and other) particular fitting to the
       IA tool requirements

10. CS: From the previous fit, estimate (x,y) offsets with respect to map centre.
    Also assess deviation from Gaussicity or any other expected feature.

   Note1: we need to decide what we call Gaussicity. In quasi-optical system, this
   is the coupling of the actual measured beam to the expected perfect Gaussian
   beam (taken in the far-field). In the general case, this is the normalised inner
   product between the beam distribution pattern and the perfect Gaussian beam
   pattern in the angular domain. Particular case is given in Eq. 4.18 of Goldsmith
   98.

   Note2: if the S/N is good enough, one might be able to detect eventual coma
   effect (error due to a misaligned sub-reflector, shifted perpendicular off the
   main reflector axis) or astigmatism (mechanical and/or thermal deformations
   on the surface).

11. CS: Check beam map against theoretical predictions obtained through quasi-
    optical simulations of the instrument (see RD27 for a description of the
    expected simulation work-packages). Assess discrepancies. If necessary, re-
    run dedicated simulations.

12. CS: return maps and parameters to calibration data-base
13. CS: File in a calibration report


                                        62
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0



Frequency of occurrence:
In PV and according to needs provided a suitable planet is visible.

Open issues:
  • what is the largest acceptable source apparent diameter ?
  • expected side-lobe level and extent ? See RD21 for first estimates
  • what is the size of the main beam (i.e. to which power level do we set it) ? A
      conservative approach is the full width between the first nulls.

Comments:
We should work out the software requirements for IA in more detail.


References:

Goldsmith P., 1998, ”Quasi-optical Systems”, IEEE Press




                                          63
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


UC-1.2.3 Pointing performances

Level:                      User
Scope:                      HIFI-ICC
Version:                    0.1
Status:                     Draft
Author:                     Frank Helmich 20 Oct 2003

General Grade: High
Grade for ILT: Low
Grade for Implementation: Medium

Brief description:

Pointing is a serious concern for the HIFI instrument. Being a single-pixel instrument,
HIFI is very sensitive to errors in the absolute pointing. Since the pointing of the
Herschel Space Observatory is not sufficient in its nominal value and barely sufficient
in its goal value, strategies to actively obtain pointing corrections are necessary.
There are two options to do so. First is to use a thermo-physical modelling of the
movements between the star-tracker and the PACS array, second is actively
obtaining pointing offsets. The first method will be described, but the second is most
within our control. Note that the pointing problems are most severe in Band 6, where
HIFI's beam is smallest. The third scenario is a placeholder for the automatic
peaking-up.

Phase:
ILT                                                                     N
IST/EET/GST...                                                          N
LEOP/Commissioning                                                      N
Calibration/PV                                                          Y
Science Demonstration                                                   Y
Routine Operations                                                      Y
Post Operations                                                         N

Actors:
   • CS: Calibration Scientist

Triggers:
   • Cal. Plan

Preconditions:
   • HIFI focal plane geometry known in-orbit, see UC 1.2.1.3
   • Beam pattern, see UC 1.2.2
   • These two are necessary to connect the mixer positions to the sky position,
     i.e. where does the mixer-beam hits the sky.
   • A data-base of pointing sources, both spectral and continuum



                                          64
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0



Minimal post conditions:
   • Accurate knowledge of pointing (within 2 TBC arcsec for 2 TBC hours) in Band
     6. For the other bands the a posteriori Herschel pointing knowledge is likely
     sufficient

Success post conditions:
  • Accurate knowledge of pointing (within 2 TBC arcsec for 22 TBC hours) in
     Band 6.

Stakeholders and interests:
   • Calibration Group
   • General astronomers

Main success scenario(s):
The 3 sketched scenarios below are likely all to be used in parallel !

Scenario 1:

Description:
Herschel uses a star-tracker to measure a field in which several stars are present.
From that field and a sky atlas (Hipparchos?) the absolute direction of the star-tracker
can be determined. This field can then be related to the PACS focal plane array in
order to determine how the star-tracker field compares to the PACS image. However,
due to differences in solar illumination, the correspondence between the two fields is
subject to change. This change will be monitored with PACS, the star-tracker and
with temperature sensors on the structure. These temperature data will be used as
input for models and compared with actual measurements of PACS and the star-
trackers. Although this method is supposed to be rather accurate there are some
drawbacks. First of all it is unknown for how long the accuracy of the model holds and
secondly, there is only ground-knowledge on the position of the HIFI beams with
respect to the PACS arrays. Note that for band 1 through 5 the a posteriori pointing
knowledge of Herschel is likely to be sufficient for our purposes. For band 6 pointing
calibration is mandatory.

Scenario 2:

       CS: take source from data-base (RD34)
       CS: Take CO J=6-5 size from data-base and calculate the CO J=16-15 (TBC)
strength from simple model. Measurements should be possible within 10 minutes
(TBC), so the strength of the line must be sufficient (note that, in Double Beam
Switch mode, 1s of integration in a 10 km/s resolution element results in a noise rms
(1 sigma) of 0.4K at 1900 GHz).
       CS: Make OTF measurement on the source. (In PV this should be checked
against a raster map as well). It should use a fine sampling (better than 0.25 beam?),
while the size of the map can be 5x5 beams at most.
       CS: Retrieve observed data from archive


                                           65
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


        CS: Do standard data reduction for map. Leave baselines in, as they may
provide extra continuum information
        CS: Plot integrated intensity of observed line. (Make sure that there were no
lines in the other side-band). Determine coordinates of peak. Make channel maps
and check visually for symmetric pattern over the source. Note that this is a
subjective measure of the roundishness of the source. We may need software that
can determine the center of ellipses instead of circles.
        CS: Check center of pointing calibration source against the position given by
the Herschel AOCS. Put centers into database. Additional information could be: the
position of the Sun in space craft x,y,z coordinates (angles are likely more useful)
and the various temperature sensor readings. Also the last and the following PACS
pointing calibration numbers should be readily available.
        CS: Make sure that the numbers are fed back into pointing models, the models
for the movements between startracker and the PACS focal plane.


Scenario 3:
This is a placeholder only!

After initial tests we have to rely on an automatic peaking-up procedure or for band 6,
(small) maps should be mandatory.

This should be checked regularly. It can be performed by making a small map in the
fashion described in the scenario above, perform a peaking-up and judge the results
from the two maps. Similarly small maps should always be checked for center
positions (but we have to be sure that we are dealing with point-like sources).

Frequency of occurrence:

Open issues:
  • Caveats are displacements of continuum and lines, strong winds in the inner
      atmospheres, which may be assymmetric. We don't know the sky at 1.9 THz!
  • OTF mapping has an intrinsic offset possibility in the continuum. Will we be
      unable to use the extra information from the continuum?

Comments:




                                          66
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0




2.4   UC-1.3 Spectral properties

Level:                  Summary
Scope:                  HIFI-ICC
Version:                0.1
Status:                 Draft
Author:                 Emmanuel Caux/Moncef Belgacem/Laurent Ravera -
                        18/09/2001

General Grade: High
Grade for ILT: High
Grade for Implementation: Medium

Brief description:
This U.C. is a summary use case dedicated to spectral calibration. It links the general
calibration strategy with the detailed measurement use cases.
Spectral calibration is one of the most important calibrations. It allows to determine
the spectral properties of the WBS, HRS, LO and the full instrument.
This calibration makes it possible to characterize the spectral instrument response.

Phase:
ILT                                                                      Y
IST/EET/GST...                                                           Y
LEOP/Commissioning                                                       Y
Calibration/PV                                                           Y
Science Demonstration                                                    Y
Routine Operations                                                       Y
Post Operations                                                          Y


Actors:
   • CS: Calibration Scientist
   • CO: Calibration Operator (?)
   • CM: Calibration Manager
   • MP: Mission (or measurement) Planner
   • IE: Instrument (System) Engineer
   • AS: Astronomer
   • DP: Data Processing pipeline

Triggers:
   • Project Scientist Team
   • P.I.
   • Requests from Herschel test manager
   • Calibration plan


                                          67
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


   •   Observers
   •   Failure recovery modes

Preconditions:
   • Availability of adequate laboratory and celestial sources
   • Knowledge of expected source profiles

Minimal post conditions:
   • Set of required spectra is measured

Success post conditions:
  • Spectral scale calibrated for both backends
  • Spectral response of the instrument calibrated
  • Spectral spurs identified and catalogued
  • Table of side-band ratios available in dedicated frequency ranges

Stakeholders and interests:
   • Calibration Group
   • Instrument System Group
   • Herschel System Group
   • General astronomers

Main success scenario:
  The following aspects of HIFI spectral calibration will be covered under section
  1.3:
  • 1.3.1: Frequency calibration
  • 1.3.2: Measure instrument line profile
  • 1.3.3: Measure spectral purity
  • 1.3.4: Measure sideband ratio
  • 1.3.5: Measure diplexer performance

Frequency of occurrence:
   • TBD

Open issues:
  • availability of celestial strong narrow lines is a concern if compactness is
      required

Comments:




                                         68
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


2.4.1 UC 1.3.1 Frequency calibration


Level:                  Summary
Scope:                  HIFI-ICC
Version:                0.1
Status:                 Draft
Author:                 Frank Helmich – 27-06-2002

General Grade: High
Grade for ILT: Medium
Grade for Implementation: High

Brief Description:
The frequency calibration of HIFI is split into the calibration of HRS and WBS
separately and by the combination of the two backends. Here only a summary of the
strategy is given, but reference to the spectral part of the HIFI framework document
(RD22) will be made as much as possible.

The frequency calibration is in principle set by the LO frequency. The resulting signal
at intermediate frequencies needs further characterization, which is done by a comb
signal for the WBS and reference to the LO signal by the HRS. See the child use
cases for the details. The transformation of the IF to the signal and image
frequencies is given in RD22. Note that the frequency calibration is only one of the
moments of the spectral line profile of HIFI. The use cases dedicated to the line
profiles should be consistent with the frequency calibration use cases.

Phase:
ILT                                                                      Y
IST/EET/GST...                                                           Y
LEOP/Commissioning                                                       Y
Calibration/PV                                                           Y
Science Demonstration                                                    Y
Routine Operations                                                       Y
Post Operations                                                          Y

Actors:
   • CS: Calibration Scientist

Triggers:
   • Cal. Plan

Preconditions:
   • Availability of adequate laboratory and celestial sources
   • Availability of a comb generator for WBS
   • Knowledge of expected source profiles


                                          69
                      HIFI calibration use cases                Doc.: ICC/2001-005
                                                                Date: 20/10/03
                                                                Issue: Draft 1.0



Minimal post conditions:
   • The measurement data required by the considered child Use Cases have
     been recorded

Success post conditions:
  • Assessment of the instrument behaviour or of the calibration and instrumental
     parameters of interest within the accuracy specified by the considered child
     Use Cases

Stakeholders and interests:
   • Calibration Group
   • General astronomers

Main success scenario:
The following Use Cases of Frequency Calibration will be covered under section
1.3.1:
   • WBS Frequency Calibration & Line Profile
   • Instrument Frequency Calibration


Frequency of occurrence:

Open issues:
  • The use cases below have to describe in much more detail how they are
      handling asymmetric profiles. For the WBS: What is the SWAS/KOSMA
      experience? For the HRS is there anything known from Odin/IRAM/JCMT? Is
      the whole procedure to come from lags to spectra guarded against
      asymmetries. Does the procedure introduce broadenings or asymmetries
      itself. How are we going to deal with that in S/S test and ILT? Will there be
      QLA procedures written for that ?

Comments:




                                         70
                      HIFI calibration use cases              Doc.: ICC/2001-005
                                                              Date: 20/10/03
                                                              Issue: Draft 1.0


2.4.1.1 UC-1.3.1.1 WBS Frequency Calibration & Line Profile

Level:                           User
Scope:                           HIFI-ICC
Version:                         0.1
Status:                          Draft
Author:                          Csaba Gal - 14/09/01

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
This use case describes the procedure to determine the calibrated frequency
measured by the WBS. The line profile check can be simultaneously evaluated from
the frequency calibration measurement.

