Docstoc

Report_WP9_meeting_June2011.docx - IS-ENES - The ENES Portal

Document Sample
Report_WP9_meeting_June2011.docx - IS-ENES - The ENES Portal Powered By Docstoc
					           Report on


IS-ENES Work-Package 9 Meeting
      Paris, 27-28 June 2011




         28th October 2011

            (Month 32)



           CNRS-IPSL
                             Report on the IS-ENES Work-Package 9 Meeting (Paris, 27-28 June 2011)




Table of content



Summary ....................................................................................................................................................... 3
Introduction ................................................................................................................................................... 4
1      Creation of support structure to facilitate model evaluation tests (Task 2) ........................................... 4
2.     Cataloguing and access services to observation datasets and evaluation tools (Task 3) ....................... 5
3.     Pilot study on the implementation of evaluation toolkits on IPSL-CM5 output (Task4) ...................... 9
Next WP9 meeting ........................................................................................................................................ 9
Milestones and Deliverables’ status .............................................................................................................. 9
Actions plan................................................................................................................................................. 10




                                                                                                                                                                2
                   Report on the IS-ENES Work-Package 9 Meeting (Paris, 27-28 June 2011)




Summary

The IS-ENES WP9 meeting (27th-28th June 2011, Paris, France) aimed at evaluating the progress and
results in the Work-Package 9 of IS-ENES, which objective is to create an interdisciplinary infrastructure
to standardize, harmonize and simplify - where possible - the tools and methodologies used in existing or
past model inter-comparison projects, in order to facilitate the evaluation of Earth System Models. The
meeting program was divided into three main sessions: i) Creation of support structure to facilitate model
evaluation tests, ii) Cataloguing and access services to observation datasets and evaluation methods, and
iii) Pilot study on the implementation of evaluation toolkits on IPSL-CM5 model output. (i) Listings of the
model output parameters, as well as formatting tools were provided, and (ii) a beta version of the
evaluation      database is      now available        on the        IS-ENES      v.E.R.C virtual       portal
(https://verc.enes.org/models/evaluation-portal). A first set of evaluation tools dealing with the evaluation
of the chemistry, precipitation, clouds and surface fluxes components has been identified. (iii) It will be
tested on the IPSL-CM5 outputs in the forthcoming months, and further released on the evaluation portal
by the end of 2012.




                                                                                                           3
                    Report on the IS-ENES Work-Package 9 Meeting (Paris, 27-28 June 2011)


Introduction

The Workpackage 9 of IS-ENES project aims to create an interdisciplinary infrastructure to standardize,
harmonize and simplify - where possible - the tools and methodologies used in existing or past model
intercomparison projects to evaluate the basic quality of an Earth System Model (ESM). This
infrastructure, accessible through the public IS-ENES virtual Earth System Modelling Resource Centre
(v.E.R.C.; https://verc.enes.org/), will include software to prepare required model diagnostics, easy-to-use
evaluation toolkits, as well as links to pre-assembled most recent state-of-the-art observational datasets.

The IS-ENES WP9 meeting that took place at Jussieu in Paris (France) on 27th-28th June 2011 aimed at
evaluating the progress and results in the workpackage so far, and at defining an actions plan by the end of
the project. The agenda (pdf), the participant list (pdf) and the presentations are available on the exchange
platform of IS-ENES portal. The meeting program was divided into three sessions corresponding to the
remaining tasks, as follows:

Session 1: Creation of support structure to facilitate model evaluation tests (Task 2)
Session 2: Cataloguing and access services to observation datasets and evaluation tools (Task 3)
Session 3: Pilot study on the implementation of evaluation toolkits on IPSL-CM5 output (Task 4)



