Ionospheric Sounder

Document Sample
Ionospheric Sounder Powered By Docstoc
					D42-1       O BLIQUE IONOSPHERIC SOUNDING                F INAL R EPORT




 THE OBLIQUE IONOSPHERIC SOUNDER




                    Project Final Report

                        October 2000



                     Dr. Ruth Bamford

              Radio Communication Research Unit,
                 Rutherford Appleton Laboratory,
           Chilton, Didcot, OXON OX11 0QX, UK

        Tel. 01235 44 6517, E-mail: R.Bamford@rl.ac.uk



                              0
D42-1                                       O BLIQUE IONOSPHERIC SOUNDING                                                       F INAL R EPORT




                                  TABLE OF CONTENTS

THE OBLIQUE IONOSPHERIC SOUNDER............................................................0
Table of Contents ..........................................................................................................1
List of Figures ...............................................................................................................3
Report Introduction ......................................................................................................6
   Overview............................................................................................................................... 6
   Project Rationale ................................................................................................................. 6
   Objectives ............................................................................................................................. 6
   The contents of This report ................................................................................................ 7
Section 1: .......................................................................................................................8
Radio propagation in the ionosphere ...........................................................................8
      The ionosphere.................................................................................................................................8
      General principles of Propagation in the ionosphere .......................................................................9
      Determining how radio waves propagate in the ionosphere by measurement ...............................11
      Vertical ionospheric sounding .......................................................................................................11
      True height and determining the electron density profile ..............................................................13
      Vertical ionogram profile inversion...............................................................................................15
      Oblique ionospheric sounding .......................................................................................................16
      Comparing oblique and vertical ionograms ...................................................................................18
Section 2: .....................................................................................................................20
The instrumentation....................................................................................................20
      The oblique ionospheric sounder ...................................................................................................20
Section 3: .....................................................................................................................25
Results from the 1997-1998 uk oblique sounding campaignes.................................25
   Introduction ....................................................................................................................... 25
   Results from the Chilbolton – Chilton campaign ........................................................... 26
      Chirp and digital instrumental comparisons ..................................................................................26
   Results from the great Baddow – Hartand point campaign .......................................... 27
      Results from short range mid-latitude path....................................................................................27
   Results from the Svalbard – Chilton measurements ...................................................... 36
      Results from 3000km range trans-auroral path..............................................................................36
   Conclusions to section 3 .................................................................................................... 40
Section 4: .....................................................................................................................42
Results from high latitude oblique sounder measurements ......................................42
   Rationale............................................................................................................................. 42
      Graphical interface added to the ray trace ..................................................................................... 43
      The Imaging Riometer ...................................................................................................................45
      The Oblique Ionospheric Sounder .................................................................................................45
   Results from the Karesuvanto – Tromsø campaign ....................................................... 46



                                                                        1
D42-1                                    O BLIQUE IONOSPHERIC SOUNDING                                               F INAL R EPORT


   Results from the combined absorption and ray tracing model ..................................... 48
   Conclusions to section 4 .................................................................................................... 53
Report Summary .........................................................................................................54
   Future work ....................................................................................................................... 55
List of References........................................................................................................57
Appendix A: List of data .............................................................................................59




                                                                   2
D42-1                                           O BLIQUE IONOSPHERIC SOUNDING                                                          F INAL R EPORT