Phase: all except post operations

Actors:
   • CS: Calibration Scientist
   • IE: Instrument (System) Engineer

Triggers:
   • Cal. plan
   • Observers

Preconditions:
   • Access to elementary integrations, zero switch, comb generator

Minimal post conditions:
   • Comb and zero spectra obtained with success

Success post conditions:
  • Calibrated frequency scale from comb measurement
  • Re-sampling of measured comb spectra (TBD)

Stakeholders and interests:
   • Calibration Group
   • General astronomers
   • Instrument System Engineer
   • System Group
   • Integration and Test Team

Main success scenario:



                                        71
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


   1. CS: switch ON "Zero" at the WBE/I, wait 100 msec, start to integrate, stop
      integrating after 1 sec, transfer data
   2. CS: switch OFF "Zero" and switch ON "Comb" at the WBE/I, start integrate,
      stop integrate after 1 sec, transfer data
   3. CS: evaluation with software at ICC (frequency scale, pixel resolution
      bandwidth, re-sampling if needed, line profile check). This part is taken care by
      WBS QLA applications and is described in RD28

Open issues:
   • Re-sampling, broadband frequency evaluation

Comments:




                                         72
                      HIFI calibration use cases              Doc.: ICC/2001-005
                                                              Date: 20/10/03
                                                              Issue: Draft 1.0


2.4.1.2 UC 1.3.1.2 Instrument Frequency Calibration

Level:                           User
Scope:                           HIFI-ICC
Version:                         0.1
Status:                          Draft
Author:                          Maryvonne Gerin - 08/31/01

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
This use case describes the procedure to check the instrument frequency accuracy
with respect to known spectra of external laboratory and celestial sources
(astronomical sources - gas cell)

Phase: all phases

Actors:
   • CS: Calibration Scientist

Triggers:
   • Cal. Plan
   • Failure notice from observations

Preconditions:
   • Availability of adequate calibration sources on the ground (gas cell at low
     pressure providing narrow lines < 1MHz at accurately known frequencies) and
     in-flight (well known astronomical sources). AGB stars could here be
     considered if their spectra are not too complex
   • Knowledge of the expected spectral lines from the source and dependence on
     beam size and frequency
   • Availability of robust centroid calculation algorithm (see e.g. RD15)

Minimal post conditions:
   • Set of required spectra measured

Success post conditions:
  • Line position assessed w.r.t. expected position

Stakeholders and interests:
   • Calibration group
   • General astronomers
   • Instrument System Group
   • Integration and Test Team


                                        73
                        HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                    Date: 20/10/03
                                                                    Issue: Draft 1.0



Main success Scenario:
  1. CS: Retrieve a source and information on the expected spectral lines from
      calibration database
  2. CS: Choose line frequencies
  3. CS: Choose a backend and adjust backend properties (bandwidth, resolution)
      to the expected signal
  4. CS: Do a frequency calibration (WBS and HRS)
  Note: the use of the HRS could be useful to monitor possible drift of the DRO up-
  converter.
  5. CS: Perform an intensity calibration /sensitivity (measure Jrec, and Jsys, see
      UC-1.1.1)
  6. CS: Integrate on the target. On the ground this consists in integrating through
      the gas cell successively filled and emptied (see RD30). In orbit, we most
      likely will use the double beam switching mode since sources with a rich
      spectra are usually compact.
  7. CS: Perform adequate data reduction to identify the spectral features present
      in the observed data and compare with the expected signals (from the
      calibration data base).
              a. (Likely) perform side-band de-convolution with adequate tools to
                   sort out lines from the respective image and signal bands
              b. Over-plot both observed and “reference” spectra after scaling to the
                   same frequency resolution
              Note: one might also want to skip the side-band de-convolution step at
              that stage and compare the measured spectra with synthesized DSB
              spectra of the expected features.
              c. Calculate line centroid(s) and/or fit expected profile(s)
  8. CS: Store results in the calibration data base
  9. CS: File in a calibration report

Frequency of occurrence:
   • During ILT, in PV and for health monitoring (?)

Open issues:
  • How many frequencies per source for this measurement ? RD23 might help in
      that. The minimum is one per LO chain. More frequencies may be needed for
      the first checks on the ground.
  • How many astronomical sources ? It is likely that we don't need a uniform
      distribution on the sky. See RD34 for tentative numbers.

Comments:
A companion document should present reference spectra for astronomical and
laboratory sources (or a documented database will be enough).

The intensity calibration may be useful in case of problems: line identification will be
easier if the intensity is calibrated.



                                           74
                       HIFI calibration use cases                   Doc.: ICC/2001-005
                                                                    Date: 20/10/03
                                                                    Issue: Draft 1.0


2.4.2 UC-1.3.2 Measure Instrument Line Profile

Level:                             Summary
Scope:                             HIFI-ICC
Version:                           0.1
Status:                            Draft
Author:                            David Teyssier - 08/31/01

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
This use case describes the general procedure to check the instrument line profile
through a series of S/S and the integrated instrument itself. This is the parent UC for
a series of dedicated line profile measurement UC's.

Phase: all except Post Operations

Actors:
   • CS: Calibration Scientist

Triggers:
   • Cal. plan
   • Failure notice from observations

Preconditions:
   • Readiness of specific child UC's to be conducted
   • Availability of identified reference line profiles to be compared to the specific
     measurement

Minimal post conditions:
   • Required line profile measured

Success post conditions:
  • Required line profile characterised
  • Accuracy TBD % depending on the child UC

Stakeholders and interests:
   • Calibration group
   • General astronomers
   • Instrument System Engineer

Main success scenario:




                                           75
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


The instrument frequency line profile results from the sharpness and stability of the
local oscillator(s), convolved to the line profile of the spectrometer in use. The use
cases included here are thus:
   • 1.3.2.1 - WBS line profile (incorporated in UC-1.3.1.1)
   • 1.3.2.2 - HRS line profile
   • 1.3.2.3 - LO line profile
   • 1.3.2.4 - Instrument line profile

Frequency of occurrence:
During ILT, in PV and for health monitoring

Open issues:
  • The commonalities between the child UC's is to be assessed.

Comments:




                                           76
                      HIFI calibration use cases             Doc.: ICC/2001-005
                                                             Date: 20/10/03
                                                             Issue: Draft 1.0




2.4.2.1 UC-1.3.2.1 WBS Line Profile

Level:                          User
Scope:                          HIFI-ICC
Version:                        0.1
Status:                         Draft
Author:                         Csaba Gal - 14/09/01

Brief description:
See UC-1.3.1.1. (WBS Frequency calibration & line profile)

Present for document consistency only!




                                         77
                      HIFI calibration use cases                Doc.: ICC/2001-005
                                                                Date: 20/10/03
                                                                Issue: Draft 1.0




2.4.2.2 UC-1.3.2.2 HRS line profile

Level:                      User
Scope:                      HIFI-ICC
Version:                    0.1
Status:                     Draft
Author:                     Laurent Ravera/Moncef Belgacem - 17/09/01

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
This use case describes the procedure to measure the HRS line profile. In this test,
we monitor for different frequencies of the input signal the power in one spectral
channel. To do this operation we need to directly feed the HRS (without the HIFI
front-end) with a signal generator. As a consequence, this measurement can be done
on the ground only with the HRS disconnected from any other subsystems of HIFI.
The hardware and the software needed to do this test will be included in the HRS
GSE. This test will be performed during the HRS performance tests. It will be done
during the ILT only if it seems that the HRS line profile has to be verified. The
procedure to measure the full instrument line profile is described in UC-1.3.2.4.

Phase: Subsystem tests and ILT

Actors:
   • (HRS) IE

Triggers:
   • HRS performance tests
   • Cal. plan

Preconditions:
   • working HRS
   • availability of a dedicated HRS GSE including:
         o a signal generator in the correct frequency range (4 - 8 GHz) with a
            sufficient precision (1kHz TBC) and stability.
         o a noise generator in the correct frequency range (4 - 8 GHz).
         o the line profile test software (HRS QLA, cf RD29).

Minimal post conditions:
   • set of required output data adquired




                                        78
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


Success post conditions:
  • Measured line-width in agreement with the HRS operating mode.

Stakeholders and interests:
   • Calibration Group
   • Instrument System Engineer
   • Integration and Test Team

Main success scenario:
  1. IE: Combine the noise signal with the single frequency signal to feed the HRS.
  2. IE: Set the noise generator power level to be in the HRS input power range: -
      95(+/-5)dBm/MHz.
  3. IE: Set the signal generator at the correct power level: 1/2 of the power in one
      channel (TBC).
  4. IE: Set the frequency range for the measurement (stop and start frequency).
  5. IE: Set the frequency step for the measurement (e.g. the channel spacing/20).
  6. IE: Set HRS-ACS in the required mode (wide band, nominal or high reslution).
  7. IE: Set HRS-IF sub-bands at the required frequencies.
  8. IE: Run the "tune_correlator_input_power" routine.
  9. IE: Set the integration time per subscan Ts.
  10. IE: repeat in the freq_range:
          • Do a subscan integration in OFF position (line switched OFF).
          • Do a subscan integration in ONf_n position (line fn switched ON).
          • Do fn = fn +freq_step
      Remark: As in On-the-Fly mapping the same OFF spectra can be used as a
      reference for several ON_n spectra (OFF ON_1 ON_2 ON_3 ON_4,...).
      The frequency of occurrence of the OFF spectra depends on the system
      stability.
  11. IE: use a dedicated routine to extract the line profile from data and to compare
      the result with a reference. Routine description is to be found in RD29.

Frequency of occurrence:
   • In HRS performance tests.
   • During ILT if the HIFI line profile is not correct and if we want to verify the HRS
     line profile.

Open issues:

Comments:




                                          79
                       HIFI calibration use cases                Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


2.4.2.3 UC-1.3.2.3 LO Line profile

Level:                           User
Scope:                           HIFI-ICC
Version:                         0.1
Status:                          Draft
Author:                          Michel Pérault - 13/09/01

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
This use case describes the measurement of the LO signal frequency profile. Both
the instantaneous line profile and the time integrated line profile may have to be
considered in case of frequency jitter. Both frequency stability and LO line width
requirements are recalled in RD22.

Phase: ILT and IST/EET

Actors:
   • CS: Calibration Scientist
   • IE: Instrument (System) Engineer

Triggers:
   • Need to check the quality of an essential part of the system
   • Abnormal line profiles detected during tests

Preconditions:
   • monitoring of the phase-lock comparison system (spectrum analyser)
   • availability of a high resolution spectrometer at the LO frequencies

Minimal post conditions:
   • production of quantitative result tables

Success post conditions:
  • specifications satisfied


Stakeholders and interests:
   • Instrument System Group

Main success scenario:
  1. IE: mount the tested LO at the spectrometer signal port and tune it to the
      selected frequency
  2. IE: tune the spectrometer at the tested LO frequency


                                         80
                      HIFI calibration use cases                Doc.: ICC/2001-005
                                                                Date: 20/10/03
                                                                Issue: Draft 1.0


   3. IE: monitor the comparison signal of the phase-lock system of the tested LO
      with a spectrum analyser. Line profile should have an intensity of TBD dB and
      a width smaller than TBD kHz. No secondary peaks or pedestal is admissible
      above TBD dB below the main peak. Check with the instrument specifications.
   4. IE: directly measure the actual LO line profile with the high-resolution
      spectrometer.

Frequency of occurrence:
Should be fully characterized at component and sub-system level over each LO sub-
band. Verifications during AIV/ILT are desirable, but need a dedicated set-up. Only
the phase-lock comparison signal can be routinely monitored at this level. No access
at all in orbit.

Open issues:
  • Availability of a spectrometer (e.g. a calibrated HIFI in its highest resolution
      mode) may be problematic. Would other instrumentation be available at some
      place ?
  • How should the frequency profiles be related, of the LO proper, and of the
      phase-lock comparison signal, once de-convolved from their respective
      measuring profile ? Is the easiest measurement enough ? NW to comment
      and specify
  • Is degradation with time of the LO profile integrated width expected at all ?

Comments:
Measurement plans at S/S level should be inquired.




                                        81
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


2.4.2.4 UC-1.3.2.4 Measure Instrument line profile


Level:                    User
Scope:                    HIFI-ICC
Version:                  0.1
Status:                   Draft
Author:                   David Teyssier/Maryvonne Gerin- 31/08/01

General Grade: High
Grade for ILT: Medium
Grade for Implementation: High

Brief description:
This use case describes the procedure to check the instrument line profile
consistency with respect to reference profiles of known "calibration" sources.

Phase: all phases
Post Operations could be done post-mortem on measurements not specifically
dedicated to this topic

Actors:
   • CS: Calibration Scientist

Triggers:
   • Cal. plan
   • Failure notice from observations

Preconditions:
   • availability of adequate calibration sources on the ground (gas cell) and in-
     flight (astronomical sources)
   • knowledge of expected source profile parameters
   • availability of line fitting routine

Minimal post conditions:
   • Required line profile measured

Success post conditions:
  • Astronomical or laboratory line profile seen by HIFI consistent with source and
     instrument properties
  • Instrument line profile characterised
  • Accuracy TBD %

Stakeholders and interests:
   • Calibration group
   • General astronomers


                                         82
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


   •   Instrument System Engineer

Main success scenario:
  1. CS: Retrieve reference for source and line profile from calibration database
      Note: it is anticipated that several types of line profile will be tested, from
      narrow to very broad. The database must sample such a range (RD34).

   2. CS: Choose backend resolution and bandwidth
   3. CS: Do a frequency calibration of the WBS
      Note: A systematic frequency calibration of the HRS is not required given the
      digital nature of the backend but should globally be done at S/S level with the
      receivers.

   4. CS: Integrate on the source. On the ground this consists in integrating through
      the gas cell successively filled and emptied (RD30). In orbit, we most likely will
      use the double beam switching mode for compact sources, while position
      switching may be suitable for more extended targets. The integration time
      depends on the accuracy level to which the profile needs to be compared with
      the reference one (as well as the expected line strength). A possible approach
      is to aim at a 5-sigma detection level of the faintest structure of interest (e.g.
      double-peak, line wings, etc). The detailed description of how the instrument
      resolution element is defined is given in RD22.