1   Creation of support structure to facilitate model evaluation tests (Task 2)

Involved partners: DLR (Leader, 6 pm), CNRS-IPSL (5 pm), AA (6 pm)
Work completed
Listings of the model output parameters
     ESMval paper (presentation not provided for confidentiality reasons)
Progress in the ESMval paper (Eyring et al., 2011, in preparation) has been presented (Veronica Eyring,
not provided because of confidentiality reasons). This paper that is partly funded by the WP9 of the IS-
ENES project will serve as a basis for the selection of diagnostics. A set of core evaluation processes
structured around three major topics (physical climate, global carbon cycle, atmospheric composition) has
been developed. Each process is associated with one or more model diagnostics and with relevant
observational datasets that can be used for the evaluation. The proposed diagnostics go beyond
climatological means and include measures of spatial patterns, trends, and characteristics of variability on
different timescales. The paper includes a “process and evaluation table” that consists of a list of key
climate feedbacks, core processes as well as diagnostics, variables, observations and references to evaluate
ESMs. The paper is planned to be submitted to Journal of Climate.

Support structures and formatting tools

In addition to listings of the model output parameters, Task 2 also aims at developing support structures
(guideline, expert network, software examples) to help ESM modeling teams to implement the appropriate
model diagnostics, and formatting tools (including CMOR, instrument simulators, variable meaning and
unit standards), so that standard evaluation tests can be easily applied. The formatting tools shall constitute
interfaces to bridge the format gaps between model and observational output. Two actions were presented
that concern the chemistry and aerosol components:
       CMOR tables for ACCMIP/IPCC by Michael. Schulz (pdf)

                                                                                                             4
                    Report on the IS-ENES Work-Package 9 Meeting (Paris, 27-28 June 2011)


M. Schulz reported the current effort of the WP9 partners involved in the ACCMIP project (CNRS-IPSL,
DLR) to merge diagnostics on aerosol, tropospheric and stratospheric chemistry forcing into one set of
CMOR tables. These tables, largely based on fields archived for HTAP, AeroCom, and/or CCMVal, also
include several newly created diagnostics. They follow standardized formats and use CF-compliant names
whenever available (http://www.giss.nasa.gov/projects/accmip/).
       AeroCom-CALIOP study by Brigitte Koffi (pdf)
Another CNRS-IPSL contribution to Task2 consisted in the application of the CALIOP Layer Product to
the evaluation of the vertical distribution of aerosol in global models. In this study, supported by the IS-
ENES project, the CALIOP Layer product 3.01 is used to evaluate the vertical distribution of aerosols in
AeroCom global models. A processing of the CALIOP Layer data including data screening, cloud
masking and vertical interpolation allows to compute annual and seasonal mean extinction profiles over
sub-continental regions that are easily comparable to the model extinction profiles. A mean aerosol
extinction height diagnostic (Zα) has been also defined to quantitatively assess the models’ performance
that could be used as a benchmark in future ESM evaluation (i.e., for assessing the (impact of)
biases(uncertainties) in the aerosol vertical distribution).

Planned work
       Finalization of Eyring et al. (2011) and Koffi et al. (2011) papers by DLR and CNRS-IPSL
        partners, respectively.
       Creation of a global 3D CALIOP gridded database of the aerosol extinction at 532 nm (CNRS-
        IPSL) to facilitate the model evaluation (aerosol component).

Publications
Eyring et al. (2011) Uncertainties in Climate Projections: The Case for a Process-Oriented Evaluation of
Earth System Models, J.Clim., in preparation.

Koffi et al. (2011) Application of the CALIOP Layer Product to evaluate the vertical distribution of
aerosols estimated by global models: Part 1. AeroCom phase I results”, submitted to JGR.


2. Cataloguing and access services to observation datasets and evaluation tools (Task 3)

Involved partners: AA (Leader, 10 pm), CNRS-IPSL (2 pm), SMHI (6 pm), DLR (2 pm), MPG (6 pm),
FMI (4 pm)

Work completed
Evaluation portal
The WP partners prepared and distributed questionnaires to all the major European ESM groups with the
objective to document the model properties that the groups evaluated against observational data, the
observational datasets that were used for the evaluation of the different parameters, the methods that were
used, and the problems that were encountered. Based on the received questionnaires, the AA partner built
an evaluation database that aims to facilitate the evaluation process of the ESM groups (D9.1 report).
Progress (pdf) and difficulties (pdf) in the development of the database were presented by George
Tselioudis and Kostas Douvis , respectively. A beta version of the database is now available from IS-
ENES v.E.R.C virtual portal: https://verc.enes.org/models/evaluation-portal.