                                              LIST OF FIGURES

Figure 1. The location of the layers of the ionosphere in altitude and a sketch showing the peaks in the
electron concentration that are referred to as the D, E and F layers of the ionosphere. .........................8
Figure 2. A sketch of the geometry of vertical ionospheric sounding. N(h) represents the profile of the
electron density variation with altitude. .................................................................................................12
Figure 3. A typical vertical ionogram. The virtual range or altitude is plotted against the receiver
frequency. (Picture thanks to the Ionosonde monitoring group at RAL)................................................12
Figure 4. A sketch representing the form of a typical day time ionospheric electron density profile. ...15
Figure 5. A vertical ionogram with a POLAN true height profile. (Picture thanks to the Ionosonde
monitoring group at RAL). .....................................................................................................................16
Figure 6. A sketch showing the geometry for oblique propagation with multi-hop propagation routes
between transmitter and receiver. ..........................................................................................................16
Figure 7. A typical oblique ionogram for propagation from Svalbard in Norway to Chilton in the UK.
The effect of propagation paths that involved multiple reflections from the ionosphere is shown.........17
Figure 8. A sketch showing the equivalent paths for a vertical and obliquely propagating waves
reflected at the same real height (B and BB) and at the same virtual heights (A and AA).....................18
Figure 9. A sketch illustrating how vertical ionograms project to oblique ionograms. The virtual
height, h’ is equal to the time-of-flight of the radio signal times c. ........................................................19
Figure 10. A sketch showing the nature of a chirp ionosonde signal. The transmitted signal is a phase
continuous with a ramping frequency at a specified rate. The return echoes from the ionosphere track
the transmitted frequencies but with a time delay (Dt) corresponding to the flight time to and from the
reflecting ionospheric layer. Multiple reflections from the ground back up to the ionosphere and down
again are also seen with time delays that are multiples of the single reflection time Dt. .......................20
Figure 11. A photograph of the IRIS oblique sounder receiver unit. .....................................................22
Figure 12. A photograph of the oblique sounder transmitter unit..........................................................22
Figure 13. A sketch of the antenna arrangement in this case for the receiver. A similar antenna was
used for the transmitter sites. .................................................................................................................23
Figure 14. A map of Europe showing the RAL oblique sounder receiver sites and the transmitter sites
throughout the course of this project (see key in Table 3). Signals from non-RAL transmitters were also
received during the course of the project. The locations of those transmitters are not shown...............24
Figure 15. A map showing the location of the oblique and vertical sounder transmitter and receiver
sites in the UK during 1997 to 1998. The monitoring vertical incidence (VI) ionosonde was located at
Chilton....................................................................................................................................................26
Figure 16. An overlay of the almost simultaneous ionograms recorded by a digital (blue and yellow
trace) and chirp (black trace) ionosondes. There is a 50 second difference in the sweep start times of
the two ionograms. .................................................................................................................................27
Figure 17. A sketch of the relative position of the oblique transmitter and receiver with the vertical
ionosonde. The electron density profile of the ionosphere is represented as N(h). ................................28
Figure 18 A vertical and oblique ionogram taken within 1 minute each other, both observing
approximately the same region of the ionosphere. .................................................................................28
Figure 19. A comparison between the measured oblique sounding data and geometrically scaled
vertical ionogram data ( fo = fv sec(f))...................................................................................................30
Figure 20. The observed diurnal variation of bandwidth for a short hop (363km) between Great-
Baddow and Hartland Point in the UK. .................................................................................................31




                                                                             3
D42-1                                            O BLIQUE IONOSPHERIC SOUNDING                                                             F INAL R EPORT