   5. CS: Perform adequate data reduction: intensity and frequency calibration
      (including likely side-band de-convolution), average if necessary, then
      compare measured and expected line profiles:
      • Over-plot the two profiles (optional)
      • Perform eventual single or multi-component fits and compare to expected
          values
      • Estimate measurement deviation to the reference profile: plot residual of
          the difference and compare with signal noise.

   6. CS: Store results in the calibration database
      Note: at this stage, we did not really address the issue of the diagnostic of the
      measurements, ie. the interpretation of a profile deviation w.r.t. the expected
      shape. Anticipated are:
      • Inconsistent profile width: is the backend resolution correct, stable
         (possible temperature dependence ?), are there strong jitter in the LO ?
         Check whether different beam sizes of reference and HSO spectra are to
         be invoked.
      • Inconsistent intensity: position or calibration accuracy. Check whether
         different beam sizes of reference and HSO spectra are to be invoked.
      • Inconsistent frequency: frequency calibration is wrong, or the LO frequency
         is bad?

Frequency of occurrence:



                                           83
                       HIFI calibration use cases              Doc.: ICC/2001-005
                                                               Date: 20/10/03
                                                               Issue: Draft 1.0


During ILT, in PV and for health monitoring

Open issues:

   •   How many sources are required and how to find this number ? Do we have to
       care about:
          o homogeneous source distribution on sky ?
          o the category of profile one wants (depending on what is to be checked)
              ?
       Note that tentative numbers are given in RD34.

Comments:
  • A companion document should present all calibration profiles (gas cell or
    astronomical sources) in order to characterise the lines as accurately as
    possible and have a kind of "ID-card" of the expected measurement. A
    documented database might do the job.

   •   This is a very difficult use case. We have a very large beam and may be
       heavily affected by different components in the line of sight. The
       observed line shape may depend on the pointing. A very much over-
       sampled large raster may help somewhat. Note that we do not expect
       many sources to have small line-widths. We need a database here.
       Tentative observing sequences and strategies are proposed in RD22.




                                         84
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


2.4.3 UC-1.3.3 Measure Spectral Purity

Level:                       Summary
Scope:                       HIFI-ICC
Version:                     0.1
Status:                      Draft
Author:                      N.D. Whyborn – 13-10-2001

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
The purpose of this Use Case and its children is to measure the spectral purity of
HIFI to ensure that the level of all significant spurious signals and spurious responses
are known. Un-catalogued instrument spectral artefacts could be misinterpreted as
astronomical features.

Phase:
ILT                                                                         Y
IST/EET/GST...                                                              Y
LEOP/Commissioning                                                          N
Calibration/PV                                                              Y
Science Demonstration                                                       N
Routine Operations                                                          Y
Post Operations                                                             N

Actors:
CS: Calibration Scientist
IE: Instrument Engineer

Triggers:
   • Calibration and AIV plans
   • Hardware failure/degradation

Preconditions:
   • Availability of adequate source equipments: harmonic generator and Hot/Cold
     internal source

Minimal post conditions:
   • Set of required spectra successfully measured

Success post conditions:
  • Measurement of level of spurious signals and responses
  • Generation of a catalogue of spurious internally generated signals and their
     strength relative to the nominal noise level


                                          85
                       HIFI calibration use cases                Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


   •   Generation of a catalogue of spurious internally generated responses and their
       strength relative to the wanted response

Stakeholders and interests:
   • Calibration Group
   • General astronomers (to ensure confidence in spectra)
   • Instrument System Engineer
   • Integration and Test Team

Main success scenario:
This UC consists of two child use cases:
   • UC-1.3.3.1: measure spurious signals
   • UC-1.3.3.2: measure spurious responses

Frequency of occurrence:

Open issues:
  • we have to decide what to do with information about instrument artefacts.
      Should this just be catalogued, should affected AOT's (frequency settings)
      generate a warning, or should the data pipeline flag suspect data in some
      way?

Comments:
It will not be possible to measure spurious spectral responses after ILT due to the
lack of access for tuneable test sources. However, spurious signals can be searched
for at any time during or after ILT.




                                         86
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


2.4.3.1 UC-1.3.3.1 Measure Spurious Signals

Level:                      User
Scope:                      HIFI-ICC
Version:                    0.1
Status:                     Draft
Author:                     N.D. Whyborn - 2001-10-13

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
Search for spurious signals in the spectra output by HIFI and determine their
frequency and strength. A spurious signal is a spectral feature which is present in the
spectrometer output in the absence of any spectral features in the signal entering
HIFI, e.g. while HIFI is looking at a black body.

Phase:
ILT                                                                         Y
IST/EET/GST...                                                              Y
LEOP/Commissioning                                                          N
Calibration/PV                                                              Y
Science Demonstration                                                       N
Routine Operations                                                          Y
Post Operations                                                             N

Actors:
CS: Calibration Scientist
IE: Instrument Engineer

Triggers:
   • Calibration and AIV plans
   • Hardware failure/degradation

Preconditions:
   • Availability of internal Hot/Cold source

Minimal post conditions:
   • Set of required spectra successfully measured

Success post conditions:
  • measurement of the frequency and level of any internally generated spurious
     signals
  • generate catalogue of spurious internally generated signals and their strength
     relative to the nominal noise level


                                          87
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0



Stakeholders and interests:
   • Calibration Group
   • General astronomers (to ensure confidence in spectra)
   • Instrument System Engineer
   • Integration and Test Team

Main success scenario:
  1. IE: Select both WBS and HRS in low resolution mode (HRS sub-band
      frequency settings TBD)
  2. IE: Tune instrument to a set of frequencies (value and number TBD).
  3. IE: For each frequency perform a calibration using the internal hot and cold
      calibration sources
  4. IE: Use the hot and cold spectra to form a spectrum of Trec.
  5. IE: Search the system noise temperature spectra for spurious lines (narrow
      peak in Trec).
  6. IE: When a line is found, try to identify the source by either calculation or by
      changing the LO and HRS sub-band frequencies. Depending on that, refine
      measurement algorithm.

Frequency of occurrence:

Open issues:
  • we have to decide which HRS configuration(s) to test
  • we have to decide what to do with information about instrument artefacts.
      Should this just be catalogued, should affected AOT's (frequency settings)
      generate a warning, or should the data pipeline flag suspect data in some
      way?

Comments:




                                          88
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


2.4.3.2 UC-1.3.3.2 Measure Spurious Responses

Level:                       User
Scope:                       HIFI-ICC
Version:                     0.1
Status:                      Draft
Author:                      N.D. Whyborn - 2001-10-13

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
Search for spurious responses in the spectra output by HIFI and determine their
frequency and strength. A spurious response is a ghost spectral feature present in
the spectrometer output which is related to a spectral feature in the signal entering
HIFI, but which appears at the wrong frequency. The image response of a double
sideband mixer is an example of spurious response.

This Use Case can only be performed during ILT since it requires an external test
signal source (see ILT Test Procedure Descriptions)

Phase:
ILT                                                                          Y
IST/EET/GST...                                                               N
LEOP/Commissioning                                                           N
Calibration/PV                                                               N
Science Demonstration                                                        N
Routine Operations                                                           N
Post Operations                                                              N

Actors:
IE: Instrument Engineer

Triggers:
   • AIV plans

Preconditions:
   • Availability of a suitable test signal source in the ILT test set-up. The test
     source should be at a known frequency or harmonics thereof and be capable
     of being square-wave modulated. A microwave synthesiser driving a comb
     generator may be suitable if the sub-mm harmonics are strong enough to
     detect easily.

Minimal post conditions:
   • Set of required spectra successfully measured



                                          89
                       HIFI calibration use cases                Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


Success post conditions:
  • measurement of the frequency and level of any internally generated spurious
     responses
  • generate catalogue of spurious internally generated responses and their
     strength relative to the wanted response

Stakeholders and interests:
   • Calibration Group
   • General astronomers (to ensure confidence in spectra)
   • Instrument System Engineer
   • Integration and Test Team

Main success scenario:
  1. IE: Select both WBS and HRS in low resolution mode (HRS sub-band
      frequency settings TBD)
  2. IE: Tune instrument to a set of frequencies (value and number TBD).
  3. IE: Optionally perform and temperature scale calibration (TBD)
  4. IE: Tune a square-wave modulated test signal to a set of frequency (value and
      number TBD)
  5. IE: For each frequency acquire a differenced spectrum in synchronism with the
      test source modulation
  6. IE: Search the resulting spectra for spurious lines (narrow peak in Trec).
  7. IE: When a line is found, try to identify the source by either calculation or by
      changing the test signal, LO and HRS sub-band frequencies.

Frequency of occurrence:

Open issues:
  • we have to decide which HRS configuration(s) to test
  • we have to decide which receiver frequencies to test
  • we have to decide which test source frequencies to measure
  • we have to decide whether calibration spectra are needed
  • we have to decide what to do with information about instrument artifacts.
      Should this just be catalogued, should affected AOT's (frequency settings)
      generate a warning, or should the data pipeline flag suspect data in some
      way?

Comments:
Spurious instrument responses have two different effects. The first is the obvious one
of causing artefacts in the resulting spectra which can be misinterpreted as
astronomical lines. The second effect is more subtle and is that the flux scale of
spectra is in error. This results because we will use black body sources almost
exclusively for flux calibration. The measured response of the spectrometers to a
change in the brightness temperature of a black body source at the HIFI input is
proportional to the sum of the wanted response plus all unwanted responses. In the
case of a double sideband mixer this is accounted for by applying knowledge of the


                                         90
                      HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


sideband ratio. By analogy we need to determine the spurious response ratio for all
spurious responses to achieve an accurate flux calibration.




                                         91
                       HIFI calibration use cases                   Doc.: ICC/2001-005
                                                                    Date: 20/10/03
                                                                    Issue: Draft 1.0




2.4.4 UC 1.3.4 Measure Side Band Ratio

Level:                       User
Scope:                       HIFI-ICC
Version:                     0.2
Status:                      Draft
Author:                      Frank Helmich 27 September 2001 draft 0.1
                             Frank Helmich 10 January 2001 draft 0.2
                             - clarifications
                              - better descriptions of method
                              - including ILT description

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief Description:
This use case aims at a determination of the side-band ratio to within a few percent.
In orbit the aim is to check the side-band ratios within 5-10% of the values
determined on the ground

Background Information:
A mixer is a device that brings down the frequency of the sky signal to lower
frequencies where amplification is much easier. It does do so by mixing the sky
signal with a signal that is very close in frequency, the so-called local oscillator (LO)
signal. In fact two (co-)sine waves are added, giving the sum frequencies (very high
frequency) and difference frequency (reasonably low frequency). Problem, or
advantage, is that in this mixing process not only the frequency of interest (signal
frequency) is mixed down, but also a mirror frequency (the image frequency). The
way in which this happens is well understood and well under control on most ground-
based observatories. HIFI, however, has extra complications. It has an enormous
tuning range, a wide bandwidth and large intermediate frequency (IF). The sensitivity
of the mixer can vary over its applicable range by over 20%. This means that the
same input power presented to both signal and image side-band will give different
outputs in the resulting spectrum (or continuum, whatever is applicable).
Characterization of the side-band ratio (i.e. gain in signal side-band divided by gain in
image side-band) is thus mandatory, especially since 10% error in side-band ratio
gives an absolute error in intensity calibration of 5%. Note that 10% error is
the base-line for HIFI (3% goal).

Phase:
ILT                                                                          Y
IST/EET/GST...
LEOP/Commissioning                                                           Y? (TBD)
Calibration/PV                                                               Y


                                           92
                      HIFI calibration use cases                Doc.: ICC/2001-005
                                                                Date: 20/10/03
                                                                Issue: Draft 1.0


Science Demonstration                                                     Y
Routine Operations                                                        Y
Post Operations                                                           N

Actors:
   • Instrument Engineer
   • Calibration Scientist

Triggers:
   • AIV plan
   • Calibration Plan

Preconditions:
 -ILT:
    • A gas cell in which absorption lines of linear molecular rotors can be seen
       against a constant background. Choices can be OCS, CO, N2O, see RD30.
    • Loads providing the background
    • Line fitting routines

- In orbit:
    • A predetermined LO stepping scheme
    • A pre-determined set of critical frequency ranges (see RD23)
    • Celestial source visible (which one(s) TBD)
    • Very accurate pointing knowledge and preferably knowledge on structure of
       the source. Point sources are preferable.