                                                                                                          5
                   Report on the IS-ENES Work-Package 9 Meeting (Paris, 27-28 June 2011)


First selected evaluation tools

Simplified versions of commonly used evaluation tools, as well as newly developed tools that will be
provided on the evaluation portal, as on line version or URL link, were presented:
  Cloud clustering, precipitation composite and satellite simulator tools by George Tseloudis (pdf)
  Cloud properties are strongly dependent on the variability of dynamic and thermodynamic atmospheric
conditions, thus it is crucial for models to be able to capture such dependencies. The process-based
evaluation of models requires the application of methodologies that stratify model and observational
outputs into regimes that have physical meaning. The main idea is to break up the complex cloud,
radiation, and precipitation fields into clusters where a certain combination of atmospheric processes
dominates the cloud and rain formation process. In that way model deficiencies in a particular
meteorological regime that are detected can be attributed to the specific process or processes that are
dominant in the deficient cloud system. This evaluates the results of GCM parameterizations and provides
more insight into sources of error. Regime separation methods that have been used recently in model
cloud evaluation include compositing and clustering. In compositing techniques, one or more atmospheric
properties are used to define atmospheric states on which cloud, radiation, and precipitation properties are
composited into different dynamic and/or thermodynamic regimes. Clustering techniques use properties of
the cloud field to define distinct groupings of cloud types that form distinct cloud systems corresponding
to particular regimes.
   CCMval diagnostic tool by Veronica Eyring (pdf)
The tool developed and handled by the DLR Institute, in the frame the CCMval activity calculates core
diagnostics for the evaluation of coupled chemistry-climate models, with a focus on the model's ability to
predict future stratospheric ozone (http://www.pa.op.dlr.de/CCMval/CCMVal_EvaluationTable.html).
Models versus observations comparison include tracer-tracer correlations, time series of the mean annual
cycle and standard deviation at selected pressure or temperature levels, within different latitude ranges,
time series of modeled zonal mean in the tropical upper stratosphere, mean vertical profiles and column
abundances. To establish the fidelity and quantify the assessment of the simulations metrics representing
many of the key stratospheric dynamical processes are calculated, which are related both to mean climate
(e.g., zonal mean climatology, extra-tropical wave driving), and to climate variability (extra-tropical and
tropical variability, stratospheric sudden warming, final warming) processes. An updated and simplified
version of the CCMval diagnostic tool is currently being developed (Gettelman, et al., 2011), which will
produce quick looks at standard diagnostic plots, output diagnostic variables, as well as climatology files
from CCMval CF‐compliant model output. In addition to stratospheric chemistry, it will also include the
evaluation of some tropospheric (gas phase and aerosols) diagnostics.
   HOAPS satellite data and simulator by Axel Andersson and Stephan Bakan (pdf)
The Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data (HOAPS) is the only
generally available satellite-based dataset with consistently derived global fields of both evaporation and
precipitation and hence of freshwater flux (evaporation minus precipitation) as well as 12 other parameters
which provide all basic state variables that are necessary to determine the evaporation. The new version of
the Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data set, HOAPS-3, contains
updated global fields of precipitation and evaporation over the global ocean and all basic state variables
needed for the derivation of the fluxes. Except for the NODC/RSMAS Pathfinder SST data set, all
variables are derived from SSM/I passive microwave satellite data over the ice free global ocean. HOAPS-
3 covers the time span from 1987 to 2005 (http://www.hoaps.zmaw.de/index.php?id=home). As a next
step, climate modelers will be given the opportunity to evaluate their model results by applying the
HOAPS climatology. For this purpose a HOAPS simulator is under finalization that provides an interface
between the spatio-temporally unevenly sampled satellite observations with the model data. The tool is to
model output with a high temporal resolution (at least 3-hourly) and mimics the satellite sampling by a