Figure 21. The extrapolated bandwidth for a short hop oblique path (360km) based on the Chilton
Digisonde DPS-1 ionogram data and a simple fo = fv sec(f) vertical-to-oblique conversion................31
Figure 22. A POLAN electron density profile (expressed as plasma frequency see Equation 1) for a
vertical ionogram recorded at Chilton on 30/4/1997 at 17:00 UT. POLAN produces discrete points which
are shown as red circles. The blue crosses mark the QPS boundary points used in the analytical
reconstruction of the profile. The solid black curve shows the QPS fit. .....................................................32
Figure 23. Ray paths calculated using the multiple quasi-parabolic (MQP) profile fitting analytical ray
trace........................................................................................................................................................33
Figure 24. A comparison between measured and MQP ray-trace reconstructed oblique ionograms. The
blue trace is the measured oblique ionogram between Great Baddow and Hartland Point (363km) on
16/08/1997 at 14:59UT. The red trace is a MQP ray analytical ray trace reconstruction of the
ionogram based upon a POLAN profile obtained from a mid-point vertical incidence ionogram
recorded at Chilton on 16/08/1997 at 15:00UT. ....................................................................................34
Figure 25. A comparison between a measured vertical incidence ionogram recorded at Chilton and the
MQP ray trace reconstruction of the vertical ionogram based upon a POLAN profile created from the
original measurement.............................................................................................................................35
Figure 26. A comparison between measured and Breit & Truve extrapolated vertical to oblique
ionogram. The blue trace is the measured oblique ionogram between Great Baddow and Hartland
Point (363km) on 16/08/1997 at 14:59UT. The red trace is the measured Chilton vertical ionogram
data recorded at the path mid point at the same time but converted into an oblique ionogram of the
same distance using the simple sec(ø) mapping. ....................................................................................36
Figure 27. A typical oblique ionogram for propagation from Svalbard in Norway to Chilton in the UK.
................................................................................................................................................................37
Figure 28. A map of Northern Europe showing the location of the IRIS oblique sounder receiver at
Chilton in the UK and the location of the transmitter on Svalbard. The location of the Kilpisjärvi
riometer (69.03N, 20.5E), which measures the ionospheric absorption, is also indicated. ...................37
Figure 29. An spectrogram of the IRIS data recorded at Chilton from the chirp transmitter located on
Svalbard . ...............................................................................................................................................38
Figure 30. The Polar Cap radio Absorption (PCA) event of November 1997. Top: The 30 MeV protons
and 0.1-0.8 nm X-ray flux as recorded by the GOES-9 satellite on the 5th-11th November 1997. Middle:
The oblique HF propagation between Svalbard and Chilton during this event. Bottom: The Lerwick
(1.2W, 60.1N) vertical ionosonde ionogram records. ............................................................................39
Figure 31. An example of the calculations performed by the ray tracing program. ..............................43
Figure 32. Multiple outputs from the modified ray tracer program. The graphical interface to the ray
tracer is written in IDL, the ray tracer itself is in C++..........................................................................45
Figure 33. The projection onto the ionosphere at 90 km altitude of the Kilpisjärvi IRIS imaging
riometer beams. The locations of the oblique sounder transmitter and receiver below the riometer's
observational region are also indicated. ................................................................................................45
Figure 34 A vertical projection of the imaging riometer beams at 90km (as shown in Figure 33.) and
the path of an oblique incidence HF radio wave between Karesuvanto and Tromsø. ...........................46
Figure 35 A typical ionogram from the Karesuvanto-Tromsø short-hop oblique HF sounder. An
absolute group delay of 1ms over the 200km ground range corresponds to a virtual height of
approximately 110km vertically. ............................................................................................................47
Figure 36 (a) Relative median signal strength variation of a 2.5MHz HF signal propagating from
Karesuvanto to Tromsø reflected from heights below 150km (E-layer). (b) The measured amount of
ionospheric absorption recorded by the imaging riometer. The baseline is derived from the quite day
reference data (see (Browne et al 1995)). ..............................................................................................48
Figure 37. A POLAN true height electron density profile obtained from a Dynasonde (Wright 1998)
vertical ionosonde measurement made at Tromsø. ................................................................................49




                                                                               4
D42-1                                           O BLIQUE IONOSPHERIC SOUNDING                                                           F INAL R EPORT


Figure 38. (a) The electron temperature profile, (b) mean electron collisional rate calculated using the
Lancaster-Sodankylä ion chemistry model overlaid on the POLAN electron density profile (dashed
curves). ...................................................................................................................................................50
Figure 39. A plot of the ionospheric absorption per km for 2.5MHz radio waves determined using the
Lancaster-Sodankylä ion chemistry model overlaid on the POLAN electron density profile (dashed
curve)......................................................................................................................................................50
Figure 40. A single 2.5 MHz ray path calculated using the RaTs numerical ray trace and the POLAN
electron density profile based on Tromsø Dynasonde observation of the 5th December 1998 at
17:42UT. ................................................................................................................................................51
Figure 41. A plot of the accumulative absorption experienced by a 2.5 MHz based on the model results
of the Lancaster-Sodankylä ion chemistry model and the RaTs ray trace..............................................51
Figure 42. A projection of the path of a 6.5MHz HF ray on the IRIS imaging riometer beams. ...........52
Figure 43. The observed ionospheric absorption by the riometer at 38 MHz........................................52