Minimal post conditions:
   • determination of side-band ratio within 5%(TBD) during ILT and verification
     within 20%(TBC) in-orbit

Success post conditions:
  • determination of side-band ratio within 2%(TBD) during ILT and verification
     within 10%(TBC) in-orbit

Stakeholders and interests:
   • Calibration group
   • Instrument System Group
   • Herschel System Group
   • General astronomers

Main success scenario:
Side-band ratios for selected frequency range (as function of LO and IF frequency)

ILT:



                                         93
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


   1. The detailed description is to be found in the test plan (see RD19). Here, only
      the method is described. From the gas cell, absorption lines are available from
      very well defined frequencies. At high (order 0.1-1 atm) pressure the lines are
      so strongly broadened that they completely fill one of the side-bands. Looking
      at a continuum in one side band and zero (continuum minus strongly
      broadened saturated line), followed by a step in which there is continuum in
      both the side bands, will give accurate power levels per side-band (see also
      RD30). From these power levels the side-band ratios can be determined.
      Strictly speaking, one line is sufficient, but many lines over a LO sub-band,
      make the procedure more reliable
   2. Determine power level with gas absorption line in signal side-band
      P(signal)=0, P(image)=P(continuum)
   3. Repeat with no absorbing gas present
   4. Shift several times by 1 GHz and repeat the above steps.
   5. Compute side-band ratio. It is simply given by:

                           Rg=Gssb/Gusb=(1-S2/S1)/(S2/S1)

   Where S1 and S2 are the Hot-Cold difference measurement through the
   respectively empty and gas-filled cell.

At some point the sum of the two power in the two side-bands, P(signal)+P(image),
should become constant, until the next absorption line moves in and in the absence
of a differing side-band ratio. Modulation of the sum of P(signal) and P(image), as
measured by the backends, in between these absorption lines is due to modulation of
the sensitivity reponse as a function of LO-frequency. Dividing the ratio of
P(signal)/P(image) gives the side-band ratio, as a function of LO frequency. The
measurement without gas cell, will provide the sum of P(signal) and P(image)
permanently as a function of LO-frequency, whereas from the measurement with gas
the value of the image and signal side-band can be extracted in an inductive fashion.

Unfortunately, standing waves occurring in a cavity formed between the mixer horn
and the LO feed make this theoretical picture untrue (see RD25). These standing
wave will create an additional pattern in the side-band ratio dependence with LO-
frequency, changing with each LO setting as LO power will be modified. One
possibility would consist in using a path-length modulator in the LO-mixer path so that
measurement of the side-band ratio at various phases of the ripple pattern may be
obtained without modifying the LO setting, then averaged to provide the required
parameter. This would increase the number of measurements by a factor of 3-4
(TBC).

In-orbit:
    1. CS: Object must be selected. Which ones are suitable? TBD, but UC HII
        regions and hot cores are very likely. A selection can be found in the
        calibration plan.
    2. CS: Plan observation of selected frequency range with predetermined LO
        stepping. LO-stepping to be determined during AIV and through study (e.g.


                                          94
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


      Comito & Schilke, 2002). Note that in first instance we can only do this per LO-
      band. However, the method should be set-up in such a fashion that all data
      from all LO-bands can be handled at the same time.
   3. CS: Set up observations
      Note: observation must be made in double beam-switch mode in order to get
      straight base lines. Signal-to-noise ratio should be high. In case of Orion 1K in
      1 MHz with a 10m telescope was needed. For the 3 times larger beam of
      Herschel the aim is 0.1K in 1 MHz for Orion. ALL other hot cores must have a
      noise level at least 5 (TBC) times smaller.
   4. CS: Obtain all the spectra. Visually examine all the spectra on spurious
      signals. Remove spurious signals and determine quality visually. If quality is
      OK, continue, otherwise use only the unaffected spectra that need to be
      arranged so that they are contiguous in LO-frequency.
      Question: Should baselines be removed? Most likely not, since there is a good
      possibility that many lines will be in absorption at higher frequencies. Can the
      method deal with absorption lines? This is in the plans.
   5. CS: Apply the Maximum Entropy Method of P. Schilke's group to the data.
      (see Comito & Schilke 2002, and references therein) . Use here as input the
      side-band ratios from AIV. Inspect single-side band spectrum versus pieces of
      the DSB spectra. Examine side-band ratio's and overplot on AIV values.
      Iterate on these results. In the maximum entropy method, the goal is to derive
      a spectrum as smooth as possible, while still explaining all the spectral
      features. Maximization of the entropy, allows for simultaneously fitting of the
      single-side-band spectrum and of the side-band ratio. If the pointing is very
      stable (likely), other calibration parameters could be fitted simultaneously. The
      measurements per LO-band consist of 160 GHz to be scanned in steps of
      about 1 GHz. The 4 GHz bandwidth thus guarantees each frequency to be
      sampled 4 times. Each 4 GHz spectrum consist of somewhat more than 4096
      pixels (for each polarization). So for each polarization the measurement vector
      consists of m=160x4x4096 elements. Our retrieved spectrum consists of
      n=160x1024 elements for the single-side band spectrum plus 160 (TBC) side-
      band ratios for LO-frequencies sampled at 1 GHz grid. In the worst case the
      retrieval matrix mxn is thus very large. This problem will get worse when the
      complete HIFI range is considered. Smart implementation is thus necessary,
      which should be guided by the de-convolution working group lead at IPAC.

Frequency of occurrence:
   • Once during ILT
   • Once during PV
   • For every acquired spectral scan

Open issues:
Should for each SINGLE spectrum a sampling of 4x1 GHz steps be chosen, so that
for each SINGLE spectrum de-convolution is possible? Note that frequency steps
should NOT be done with regular increments (see Comito & Schilke 2002).




                                          95
                      HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


Comments:
The complete procedure is described in the: "Detailed test plan of HIFI instrument
Tests". SRON-G/HIFI/PL/2001-001. For draft 0.2 of this use case the draft version
from 8 November 2001 is used

References:
Comito C., Schilke P., 2002, A&A 395, 357
Sutton, E.C., et al. 1994. ApJS 97, 455




                                         96
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0




2.4.5 UC 1.3.5 Measure diplexer performance

Level:                      User
Scope:                      HIFI-ICC
Version:                    0.1
Status:                     Draft
Author:                     D. Teyssier

General Grade: Low
Grade for ILT: Low
Grade for Implementation: Medium

Brief description:

This use case is still to be written and may best be filled in by KOSMA based on their
recently held tests. The final Use case must be given by the Focal Plane SubSystem.

One issue to be addressed here is the optimisation of diplexer position and its
consequences in terms of losses (Tsys increase) and side-band ratio changes. In this
later case, this is particularly significant as side-band ratio will be measured on a
limited amount of points and interpolated to the whole frequency range. Diplexer
position changes across the band may thus not be compatible with the interpolated
values.

Phase:
ILT                                                                     Y
IST/EET/GST...                                                          N
LEOP/Commissioning                                                      N
Calibration/PV                                                          N
Science Demonstration                                                   N
Routine Operations                                                      N
Post Operations                                                         N

Actors:
   • CS: Calibration Scientist

Triggers:
   • Cal. Plan

Preconditions:
   • …

Minimal post conditions:
   • …



                                          97
                    HIFI calibration use cases   Doc.: ICC/2001-005
                                                 Date: 20/10/03
                                                 Issue: Draft 1.0



Success post conditions:
  • …

Stakeholders and interests:
   • Calibration Group
   • General astronomers

Main success scenario:

Frequency of occurrence:

Open issues:
  • …

Comments:




                                 98
                        HIFI calibration use cases                   Doc.: ICC/2001-005
                                                                     Date: 20/10/03
                                                                     Issue: Draft 1.0




2.5   UC-1.4 Intensity properties

Level:                             Summary
Scope:                             HIFI-ICC
Version:                           0.1
Status:                            Draft
Author:                            Michel Pérault - 13/09/01

General Grade: High
Grade for ILT: High
Grade for Implementation: Medium

Brief description:
This U.C.is a summary of the use cases dedicated to intensity (photometric)
calibration. It links the general calibration strategy described in the leading sections
of the calibration plan with the details of the calibration measurements described in
the dependent use cases.

Intensity calibration is the basic astronomical calibration. For a high resolution
spectrometer it primarily covers the actions allowing the delivery of calibrated of
(narrow) line fluxes, but broad band spectra may also be requested, as well as
calibrated continuum fluxes. As a matter of fact providing calibrated fluxes in each
spectrometer resolution element at once yields spectral profiles and fluxes for lines or
features of any width, within the analysed band, as well as continuum fluxes.

Intensity calibration assumes a proper calibration of the angular and frequency
coordinates. In addition the knowledge of the instrumental angular and frequency
profiles is often requested for intensity calibration, which in turn provides the absolute
normalization of these profiles.

Calibrated fluxes should be independent of the observing mode, but calibration
operations and flux derivation are likely to be observing mode dependent.

Phase: all phases

Actors:
   • CS: Calibration Scientist
   • CO: Calibration Operator (?)
   • CM: Calibration Manager
   • MP: Mission (or measurement) Planner
   • SE: System Engineer
   • GO: General Observer
   • DP: Data Processing pipeline




                                           99
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


Triggers:
   • Project scientist
   • P.I.
   • Requests from Herschel test manager
   • Calibration plan
   • Observers
   • Failure recovery modes

Preconditions:
   • Availability of absolute photometric calibrators in the lab. (external black
     bodies) and in the sky (planets, asteroids)
   • Availability of internal hot and cold loads
   • Availability of variable test signals

Minimal post conditions:
   • Set of required spectra and outputs successfully measured

Success post conditions:
  • Successful completion of all child use cases

Stakeholders and interests:
   • Calibration Group
   • System Group
   • Herschel System Group
   • Astronomers

Main success scenario:
The following aspects of HIFI intensity calibration will be covered under section 1.4:
   • calibration of the internal calibration source
          o 1.4.1: measure internal calibrator radiometric properties
          o 1.4.2: measure internal calibrator coupling
   • calibration of the spectrometer coupling to the sky
          o 1.4.3: measure telecope aperture efficiency
          o 1.4.4: measure telescope beam efficiency
          o 1.4.5: do intensity calibration
   • intensity distorsions and related issues
          o 1.4.6: measure dynamic range
          o 1.4.7: measure non-linearity

Frequency of occurrence:
   • 1.4.1: only at S/S level, once for every model
   • 1.4.2: only at S/S level, once for every model
   • 1.4.3: each time a band-pass calibration is performed
   • 1.4.4: each time am adequate solar system body is available
   • 1.4.5: each time an observation is performed



                                          100
                      HIFI calibration use cases      Doc.: ICC/2001-005
                                                      Date: 20/10/03
                                                      Issue: Draft 1.0


  •   1.4.6: on the ground only (?), once per model
  •   1.4.7: on the ground only (?), once per model

Open issues:
  • see child use cases

Comments:
  • see child use cases




                                       101
                      HIFI calibration use cases               Doc.: ICC/2001-005
                                                               Date: 20/10/03
                                                               Issue: Draft 1.0


2.5.1 UC-1.4.1 Measure internal calibrator radiometric properties

Level:                      User
Scope:                      HIFI-ICC
Version:                    0.1
Status:                     Draft
Author:                     David Teyssier & Michel Pérault - 14/09/01

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
This use case describes procedures needed to characterise the calibration source
assembly (hereafter CSA) and assess its radiometric properties. These
measurements are expected to be conducted at S/S level. They will help
understanding the integrated instrument and are needed for the HIFI intensity
calibration.

These measurements should fully characterize the CSA S/S. Modelling of the CSA
will be required in order to interpolate the CSA properties to conditions (e.g.
frequencies) for which the CSA will not have been calibrated in the instrument
(neither in the lab, nor in orbit, cf UC-1.4.2). These properties are:
    • Broad-band emissivity / reflectivity
    • Uniformity of the aperture
    • Angular dependence
    • Emission spectrum
    • Stability

Phase:

S/S tests                                                                Y
ILT                                                                      N
IST/EET/GST...                                                           N
LEOP/Commissioning                                                       N
Calibration/PV                                                           N
Science Demonstration                                                    N
Routine Operations                                                       N
Post Operations                                                          N


Actors:
   • CS: Calibration Scientist
   • IE: Instrument (System) Engineer

Triggers:


                                        102
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


   •   Cal. plan

Preconditions:
   • availability of adequate sensors to measure physical temperatures of hot and
     cold loads
   • availability of the measurement instrumentation described below

Minimal post conditions:
   • set of required output successfully measured

Success post conditions:
  • Measure of emissivity of the CSA (as a function of physical temperature,
     frequency, position in the aperture, and angle)
  • Accuracy TBD % relative, TBD % absolute To be compared with the Specs

Stakeholders and interests:
   • Calibration Group
   • Integration and Test Team
   • System Group
   • Instrument System Engineer

Main success scenario:

Emissivity / reflectivity measurement:
  1. IE: use a reflectometry setup to derive (mean) reflectivity of the source
      aperture at 1 frequency per HIFI sub-band
  2. IE: derive emissivity = 1 - reflectivity

CSA angular mapping:
  1. IE: use a calibrated detector + cold filter with a 2-d scanner, in order to
     measure the integrated emission of the CSA over its aperture, as a function of
     viewing angle. This measurement is an indirect verification of the source
     "blackness" and is needed for deriving the coupling with the different HIFI's
     beams. To be done at a few physical temperatures, 1 freq. per sub-band of
     HIFI.
  2. IE: derive (calibrated) emission characteristic. The absolute calibration is only
     desirable. The detector + cold filter assembly may be calibrated with the
     submm Absolute Hot Black Body source.