                                                                                                          6
                    Report on the IS-ENES Work-Package 9 Meeting (Paris, 27-28 June 2011)


retrieval based filter as well as an orbit mask for the satellite positions. Hence, it is possible to quantify the
systematic biases between model simulations and satellite based data due retrieval properties and/or the
temporal coverage of the satellites. A first study has been conducted by applying the simulator to
evaporation fields from the ECHAM-5 model. Furthermore the simulator is currently used for the
evaluation of ocean surface fluxes from the MPI-ESM CMIP5 experiments.
  AeroCom tool by Michael Schulz (pdf)
The AeroCom project intends to document differences of aerosol component modules of global models
and to assemble data-sets for aerosol model evaluations. It focuses on the five most important aerosol
components, i.e. dust (DU), sea salt (SS), sulfate (SO4), black carbon (BC), and particulate organic matter
(POM), and on the sum of these components. The AeroCom tool allows evaluating the model outputs by
comparison to different observational datasets. Observed surface concentrations, aerosol optical depths
(AOD) and extinction vertical profiles are computed from measurements from global and regional
networks (GAW, AERONET, IMPROVE, EMEP, Aeroce and Earlinet) and satellite observations
(POLDER/Parasol, MODIS, MISR, TOMS and AVHRR). A CALIOP 3D gridded product of aerosol
extinction profiles is also currently being built, that will be included into the AeroCom tool, together with
the calculation of the related Zα diagnostic (see section 1). The AeroCom tool is not directly available as
an online software package but the modelers are rather requested to provide their model outputs following
a standard data-protocol. Evaluation results are then provided to the modelers through the graphical
evaluation environment of the AeroCom website (http://aerocom.met.no/aerocomhome.html).

Planned work
Evaluation portal: improvements/updates
    Date of revision of database should be mentioned on front page
    Name of model should agree with IS-ENES website
    Entry page should have a contact name and institutions contributing
    Model description should refer to IS-ENES model description resources
    Usage of datasets by MIPs could be marked, mentioned
    Reordering of dataset entries according to weighting of importance
    Should we make the questionnaire public so that other ESMs could contribute?
    Name of responsible for each model component
    Responsible for each component should go over datasets one by one (e.g. ARGOS is missing)
    Single entry - i.e. the most recommended URL link - for each dataset
    To be updated with Ocean and Chemistry components for the IPSL model (questionnaires completed)
    Refer to tools in three categories on the evaluation portal: data preparation/generic comparison
     /specific MIP type evaluation tool (Figure 1). Tools to be linked to relevant diagnostics and datasets
    Refer to reference papers on datasets and tools (e.g, Eyring et al. 2011; Gettelman, et al., 2011)
Evaluation tools
    Release to CNRS-IPSL partner of a beta version of the selected evaluation toolkits (cloud clustering,
     precipitation composite, CCMval, HOAPS), together with tutorials
    Proposition of an harmonized documentation of the evaluation tools (CNRS-IPSL partner)
    Selection of a suite of core diagnostics (all partners) to be evaluated from the selected evaluation
     tools

Publications
Gettelman, et al., 2011, in preparation for ACPD or GMDD

Tselioudis et al., 2011, in preparation for BAMS

                                                                                                                7
                  Report on the IS-ENES Work-Package 9 Meeting (Paris, 27-28 June 2011)




                                            Surface Fluxes

                                    Please set your search criteria...



                                                                         SELECT BY ESM

                         Properties Evaluated:
                                                                         COMPLETE LIST




                                                                         SELECT BY ESM



                      Datasets used for evaluation:                  SELECT BY DATASET



                                                                         COMPLETE TABLE




                                                             Data/model output preparation

                      Tools used for evaluation:                         Generic tools

                                                                         Specific tools


                                         Back to main page......