ICS aperture mapping:
   1. IE: use the above set-up with a 1:1 re-imager, in order to map the CSA
      aperture, integrated over angles. This measurement is needed for deriving the
      coupling with the different HIFI's beams. To be done at a few source physical
      temperatures, 1 freq. per sub-band of HIFI.
   2. IE: derive (calibrated) emission map. Same remark.



                                         103
                        HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                    Date: 20/10/03
                                                                    Issue: Draft 1.0


CSA frequency scan:
  1. IE: use a calibrated spectrometer (FTS ?) in order to try and estimate eventual
      deviations from black-body spectrum. The spectral calibration could use the
      submm Absolute Hot Black Body source as a reference. To be done at a few
      source physical temperatures. Frequency resolution TBD.

Stability:
   1. IE: use the photometric setup for long records. Repeat at various time
        intervals.

Note as an example (sub-scenario): measure emissivity
   1. CS: measure power output on a reference aluminium plate (supposed to be
      perfectly reflective)
   2. CS: measure output power on the internal load
   3. CS: Divide the two measurements to calibrate out the transmission function
      and derive reflectivity
   4. CS: compute emissivity: emissivity = 1 - reflectivity

Frequency of occurrence:
   • During S/S level tests

Open issues:
  • Detailed setups
  • Range of parameters to be used for these tests
  • ICS physical model specification

Comments:
After the December 2001 consortium meeting, it appeared that there was no plans for
such tests at S/S levels. We will thus very likely rely on theoretical predictions of both
the photometric properties of the designed loads, as well as on quasi-optical
characterisation of the CSA optics and expected beam couplings (outcomes of
RD27).




                                           104
                       HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                   Date: 20/10/03
                                                                   Issue: Draft 1.0


2.5.2 UC-1.4.2 Measure internal calibrator coupling

Level:                 User
Scope:                 HIFI-ICC
Version:               0.1
Status:                Draft
Author:                David Teyssier/Michel Pérault - 14/09/01
                       DT/MP - Update 15/01/02
                       DT/MP - Update 15/05/03

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
This use case describes the procedure to measure the mixer beam coupling to the
integrated internal calibrator sources (hereafter CSA, consisting of the Hot and Cold
Black Bodies, HBB and CBB), including hot and cold loads. It aims at knowing what
the radiation temperature seen by the mixers on each load actually is. The frequency
dependence needs to be assessed, as it affects the baseline quality (cf UC-1.1.6).
Note that a 1K uncertainty on the effective load radiation temperature used to
calibrate the data intensity scale translates into a 2% error on the overall calibration
accuracy (assuming a hot load of 100K).

Phase:

ILT                                                                          Y
IST/EET/GST...                                                               Y
LEOP/Commissioning                                                           N
Calibration/PV                                                               N
Science Demonstration                                                        N
Routine Operations                                                           N
Post Operations                                                              N


Actors:
   • CS: Calibration Scientist
   • IE: Instrument (System) Engineer

Triggers:
   • Cal. plan
   • alignment plan ?

Preconditions:
   • availability of adequate sensors to measure physical temperatures of hot and
     cold loads


                                          105
                        HIFI calibration use cases                   Doc.: ICC/2001-005
                                                                     Date: 20/10/03
                                                                     Issue: Draft 1.0


   •   knowledge of optical losses along the concerned paths
   •   knowledge of the CSA characteristics (quasi-optical, temperature behaviour)
       addressed by UC-1.4.1 (likely from model)
   •   availability of absolute photometric references (Hot and Cold)
   •   others..(depending on the measurement procedure still TBD) ?

Minimal post conditions:
   • calibrated measurement of the power detected on either of the loads

Success post conditions:
  • Measure of the efficiency coupling to internal calibrator
  • Accuracy TBD % (requirement: better than 1 K)

Stakeholders and interests:
   • Calibration Group
   • Integration and Test Team
   • System Group
   • Instrument System Engineer

Note:
These measurements are the second phase of the CSA characterization: it is
performed once integrated inside HIFI. The first phase (cf UC-1.4.1) should have
provided a characterisation of the CSA S/S as a black body. The second phase
includes:

1. Laboratory measurements of the CSA coupling to the mixers: the internal loads are
calibrated against an external beam filling source serving as an absolute calibrator. In
order to avoid contamination from the ambient 300K level, the absolute source shall
have an aperture larger than 4 times the waists at the cryostat gate. It will be located
immediately at the cryostat output. In the current set-up, in includes a heat filter at the
cryostat gate consisting of a tilted polyethylene window.

2. Theoretical understanding of the coupling and comparison of a model approach
with the measurements performed on an absolute calibrator. This is the purpose of
the study described in RD27.
This characterisation is of primary importance in order to extrapolate the internal load
coupling to frequency and load temperature ranges that will not be measured for
restricted test time reason. We assume that the internal load has been accurately
defined by a set of parameters to be measured at the S/S level, and that adequate
models are designed. Based on this CSA description, one should be able to design a
model of the expected mixer coupling to the internal loads. This approach relies on a
good knowledge of the mixer beam patterns in the internal calibrator plane, as well as
of the optical losses along the probed path. It assumes that a reasonable
measurement of the beam misalignment is available and that consequences on
beam coupling (better said "un-coupling") can be accurately computed (TBC). The
integrated measurements will be used to globally scale the model.


                                           106
                        HIFI calibration use cases                                     Doc.: ICC/2001-005
                                                                                       Date: 20/10/03
                                                                                       Issue: Draft 1.0



Main success scenario:

Laboratory measurements: The se-tup makes use of two external absolute
photometric references, so-called Absolute Hot Black Body (AHBB, see RD31) and
Absolute Cold Black Body (ACBB). While the AHBB sits at the cryostat gate, the
ACBB currently consists of an absorbing shutter at sink temperature located inside
the cryostat above M3.

   1. IE: Set hot load to required temperature. Wait MM seconds (as derived from
      UC-1.1.2.5)
      Note: If available, this check should be performed for several hot load
      temperatures.
   2. IE: Tune receiver(s)
      Question: How many receiver frequencies do we have to check ? We suggest
      3 to 5 measurements at regularly spaced frequencies for each channels.

   3. IE: External "Hot" and "Cold" load exposure (integration time TBD, likely some
      seconds). The differenced signal writes:

                      ∆V1 = VAHBB − VACBB = γ [η AHBB J eff , AHBB − η ACBB J eff , ACBB ]

      where γ is the instrumental gain, Jeff = GsJs +(1-Gs)Ji is the effective radiation
      temperature seen in both image and signal side-bands, and ηAHBB and ηACBB
      are the global transmission along the respective corresponding optical paths.
      All Rayleigh-Jeans temperatures can be deduced from the known physical
      temperatures of the loads (including emissivity correction).

   4. IE: Internal hot load exposure (integration time TBD, likely some seconds),
      possibly chopped against the external AHBB. The differenced signal writes:
                     ∆V2 = VHBB − VACBB = γ [ηhotη HBB J eff ,HBB − η ACBB J eff , ACBB ]

      where ηHBB is the transmission along the HBB optical path and ηhot represents
      the radiative coupling to the physical temperature featured by the HBB (and
      expressed here by Jeff,HBB).

   5. IE: Compute effective HBB radiation temperature:

                                         ∆V1 η AHBB J eff , AHBB − η ACBB J eff , ACBB η ACBB
                     ηhot J eff ,HBB =       ×                                        +       J
                                         ∆V2                   η HBB                    η HBB eff , ACBB

      If required, isolate ηhot.

      Note: In order to derive the effective HBB temperature, one needs to make
      some assumptions on the optical transmissions involved in the equation. While


                                                     107
                           HIFI calibration use cases                                   Doc.: ICC/2001-005
                                                                                        Date: 20/10/03
                                                                                        Issue: Draft 1.0


      it is likely justified to assume ηHBB ≈ ηACBB, there might be extra transmission
      losses induced by the polyethylene window isolated the AHBB from the
      cryostat. Can this loss be estimated (likely frequency dependent) ?

   6. IE: Repeat step 4 for CBB.

                       ∆V2 = VCBB − VACBB = γ [ηcoldηCBB J eff ,CBB − η ACBB J eff , ACBB ]

      and

                                     ∆V1 η AHBB J eff , AHBB − η ACBB J eff , ACBB η ACBB
                ηcold J eff ,CBB =       ×                                        +       J
                                     ∆V2                   ηCBB                     ηCBB eff , ACBB

   If required, isolate ηcold.

   Note: again, some assumptions on the transmissions need to be made.

   Computational/modelling work (we can use the parameters characterising the
   internal loads and design/validate a model of the coupling to the mixers from a
   comparison between expected values and measured coupling as given by the
   external calibrator)

   1. CS: Choose internal calibrator physical temperatures (hot and cold),
   2. 2.CS: Compute expected radiation temperature from model
   3. 3.CS: if absolute measurement (Sub-scenario 1) available at this
      frequency/temperature:
             a. CS: Compare to absolute measurement
             b. CS/IE: (if applicable) adjust model
             c. CS: Return "exact" internal load temperatures and store in
                 calibration data-base
   4. 4.CS: if no absolute measurement available at this frequency/temperature:
             d. CS: Extrapolate from the model
             e. CS: Return "exact" internal load temperatures and store in
                 calibration data-base

Frequency of occurrence: In ILT/AIV

Open issues:
  • basis of the modelling effort
  • which black bodies to be used, location...

Comments: See note above.




                                                      108
                       HIFI calibration use cases                Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


2.5.3 UC 1.4.3 Measure Telescope Aperture Efficiency

Level:                           User
Scope:                           HIFI-ICC
Version:                         0.1
Status:                          Draft
Author:                          Maryvonne Gerin - 31/08/01
                                 Update D. Teyssier – 15/05/03

General Grade: High
Grade for ILT: Low
Grade for Implementation: Low

Brief description:
This use case describes the procedure to measure the instrument aperture efficiency
at different frequencies using astronomical sources, or from theoretical computation.

Phase:
ILT                                                                            N
IST/EET/GST...                                                                 N
LEOP/Commissioning                                                             Y
Calibration/PV                                                                 Y
Science Demonstration                                                          Y
Routine Operations                                                             Y
Post Operations                                                                Y
(post-mortem or non-dedicated observations)

Actors:
   • CS: Calibration Scientist

Triggers:
   • Cal. Plan
   • contingency

Preconditions:
   • Calibrated Hot and Cold loads (known effective radiation temperatures). The
     Hot load temperature should preferably be larger or similar to the expected
     brightness temperature of the continuum source.
   • Availability of adequate calibration sources in-flight: strong continuum POINT
     sources.
   • For theoretical aspects: quasi-optical description of the whole optics

Minimal post conditions:
   • Set of required measurements successfully performed

Success post conditions:


                                         109
                        HIFI calibration use cases               Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


   •   Aperture efficiencies derived at a TBD number of frequencies with a 10%
       (TBC) accuracy

Stakeholders and interests:
   • Calibration group and astronomers

Main success scenario:

In-orbit measurement:

   1. CS: Retrieve a point source and information on the expected continuum
      intensity at the peak from calibration database
   2. CS: Choose observed frequencies
   3. CS: Choose a backend and adjust backend properties (bandwidth, resolution)
      to the expected signal
   4. CS: Perform an intensity calibration /sensitivity (measure Trec and Tsys)
   5. CS: Peak-up on target
      Note: It is absolutely necessary that the source is well aligned with the
      telescope.
   6. CS: integrate on target in double beam switching (switching rate likely 3.5Hz to
      optimise effective noise bandwidth). Time estimates are discussed in RD21.
   7. CS: Perform adequate data reduction to determine the continuum level and
      compare with the expected signal for a perfect system. Equations are
      described in RD21. Derive aperture efficiency accordingly
   8. CS: Store results in the calibration data base
   9. CS: File in a calibration report

Theoretical predictions:

   Predictions of the aperture efficiencies could be provided by the quasi-optical
   characterisation of HIFI described in RD27. Ideally, model could be run in order to
   compute efficiencies at any frequencies. These predictions rely on an adequate
   description of the telescope though.

Frequency of occurrence:
   • in PV and for health monitoring, but also likely each time an adequate source
     is visible

Open issues:
  • How many frequencies for this measurement ? A possible approach consists
      in considering the most observed frequencies by HIFI (highest priorities as
      compiled in RD23).
  • Another possible approach consists in doing measurements at 3-4 frequencies
      per band at regular intervals, and use the theoretical predictions to connect
      these points across the bands.




                                         110
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


   •   The measurement may be done on bright planets only provided they have
       very small size w.r.t. the beam (Mars, Uranus), or even asteroids. Saturation
       of the signals must not be a concern if one want to ensure accurate
       measurement.
   •   In case the source is considered an absolute photometric calibrator, the
       measurements described in UC-1.2.2 also allows to derive the aperture
       efficiency.

Comments:
  .
  • Comparisons of theoretical predictions with in-orbit measurements should
    address the issue of whether theory or observations provide the lowest
    uncertainties in computing this efficiency
  • A companion document should present reference spectral energy distribution
    for the astronomical sources (bright planets), and formulae for calculating the
    expected continuum signal from these astronomical sources viewed with a
    perfect telescope with a given beam.




                                         111
                       HIFI calibration use cases                Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


2.5.4 UC-1.4.4 Measure telescope beam efficiency

Level:                           User
Scope:                           HIFI-ICC
Version:                         0.1
Status:                          Draft
Author:                          David Teyssier – 15/05/03

General Grade: High
Grade for ILT: Low
Grade for Implementation: Low

Brief description:
This use case describes the procedure to measure the instrument beam efficiency at
different frequencies using astronomical sources, or from theoretical computation.

Phase:
ILT                                                                            N
IST/EET/GST...                                                                 N
LEOP/Commissioning                                                             Y
Calibration/PV                                                                 Y
Science Demonstration                                                          Y
Routine Operations                                                             Y
Post Operations                                                                Y
(post-mortem or non-dedicated observations)

Actors:
   • CS: Calibration Scientist

Triggers:
   • Cal. Plan
   • contingency

Preconditions:
   • Calibrated Hot and Cold loads (known effective radiation temperatures). The
     Hot load temperature should preferably be larger or similar to the expected
     brightness temperature of the continuum source.
   • Availability of adequate calibration sources in-flight: strong continuum sources
     of size comparable to the beam.
   • For theoretical aspects: quasi-optical description of the whole optics

Minimal post conditions:
   • Set of required measurements successfully performed

Success post conditions:



                                         112
                        HIFI calibration use cases               Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


   •   Beam efficiencies derived at a TBD number of frequencies with a 10% (TBC)
       accuracy

Stakeholders and interests:
   • Calibration group and astronomers

Main success scenario:

In-orbit measurement:

   1. CS: Retrieve a point source and information on the expected continuum
      intensity at the peak from calibration database
   2. CS: Choose observed frequencies
   3. CS: Choose a backend and adjust backend properties (bandwidth, resolution)
      to the expected signal
   4. CS: Perform an intensity calibration /sensitivity (measure Trec and Tsys)
   5. CS: Peak-up on target
      Note: It is absolutely necessary that the source is well aligned with the
      telescope.
   6. CS: integrate on target in double beam switching (switching rate likely 3.5Hz to
      optimise effective noise bandwidth). Time estimates are discussed in RD21.
   7. CS: Perform adequate data reduction to determine the continuum level and
      compare with the expected signal for a perfect system. Equations are
      described in RD21. Derive beam efficiency accordingly.
   8. CS: Store results in the calibration data base
   9. CS: File in a calibration report

Theoretical predictions:

   Predictions of the beam efficiencies could be provided by the quasi-optical
   characterisation of HIFI described in RD27. Ideally, model could be run in order to
   compute efficiencies at any frequencies. These predictions rely on an adequate
   description of the telescope though.


Frequency of occurrence:
   • in PV and for health monitoring, but also likely each time an adequate source
     is visible

Open issues:
  • How many frequencies for this measurement ? A possible approach consists
      in considering the most observed frequencies by HIFI (highest priorities as
      compiled in RD23).
  • Another possible approach consists in doing measurements at 3-4 frequencies
      per band at regular intervals, and use the theoretical predictions to connect
      these points across the bands.



                                         113
                      HIFI calibration use cases                Doc.: ICC/2001-005
                                                                Date: 20/10/03
                                                                Issue: Draft 1.0


   •   The measurement may be done on bright planets only provided they have a
       comparable size w.r.t. the beam (Mars, Uranus). Saturation of the signals
       must not be a concern if one want to ensure accurate measurement.
   •   In case the source is considered an absolute photometric calibrator, the
       measurements described in UC-1.2.2 also allows to derive the beam
       efficiency.

Comments:
  .
  • Comparisons of theoretical predictions with in-orbit measurements should
    address the issue of whether theory or observations provide the lowest
    uncertainties in computing this efficiency
  • A companion document should present reference spectral energy distribution
    for the astronomical sources (bright planets), and formulae for calculating the
    expected continuum signal from these astronomical sources viewed with a
    perfect telescope with a given beam.




                                        114
                      HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0




2.5.5 UC-1.4.5 Intensity calibration

Level:                           User
Scope:                           HIFI-ICC
Version:                         0.1
Status:                          Draft
Author:                          Thierry Jacq - 17/09/01
                                 Update d. Teyssier – 15/05/03

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
This use case calibrates the group receivers + spectrometers on loads serving as
known references. It aims at scaling spectrometers outputs given for each channel in
spectrometers units to values given in "physical" units like Kelvin. Astronomical
calibrations are to be applied afterwards using dedicated efficiencies calibrated
against celestial bodies.

Phase: all phases

Actors:
   • IE: Instrument (System) Engineer
   • CS: Calibration Scientist
   • CM: Calibration Manager
   • DP: Data Processing Pipeline

Triggers:
   • Cal. Plan
   • Proposal observing sequence

Preconditions:
   • Working spectrometers
   • Known effective load radiation temperatures (completion of UC-1.4.2)
   • Working interactive analysis system
   • Known side-band ratios at the frequencies of interest (completion of UC-1.3.4)

Minimal post conditions:
   • Set of required spectra successfully measured

Success post conditions:
  • Both HRS and WBS spectra are calibrated at the instrument level



                                        115
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


   •   If applicable, derive telescope forward efficiency ηl (assuming a known
       telescope temperature)

Stakeholders and interests:
   • Calibration Group
   • General astronomers
   • Instrument System Engineer
   • Herschel System Group


Main success scenario:

The detailed discussion of the intensity calibration of HIFI is given in RD20. It
proposes a calibration scheme allowing to cope with the presence of standing waves
in the system, as well as with the fact that responses differ in the respective image
and signal side-bands. This scheme consists of two calibration steps, namely a so-
called load calibration aimed at calibrating the receiver band-pass and using two
internal loads, and a so-called OFF or baseline calibration (see RD20 for details)
making use of a blank sky integration to correct from standing waves affecting the
spectra. RD20 extensively discusses uncertainties and time estimates associated to
each of these steps.

   1. CS: Perform load calibration
            a. Set hot load to the required temperature (highest temperature are
               likely to decrease uncertainties) and get hot and cold load effective
               radiation temperatures from database (output of UC-1.4.2)
            b. Integrate on Hot and Cold loads. See integration time estimates in
               RD20. They are designed so that the load measurement contribute
               to less than 1% in the total spectra noise.
            c. Compute receiver band-pass γlrec and receiver temperature Jlrec (see
               RD20)
            d. Compute uncertainties on the above values
   2. CS: Perform OFF-calibration
            a. Integrate on a blank sky position, likely close to the target.
               Integration time are discussed in RD20. They are based on stability
               times in the instrument.
            Note: this OFF measurement is also required to measure Tsys and is
            assimilated to the Reference position in observing modes such as
            Position Switching and OTF.
            b. Use adequate equations of RD20 to investigate nature of standing
               waves and derive the contribution from the telescope pick-up (1-
               ηl)Jeff,T (see RD20 for details). If possible, extract the forward
               efficiency ηl.
            Note: this investigation requires spectral smoothing utilities to be
            available in IA.
            c. Compute uncertainties on the above value



                                         116
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


   3. CS: Calibrate spectra
              a. If applicable, smooth OFF-calibration spectrum to the adequate
                 “standing wave” frequency resolution (see RD20)
              b. Combine load-calibration and OFF-calibration measurements to
                 obtain a calibrated spectrum (so far in the HIFI proper temperature
                 scale, can this be called T A * ?). See equations in RD20.
              c. Compute global calibration uncertainty of the measurement
   4. CS: Calibrate into an absolute temperature scale
      Note: this is a non-trivial exercise depending on the source brightness
      temperature distribution with respect to the telescope beam. In RD20, this is
      accounted via a global coupling factor labelled ηsf. The conservative approach
      will consist in scaling observations in Tmb scale and account for source dilution
      if applicable (case of source size smaller than the beam size)
              a. Scale spectrum to Tmb scale (multiply by ηl/ηmb). See RD21 for
                 details.
              Note: This factor is a possible approximation of the general source
              coupling efficiency ηsf used in RD20.
              b. For extending sources, correcting algorithm TBD
   5. CS: Compute final calibration uncertainty accounting for coupling coefficient
      errors.

Frequency of occurrence:
   • Each time a calibration sequence is requested. This will depend on the
     observing mode sequencing. We here refer to RD32.

Open issues:
  • should "bad channels" be checked for here for marking ?
  • see RD20 for additional comments

Comments:




                                         117
                        HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                    Date: 20/10/03
                                                                    Issue: Draft 1.0




2.5.6 UC-1.4.6 Measure dynamic range

Level:                        User
Scope:                        HIFI-ICC
Version:                      0.1
Status:                       Draft
Author:                       Laurent Ravera/Moncef Belgacem - 17/09/01

General Grade: High
Grade for ILT: High
Grade for Implementation: High.

Brief description:
This use case describes the procedure to measure the HIFI dynamic range with the
HRS. The dynamic range is checked by a measure of the sensitivity at the limits of
the input range.

Phase:

ILT                                                                                Y
IST/EET/GST...                                                                     Y
LEOP/Commissioning                                                                 Y
Calibration/PV                                                                     Y
Science Demonstration                                                              Y
Routine Operations                                                                 Y
Post Operations                                                                    N
(post-mortem or non-dedicated observations)

Remark:
The dynamic limitation is not the same in the HRS and in the WBS.
This UC should be divided in two parts:
      UC-1.4.5.1 Measure HIFI dynamic range with the HRS
      UC-1.4.5.2 Measure HIFI dynamic range with the WBS.

Question:
What is the HIFI dynamic range?
  • the range of the input power for a given setting of the instrument (a given HRS
      attenuator setting, ...). This range is the limitation during a single observation.
  • the full input power range (taking into account that the HIFI settings can be
      changed). This range is the limitation for all the observations

Actors:
   • IE
   • CS



                                          118
                    HIFI calibration use cases   Doc.: ICC/2001-005
                                                 Date: 20/10/03
                                                 Issue: Draft 1.0



Triggers:
   • Cal. plan

Preconditions:

Minimal post conditions:

Success post conditions:

Stakeholders and interests:
   • Calibration Group
   • General astronomers
   • Instrument System Engineer
   • Integration and Test Team

Main success scenario:

Frequency of occurrence:

Open issues:

Comments:
NW to comment on the questions.




                                  119
                      HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


2.5.7 UC-1.4.7 Measure non-linearity

Level:                       User
Scope:                       HIFI-ICC
Version:                     0.1
Status:                      Draft
Author:                      Laurent Ravera/Moncef Belgacem - 17/09/01

General Grade: High
Grade for ILT: High
Grade for Implementation: High

Brief description:
This use case describes the procedure to measure the HIFI non-linearity.
The non-linearity will be measured via the HRS and the WBS. The result can be
different between the two spectrometers.

Phase:

ILT                                                                           Y
IST/EET/GST...                                                                Y
LEOP/Commissioning                                                            Y
Calibration/PV                                                                Y
Science Demonstration                                                         Y
Routine Operations                                                            Y
Post Operations                                                               N
(post-mortem or non-dedicated observations)


Actors:

   •   IE
   •   CS

Triggers:
   • Cal. plan

Preconditions:
   • Availability of 3 adequate sources (S1, S2, S3) with known power levels in a 3
     dB range (Pin_3 ~ Pin_2+1.5dB ~ Pin_1+3dB) (with a TBD accuracy).
     Note: The HIFI specification is a non-linearity smaller than 1% in a 3dB range.

Question: where in the HIFI input power range do we have to do this test? Is non-
linearity most expected at high power levels ?

Minimal post conditions:


                                        120
                      HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


   •   Outputs successfully measured at all three power levels

Success post conditions:
  • Level of non-linearity measured within the considered input power range with a
     better than 1% accuracy (TBC)

Stakeholders and interests:
   • Calibration Group
   • General astronomers
   • Instrument System Engineer
   • Integration and Test Team

Main success scenario:

   1. CS: Run the "tune correlator input power" routine with the source S2 (about
      the middle of the 3dB range) (for the HRS only).
      Question: what is the equivalent command for the WBS (“set attenuators” ?)

   2. CS: repeat on each source (i=1,2,3):
      • Do an integration on source Si.
        Note: as only the total power is measured, a short integration time (value
        TBD) is sufficient.
      • Compute the total power measured by the spectrometer (Pout,i).

   3. CS: Display the output power as a function of the input power:
      • Pout,i =fct(Pin,i)
      • Measure the non-linearity. This is likely done through estimates of the
        slopes at low and high input power level respectively. Extrapolation of the
        low input power level slope to the high input power level domain should
        result is a less than 1% discrepancy.

Frequency of occurrence:

Open issues:

Comments:




                                        121
                       HIFI calibration use cases                Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0




2.6   UC-1.5 Validation of observing modes

Level:                 Summary
Scope:                 HIFI-ICC
Version:               0.1
Status:                Draft
Author:                Frank Helmich- 16/09/01

General Grade: High
Grade for ILT: Low
Grade for implementation: Medium

Brief description:
In this use case the validation of the observing modes is discussed. By validation we
understand: check that a given observing mode is behaving as expected, fills in the
properties for which it was chosen (e.g. capacity of measuring continuum, mapping),
and eventually that it is indeed calibratable (although that should have been
assessed before and reported in RD17 and RD32). Details on the description, the
implementation and the calibration of each observing mode can be found in RD35,
RD36 and RD37 respectively.