Figure 1: Illustration of the Evaluation portal (example of the Surface Fluxes component entry page): to
be updated with links to the evaluation tools and tutorials, for each relevant component, diagnostic, and
dataset.




                                                                                                       8
                   Report on the IS-ENES Work-Package 9 Meeting (Paris, 27-28 June 2011)


3. Pilot study on the implementation of evaluation toolkits on IPSL-CM5 output (Task4)

Involved partners: CNRS-IPSL (Leader, 10 pm), AA (8 pm), DLR (2 pm), MPG (2 pm)

Work completed
Initially planned to begin on month 30 (August 2011). Beginning is postponed on month 33.

Planned Work
The pilot study will begin as soon as (links to) the toolkits will be provided to the IS-ENES web portal and
a list of crucial diagnostics will be set. CMIP5 outputs from the IPSL-CM5 model will be used to test the
collection of toolkits and relevant observational datasets. The objective is to test the functionality of the
assembled software tools and to iron out any potential problems before releasing the evaluation portal to
the ESM community. At the time of writing, the identified tools (and diagnostics) include HOAPS (fresh
water flux), CCMval (e.g., tropospheric and stratospheric ozone), AeroCom (AOD, Zα), and the cloud
clustering and precipitation composite tools.


Next WP9 meeting

The next WP9 meeting will be hold on December 2011 as part of the next IS-ENES General Assembly
(12-14 December, Lecce, Italy).


Milestones and deliverables’ status

D9.1: Delivered (pdf)

M9.2 (Task 3, mo 30) Release to IS-ENES web portal of beta version of the evaluation toolkits
Postponed to month 33 for the earliest available tools.

D9.2 (Task 2, mo 30) Listings of required model fields and interface software
In progress as part of the evaluation portal. Postponed to month 34.

D9.3 (Task 3, mo 30) Listings and access services to observational datasets and evaluation tools
In progress as part of the evaluation portal. It has notably to be updated with links to selected evaluation
tools and their documentation (tutorial, standard description template). Postponed to month 34.

D9.4 (Task 4, mo 48) Report on pilot study for the evaluation of CNRS-IPSL ESM output using the
collected toolkits and datasets.




                                                                                                           9
                  Report on the IS-ENES Work-Package 9 Meeting (Paris, 27-28 June 2011)


Actions plan

   What                       Who                 When (Mo)           Comments
   Update/improvements        AA                  On a      regular Including new contributions from
   of the evaluation portal                       basis             all WP9 partners
   Provision    of   the MPG                      33                  To CNRS-IPSL partner
   HOAPS simplified tool
   and tutorial
   Provision of the cloud AA                      33                  To CNRS-IPSL partner
   clustering          and
   precipitation composite
   tools and tutorials
   Pilot study on        the CNRS-IPSL            > 33                Postponed from Month 30
   implementation         of
   evaluation toolkits
   Provision    of    the DLR                     34                  To CNRS-IPSL partner
   CCMval simplified tool
   and tutorial
   Provision    of   an CNRS-IPSL                 34                  Draft to all partners on month 33.
   harmonized
   documentation of the
   tools
   Selection of a suite of All partners           34
   diagnostics to be tested
   on IPSL-CM5
   Preparation of the 2nd AA                      34                  With contributions of all partners
   GA of IS-ENES
   D9.2                       DLR                 34                  Postponed from Month 30
   D9.3                       AA                  34                  Postponed from Month 30
   Reporting                  AA                  36                  With contributions of all partners
   Provision of a CALIOP CNRS-IPSL                40
   3D     global    aerosol
   extinction dataset
   D9.4                       CNRS-IPSL           48
   Publication of ESMval DLR                      48
   paper
   Publication    of CNRS-IPSL                    48
   CALIOP-AeroCom
   paper




                                                                                                           10

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:6
posted:2/27/2012
language:
pages:10