Phase:
ILT                                                                           N
IST/EET/GST...                                                                N
LEOP/Commissioning                                                            Y
Calibration/PV                                                                Y
Science Demonstration                                                         Y
Routine Operations                                                            N
Post Operations (post-mortem or non-dedicated observations)                   Y

Actors:
CS: Calibration Scientist

Triggers:
   • Calibration Plan
   • HIFI Scientists and general astronomers
   • Ground Segment

Preconditions:
   • availability of known sources and known line fluxes from ground-based
     observations at the Herschel beam-sizes.
   • availability of fitting routines for lines
   • availability of map making procedures




                                        122
                     HIFI calibration use cases              Doc.: ICC/2001-005
                                                             Date: 20/10/03
                                                             Issue: Draft 1.0


Minimal post-conditions
   • Validation of the observing modes at a 10% (TBC accuracy level)

Success post conditions:
  • Validation of the observing modes at 5% (TBC) level. Confidence should exist
     that the modes deliver what they are supposed to deliver, i.e. stable base
     lines, stable pointing, fast scanning, etc.

Stakeholders:
   • Calibration Group
   • Instrument System Group
   • General astronomers
   • Ground Segment

Main success scenario:
  1. CS: Create the required observations. Modes considered so far are:
            • UC-1.5.1: Validation Double Beam Switching mode
            • UC-1.5.2: Validation On-The-Fly mapping mode
            • UC-1.5.3: Validation load-chop with OFF calibration
            • UC-1.5.4: Validation Frequency survey with chopper
  2. CS: Observe
  3. CS: Analyse the spectra and/or maps
  4. CS: update calibration data base and file reports
  5. CS: Trigger higher level analysis
  6. CS: If observing mode cannot be validated, abandon this mode from general
      use

Frequency of occurrence
   • Mainly in PV and maybe the first 2 months of observing. Sometimes checks
     will have to be repeated, sometimes general observations will be used.

Open issues:

Comments:




                                      123
                      HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0




2.6.1 UC 1.5.1 Validation Double Beam Switching Mode

Level:                  User
Scope:                  HIFI-ICC
Version                 0.1
Status:                 Draft
Author:                 Frank Helmich 28-01-2002

General Grade: High
Grade for ILT: Medium
Grade for implementation: Medium

Brief description:
This use case describes the validation of the Double Beam Switching (DBS) mode.
At the time of writing the double beam-switch mode is the same as the "nodding" in
the ESA pointing document. While doing the slow (=normal) chopped observations,
the chopper-off beam is moved from the off position to the source-on position.
This mode can be validated against ground-based observations (in a map or beam
similar to HSO's beam), or through mapping of HIFI itself. This mode is validated
when the resulting spectra are within a few percent with HIFI maps or within 20% with
a ground based map (absolute) or within 5% with a the relative intensities of the
ground-based map (in this latter case emission should extend over TBC arcmin.).
Note that this mode can only be validated in-orbit.

Phase:

ILT                                                                           N
IST/EET/GST...                                                                N
LEOP/Commissioning                                                            Y
Calibration/PV                                                                Y
Science Demonstration                                                         Y
Routine Operations                                                            Y
Post Operations                                                               N

Actors:
AS: General Astronomer
CS: Calibration Scientist
IE:   Instrument Engineer


Triggers:
   • Calibration Plan
   • Observing modes definition

Preconditions:


                                        124
                       HIFI calibration use cases                Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


   •   Functional Common Science System
   •   Accurate a posteriori Pointing knowledge
   •   Ground-based maps in strong (CO) lines covering at least the 3 arcmin
       chopper throw, although the emission may be a lot more compact.

Minimal post conditions:
   • Agreement with HSO maps within 10% and with ground-based maps within
     20%.

Success post conditions:
  • Agreement with HSO maps within 3% (TBC) and with ground-based maps
     within10%.

Stakeholders and Interests:
   • AIV engineer
   • System engineer
   • Calibration Group
   • General Astronomers

Main success scenario:
  1. CS/AS: Create observation for raster map. Use oversampled (Nyquist/3) and
      integration such that the noise level is 5 times lower than 1/50th (TBC) of the
      peak flux (take values from ground-based observations). This guarantees that
      we can validate the mode internally for HIFI within the goal 3%. The map size
      can be 3 beams wide, but unknown chop direction probably requires that we
      make a square map. With a chopper-throw of maximum 3 arcminutes, the
      map-size in the chopper direction is at least 3 arcminutes. Use the WBS and
      the HRS together. Make sure that our main line is not in between two WBS
      arrays.
  2. HCSS: Schedule this observation and put data in the archive.
  3. CS/AS: Create double beam-switch observation. Throw will be the maximum
      and one throw halfway. Care should be taken that the single pointings have
      their counterpart within the raster map, to avoid smoothing effects. Use high
      signal-to-noise ratio and thus large integration times. Both WBS and HRS will
      be used. The frequency settings must be exactly the same as for the raster
      map.

   Question: Should the integration time for map and for DBS be exactly the same?

   4. HCSS: Schedule the observation and put data in the archive
   5. CS/AS: Retrieve data from the archive
   6. CS/AS: Do normal calibration steps for WBS and HRS data. (Zero, comb,
      hot/cold, windowing etc.) Intercompare the two results from the two
      spectrometers. Check that peak fluxes agree, when smoothed to the same
      resolution. This should agree within better than 1%.




                                         125
                      HIFI calibration use cases                Doc.: ICC/2001-005
                                                                Date: 20/10/03
                                                                Issue: Draft 1.0


   7. CS/AS: Compare data with the raster map. Checks should be made that the a
      posteriori pointings are the same for map and DBS observation. Overlay DBS
      spectrum over spectrum from map point that is closest to the DBS
      observation. Compare also with map value at exactly the right location.
      Repeat procedure with spectrum with half the chopper throw. All results should
      agree within about 1%
   8. CS/AS: Compare the results with ground-based maps. It is preferred that the
      ground-based map is created in OTF-mode, to avoid mispointings and drifts in
      ground-based observations. The signal-to-noise ratio is only important in the
      inner part of the map, where the HSO main beam must be carefully mapped.
      Strong lines are favoured. Outflows hinder the comparison, but are likely not
      critical in the lower CO-lines. The HIFI DBS observation and ground-based
      observation should agree within 20%
   9. CS/AS: File extensive report. Consider the mode validated whenever the
      internal agreement HRS-WBS is within 1%; Map-DBS within 2% and Ground
      Based-HIFI DBS are within 20%

Open Issues:
  • Should validation be done for every LO subband? I don't think so, but on the
      other hand this mode is the most important one, so it better be good.
  • Ground-based maps in the higher frequencies can only be done with Sofia or
      SWAS. Intercomparison will be difficult because of differences in beam size.




                                        126
                      HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0




2.6.2 UC-1.5.2 Validation On-the-fly Mapping

Level:                User
Scope:                HIFI-ICC
Version:              0.2
Status:               Draft
Author:               Carsten Kramer - 24/07/01 - Update - 16/09/01
                      Update D. Teyssier – 08/08/2003

Description:
This use case describes the validation of the spectra-line On-the-Fly observing mode.
In this mode, the telescope observes a complete area centred on a given field centre.
After observing an OFF source reference position, the telescope slews to the source
and then scans it in one coordinate at a constant speed. During scanning, data are
acquired (dumped) at a constant rate. The other coordinate is kept constant during
each scan. Any scanning direction can be chosen and a full map is compiled by
observing one scan line after the other.

Restrictions (as of RD10):
The maximum size of OTF maps is 2 degrees times 32 lines at a maximum
separation of 8 arcmin. The scanning velocity can be varied between 0.1"/sec and
1'/sec.

Actors:
AS: General Astronomer
CS: Calibration Scientist
IE:   Instrument Engineer

Triggers:
   • Calibration Plan
   • Observing modes definition

Preconditions:
   • Functional Common Science System
   • A priori knowledge of required calibration parameter (as listed in RD37)
   • Availability of ground-based spectra from adequate sources (as assessed by
     RD34)

Minimal post conditions:
   • Agreement with HSO data within 10% and with ground-based data within 20%.

Success post conditions:
  • Agreement with HSO data within 3% (TBC) and with ground-based data within
     10%.


                                        127
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0



Stakeholders and Interests:
   • AIV engineer
   • System engineer
   • Calibration Group
   • General Astronomers

Main success scenario:

   1. CS/AS: Choose an adequate source. This observing mode is foreseen for
      extended sources with very likely structure. Ideal targets must have been
      identified in RD34, and ground-based preparatory observations at the low HIFI
      frequency should be available
   2. CS/AS: Create observation for On-The-Fly mapping. RD36 for that purpose.
      We consider here that OFF positions are observed via a slew of the telescope
      (see comments for the alternatives). Both WBS and HRS will be used.
   3. HCSS: Schedule this observation and put data in the archive.
   4. CS/AS: Retrieve data from the archive
   5. CS/AS: Do normal calibration steps for WBS and HRS data. (Zero, comb,
      hot/cold, windowing etc.) Details are given in RD37.
          o Project the data on a regular grid and check reconstructed map centre
              coordinate with respect to the aimed position (check of pointing)
          o Check in particular the capability of suppressing standing waves
          o Investigate the optimal way to use OFF measurements performed
              between the rows (time average, weighted mean, etc. See RD37).
   6. CS: Inter-compare the two results from the two spectrometers. Check that
      peak fluxes agree, when smoothed to the same resolution. This should agree
      within better than 1% (TBC).
   7. CS/AS: Compare the results with ground-based data. Take the respective
      beam resolutions into account for an accurate comparison. For extended
      sources it is already anticipated that some data will be affected by severe error
      beam pick-up on some ground-based telescopes. However, this may very
      much depend on the instrument used to gather the ground-based reference
      maps. The HIFI Load-Chop observations and ground-based observation
      should agree within 20% (TBC). Among the things to check are:
          o Noise distribution across the map, and comparison to expected noise
              level. Check predictions from RD37
          o Consistency with structures as exhibited by the ground-based map
          o Existence of stripes (indicating uncorrected gain variations in the course
              of the mapping) or other distortions in the map (e.g. consecutive of
              beam smearing). If existing, check the capability of dedicated
              algorithms (e.g. Emerson 1988) to clean them. If of any sense, perform
              the map again in the orthogonal direction.
          o Check spectral shape of the profiles
   8. CS/AS: File extensive report. Consider the mode validated whenever the
      internal agreement HRS-WBS is within 1%, when two HIFI maps are within,
      and Ground Based-HIFI maps are within 20%


                                         128
                       HIFI calibration use cases                Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0



Comments:
  • Simultaneous Beam Switching or Frequency switching may be specified
    independently.
  • Taking OFF source scans before and after the ON source OTF scans may
    reduce the slewing times of the observatory. This technique is used at several
    ground-based telescopes, among which the IRAM 30-m and at the CSO
    telescopes. For data reduction there should be full flexibility to use a weighted
    average of OFF scans. For reduction of each ON source dump, the two OFF
    source scans may for example be weighted according to their distance in time
    to the ON source dump. I.e. an ON source dump following directly the first
    OFF scan is almost entirely calibrated with the first OFF scan, while an ON
    source dump at mid position during an OTF scan is calibrated by the average
    of both OFF positions. These issues are however addressed in RD37.

References:




                                         129
                      HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                 Date: 20/10/03
                                                                 Issue: Draft 1.0


2.6.3 UC 1.5.3 Validation load-chop with OFF calibration

Level:                User
Scope:                HIFI-ICC
Version:              0.1
Status:               Draft
Author:               David Teyssier – 08/08/03

General Grade: High
Grade for ILT: Medium
Grade for implementation: Medium

Brief description:
This use case describes the validation of the load-chop observing mode. In this
mode, the source emission is modulated against the internal cold load, then an OFF
position is observed, again against the cold load, in order to perform an baseline
calibration and correct from potential standing waves (see description in RD32). Note
that this mode is the only one that can be mimicked from the ground, expect from any
effect arising from M2.

Phase:

ILT                                                                           Y
IST/EET/GST...                                                                N
LEOP/Commissioning                                                            Y
Calibration/PV                                                                Y
Science Demonstration                                                         Y
Routine Operations                                                            Y
Post Operations                                                               N

Actors:
AS: General Astronomer
CS: Calibration Scientist
IE:   Instrument Engineer

Triggers:
   • Calibration Plan
   • Observing modes definition

Preconditions:
   • Functional Common Science System
   • A priori knowledge of required calibration parameter (as listed in RD37)
   • Availability of ground-based spectra from adequate sources (as assessed by
     RD34)
   • Availability of laboratory sources and of simulated spectrum for the gasses of
     interest during ILT tests


                                        130
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0



Minimal post conditions:
   • Agreement with HSO data within 10% and with ground-based data within 20%.

Success post conditions:
  • Agreement with HSO data within 3% (TBC) and with ground-based data within
     10%.

Stakeholders and Interests:
   • AIV engineer
   • System engineer
   • Calibration Group
   • General Astronomers

Main success scenario:

- ILT:

This observing mode can be mimicked using the gas cell fed with a Hot/Cold switch.
Any of the gas foreseen with the gas cell (from scarce to rich spectrum) can be used
a priori. Pressure should be of the order of some mbars (TBC). While the hot source
will provide an emission background stimulating the source when the cell is filled with
gas, the cold source mimics the OFF position, acting as a blank sky when the cell is
empty. The time to empty the cell might be representative of the time required to
perform the slew to the OFF position once in flight.

   1. IE: Set up gas cell
   2. IE/CS: Set up observation. This includes the chopper rate, and the respective
      times to be spent on the various phases. They somewhat depend on the dead-
      times in the observation scheme, so they may need to be adapted to the
      laboratory condition w.r.t. to what is anticipated in space. Details are to be
      found in RD36.
   3. IE/CS: Tune to the LO frequency.
      Question: do we expect a dependence on frequency ? In any case, what
      frequency(ies) should be used ?

   4. IE/CS: Perform an internal calibration (internal Hot/Cold). See times in RD36.
   5. IE/CS: Fill in the cell at the required pressure and switch to the Hot load
   6. IE/CS: Modulate the Hot source emission seen through the filled cell against
       the internal cold load
   7. IE/CS: Empty the cell (takes up to 1 min., TBC) and switch to the Cold load
   8. IE/CS: Modulate the Cold source emission seen through the empty cell
       against the internal cold load
   9. IE/CS: Calibrate the data: see recipes in RD37. If necessary, perform a side-
       band de-convolution
   10. CS: Examine the spectra visually.



                                         131
                        HIFI calibration use cases                    Doc.: ICC/2001-005
                                                                      Date: 20/10/03
                                                                      Issue: Draft 1.0


                a. Overlay simulated spectrum and HIFI SSB (single side-band)
                   spectrum.
                b. Determine peak ratios for the same lines from the two different
                   (model vs. HIFI) spectra.
                c. Check the line-frequencies against catalogue (e.g.
                   http://spec.jpl.nasa.gov or http://www.cdms.de) and against
                   simulated spectrum. Use fits to the lines to determine the precise
                   frequency. Explain differences.
                d. Check for ghosts and other spurious signals in the HIFI spectrum.
                   Trace spurious signals in the original scans. If applicable, check if
                   MEM creates spurious contributions. Vary the MEM input
                   parameters (side-band ratios, noise levels, etc.). Rerun the MEM
                   code. Re-inspect etc.
                e. File report, with inclusion of all spurious signals, all determined peak
                   ratios, all MEM settings etc.
    11. CS: If the simulated and the HIFI spectra are well in agreement with each
        other (within uncertainties of the model spectrum), consider this mode
        validated.

- In-orbit:

    9. CS/AS: Choose an adequate source. This observing mode is foreseen for
        sources with no obvious close-by emission free OFF position. Extended
        sources with e.g. known peak(s) could be of interest, as well as targets close
        to the Galactic Plane. Source with a known continuum could also be of interest
        as this mode should be able to measure continuum level (TBC). For the sake
        of simplicity, the spectra should not be too rich (TBC).
    10. CS/AS: Create observation for load-chop with OFF calibration. Use RD36 for
        that purpose. Preferably, aim at Use high signal-to-noise ratio and thus large
        integration times. Both WBS and HRS will be used.
    11. HCSS: Schedule this observation and put data in the archive.
    12. CS/AS: Retrieve data from the archive
    13. CS/AS: Do normal calibration steps for WBS and HRS data. (Zero, comb,
        hot/cold, windowing etc.) Details are given in RD37. Check in particular the
        capability of suppressing standing waves as this mode implies combination of
        measurement performed through different optical paths.
    14. CS: Inter-compare the two results from the two spectrometers. Check that
        peak fluxes agree, when smoothed to the same resolution. This should agree
        within better than 1%.
    15. CS/AS: Compare the results with ground-based data. It is anticipated that the
        ground-based data were obtained in Position Switching is the source was
        extended. Take the respective beam resolutions into account for an accurate
        comparison. For extended sources this may result complex as some ground-
        based telescopes have severe error beams at their highest frequencies of
        operation (coincident with the HIFI ones). The HIFI Load-Chop observations
        and ground-based observation should agree within 20% (TBC)



                                           132
                     HIFI calibration use cases              Doc.: ICC/2001-005
                                                             Date: 20/10/03
                                                             Issue: Draft 1.0


   16. CS/AS: File extensive report. Consider the mode validated whenever the
       internal agreement HRS-WBS is within 1% and Ground Based-HIFI DBS are
       within 20%

Open Issues:
  • Should validation be done for every LO subband? In other word, should there
      be a dependence on the frequency ? I don't see why, but on the other hand
      this mode is a fallback mode, so it might be safe.




                                      133
                      HIFI calibration use cases                Doc.: ICC/2001-005
                                                                Date: 20/10/03
                                                                Issue: Draft 1.0




2.6.4 UC 1.5.4 Validation Frequency survey with chopper

Level:                User
Scope:                HIFI-ICC
Version:              0.1
Status:               Draft
Author:               Frank Helmich 28-01-2002
                      Up-date D. Teyssier 06-08-2003

General Grade: High
Grade for ILT: Medium
Grade for implementation: Medium

Brief description:
This use case describes the steps to take to derive a single side-band spectrum from
several DSB scans within 1 LO-band. In fact this is more a validation use case than
anything else, although the output should be usable for UC 1.3.4 "Determine Side-
band Ratio" and UC 1.3.1.4 "Instrument Frequency Calibration". However, especially
the latter needs detailed astronomical knowledge of the source. Note that the
treatment of the spectra is very similar to that of UC 1.3.4.

Phase:
ILT                                                                           Y
IST/EET/GST...                                                                N
LEOP/Commissioning                                                            Y
Calibration/PV                                                                Y
Science Demonstration                                                         Y
Routine Operations                                                            Y
Post Operations                                                               N

Actors:
AS: General Astronomer
CS: Calibration Scientist
IE:   Instrument Engineer

Triggers:
   • Calibration Plan
   • Observing modes definition

Preconditions:
   • Functional Common Science System
   • A priori knowledge of required calibration parameter (as listed in RD37)
   • FTS spectrum of the HIFI frequency range with a water (or other gas, TBD)
     spectrum of the right pressure for comparison with the HIFI spectrum Spectral



                                        134
                       HIFI calibration use cases                 Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


      survey from the ground in selected area. In strong lines maps should be
      available.

Minimal post conditions:
   • Single side-band spectrum for a LO sub-band. Ghost features below 5% of
     strongest line (TBD)

Success post conditions:
  • Single side-band spectrum for a LO sub-band. Ghost features below 1% of
     strongest line (TBD)

Question: should this use case aim for glueing all the LO sub-bands to one
composite spectrum. This may mean thinking about all the resample steps.

Stakeholders and Interests:
   • AIV engineer
   • System engineer
   • Calibration Group
   • General Astronomers

Main success scenario:

 - ILT:
We will use a gas cell positioned in front of the instrument. Use will be made of gases
offering rich spectra, like e.g. water or methanol. Room temperature will indeed
create populations of many energy levels in these molecules, causing many spectral
lines to be present. In order to avoid pressure broadened lines and too many blended
features, the pressure has to be lowered to at some mbars (TBC). A hot/cold switch
source should provide the background modulation as there is no possibility to mimic
a wobbling mirror in the lab. As the whole system will be pumped out, very little water
absorption will be present on the line-of-sight, decreasing significantly the noise
temperature and thus the required integration times.

   1. IE: Set up gas cell
   2. IE: Set up frequency survey - Choose LO-bands and integration time. LO-
      bands may be assessed on the experience of the de-convolution working
      group, as well as following the results reported in Comito & Schilke 2002.
      Integration times of the order of some sec. per phase (i.e. on both Hot and
      Cold loads) should already allow signal-to-noise ratio of the order of 50.
   3. IE: Tune to the first LO frequency, let the system stabilize.
   4. IE/CS: Do the frequency survey within one LO sub-band, gather scans every 4
      (TBC) seconds (Only WBS data). The LO stepping scheme, may be optimised
      here by trying different schemes (cf Comito & Schilke 2002). Do hot/cold
      calibration and remove spurious signals:
                     Observe Hot and Cold phases with the cell filled with the gas.
                     Repeat this for each LO step



                                         135
                      HIFI calibration use cases                  Doc.: ICC/2001-005
                                                                  Date: 20/10/03
                                                                  Issue: Draft 1.0


                    Calibrate the data via the observation of the emptied cell (this
                    should take no more than a minute). In principle this should be
                    done again for each LO step (the baseline pattern changes with
                    LO setting) although stability may be of concern.
                Note: The calibration spectra could also be provided by an internal
                hot/cold measurement, although it may not be able to correctly
                calibrate for the optical losses across the gas cell path.
                Fundamentally, the only difference with the standard chopper wheel
                equation (e.g. RD16) stands in an additional factor ηCSA/ηcell where
                these factors represent the respective coupling through the CSA
                path (internal load) and the cell path. This should be investigated.

  5. IE/CS: Use Maximum Entropy Method from P. Schilke (MPIfR) for glueing the
     spectra and de-convolve the respective image and signal bands. An adequate
     tool should be provided by the de-convolution working group lead by IPAC.
  6. CS: Examine the spectra visually.
                   If available, overlay FTS spectrum and HIFI SSB (single side-
                   band) spectrum. Determine peak ratios for the same lines from
                   the two different (FTS vs. HIFI) spectra. Check the line-
                   frequencies against catalogue (e.g. http://spec.jpl.nasa.gov or
                   http://www.cdms.de) and against FTS spectrum.
                   Use fits to the lines to determine the precise frequency. Explain
                   differences. Check if the peak ratio compares well with the
                   difference in resolution. The integral under the line should be the
                   same for the two instruments.
                   Check for ghosts and other spurious signals in the HIFI
                   spectrum. Trace spurious signals in the original scans. If not
                   present, check if MEM creates spurious contributions. Vary the
                   MEM input parameters (side-band ratios, noise levels, etc.).
                   Rerun the MEM code. Re-inspect etc.
                   File report, with inclusion of all spuriuous signals, all determined
                   peak ratios, all MEM settings etc.
  7. CS: If the spectra of the the FTS and HIFI are well in agreement with each
     other, consider this mode validated.

  Question: Who would provide the FTS spectra ?

- In-orbit:
    1. CS/AS: Create observation for spectral survey. Choose a rich source (hot
       cores from the calibration plan are the most likely candidates). The observing
       mode is here double beam switching, so the source should be compact
       enough. General integration times are expected to be about 8 seconds. Only
       WBS required.
    2. CS/AS: Perform the observation: the LO-stepping strategy should be derived
       from our experience during ILT.

  After operation is conducted and stored in the HCSS:


                                         136
                       HIFI calibration use cases                   Doc.: ICC/2001-005
                                                                    Date: 20/10/03
                                                                    Issue: Draft 1.0


   3. CS/AS: Get data from archive
   4. CS/AS: Create spectra in the usual way (check complete calibration scheme
      of the considered mode in RD37): Remove zero, fit comb spectrum, do
      hot/cold calibration. Inspect visually for spikes and remove these.
   5. CS/AS: Use MEM software to derive the single SSB spectrum
   6. CS/AS: Inspect visually for spurious signals.
   7. CS/AS: Overlay existing ground-based survey spectrum multiplied by the ratio
      of the Herschel-beam and the ground-based beam solid angles ( ΣHSO/Σground).
      Check the peak fluxes of strong lines. See if the ratio is either 1 or can be
      explained by difference in beam-filling. The maps in the strong lines can serve
      as test for this. Pointing may also be an issue.
   Note: this can only be done in the lowest frequency band as no ground-based
   survey will likely be available above 900 GHz.
   8. CS/AS: Change parameters of MEM analysis. Check if spectra are improved.
      Obtain (subjective) best parameters.
   1. CS/AS: Check all the recorded ground-based frequencies against the HIFI
      spectra. Consider the maps when only a few frequencies are off. These may
      be attributed to differences in velocities of the object under study. Use
      position-velocity (PV) diagrams to study these effects.
   2. CS/AS: Consider the mode validated when no spurious signals are present
      and if the peak fluxes are explainable. The latter is subjective. At present no
      objective criteria have been formulated.

Both ILT and in-orbit:
File extensive reports with details on source, ground-based data and instrument and
analysis parameters

Open Issues:
It is unclear if we can really validate this mode in an objective fashion. Can we put
objective criteria, while checking spectra from different instruments and/or different
telescopes?

It is impossible to have a ground-based spectral survey in Band 6 and likely in
Band 5. We have to judge the mode from ILT and by induction from the other
bands. For some frequencies SOFIA spot checks are possible, but the beam
difference may be a concern.

References:
Comito C., Schilke P., 2002, A&A 395, 357




                                          137

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:3
posted:9/5/2011
language:English
pages:137