                                                                             5
D42-1                          O BLIQUE IONOSPHERIC SOUNDING                             F INAL R EPORT



                    REPORT INTRODUCTION

                                        OVERVIEW
The ionosphere, which is that part of the Earth’s upper atmosphere that is ionised by radiation
from the sun, effects radio propagation from the extremely low frequencies (<3kHz) to super
high frequencies (30GHz). Below 30 MHz the ionosphere is an essential part of the propagation,
whereas above 30MHz the ionosphere is a source of band pollution particularly at night in the
LF and MF bands (30kHz to 3MHz) and system disruption for Earth-space communications
such as navigation systems. The nature of the ionosphere is intimately linked with the
fluctuations in the emissions from the sun as well as the state of the Earth’s atmosphere. This
makes it highly variable and level of disruption to communications difficult to predict at the
moment.

The changing state of the ionosphere is generally monitored by networks of vertical ionosondes.
These radars transmit and receive HF (1- 30MHz) radio waves to and from the ionosphere
directly above the monitoring station. The result is real-time information on the state of the
ionosphere needed by communications users allowing them to adapt their operational systems
accordingly. The oblique ionospheric sounder extends this idea. The transmitter and receiver of
the oblique sounder are not co-located like the vertical but are generally hundreds thousands of
kilometres apart. So the instrument is able to study how the radio signals of real communications
(below 30MHz) propagate via the ionosphere under a variety of conditions. However the
interpretation of the oblique propagation is also significantly harder than the case of the vertical..


                             PROJECT RATIONALE
Many radiowave applications require real-time determination of the 2-dimensional structure of
the ionospheric electron density along the path, or at least of the mid-path electron density
profile. Since it is often impractical to deploy a vertical ionosonde at the point of interest, such as
across oceans, the idea of reconstructing an average mid-path profile from oblique incidence
measurements is very appealing. Oblique incidence sounding also provides an excellent test of
ionospheric forecasting tools, since ray tracing through the predicted electron density structure
can be compared directly to the measured group delays on the oblique path. In addition the
measurements can be used to improve absorption modelling, which is of particular interest to
broadcasters wishing to limit spectrum pollution. Through knowledge of the state of the
ionosphere oblique sounding can be used to improve the accuracy of single frequency GPS
(Global Positioning System) navigational information.

A recently developed oblique sounding system, providing absolute ionospheric group delays
using GPS-timing, is used to carry out experimental campaigns at mid and high latitudes. The
project involves collaboration with the Radio Science and Propagation Group at DERA Malvern
and with a number of British universities. This work fits in well with the objectives of European
project COST 251, and is part of collaborative activities with several participating countries.

                                      OBJECTIVES
(i)     Organisation of experimental campaigns including suitable vertical incidence control
        measurements. At some stage this will involve other European countries, probably
        within the framework of COST 251.




                                                  6
D42-1                        O BLIQUE IONOSPHERIC SOUNDING                         F INAL R EPORT


(ii)    Development of inversion algorithms to obtain vertical electron density profiles from
        oblique soundings.

(iii)   Development of ionospheric ray-tracing algorithms.

(iv)    Testing of inversion and ray-tracing techniques using the oblique sounder measurements
        and the control vertical measurements.

               THE CONTENTS OF THIS REPORT
This report contains a description of the nature of the oblique ionospheric sounding and the
measurement campaigns conducted in the UK and abroad. Also included is description of the ray
tracing and other modelling used in the analysis of the propagation along with a presentation of
some of the results. Included as Appendices are a full listing of the dates and propagation paths
recorded during the period of the project and a CD is included of the full data set.




                                               7
D42-1                           O BLIQUE IONOSPHERIC SOUNDING                                F INAL R EPORT



                                        SECTION 1:
               RADIO PROPAGATION IN THE
                      IONOSPHERE
This section outlines the nature of the ionosphere and the general principles behind propagation
in an ionised media like the ionosphere before considering the radio propagation technique of
oblique sounding.

THE IONOSPHERE
The ionosphere extends from about 50km altitude to about 600km where it merges with near
Earth space environment. The ionosphere is created by radiation, both electromagnetic and
particles, coming from the sun which ionise the Earth's upper atmosphere into free electrons and
ions. The reason the free electrons persist in the ionosphere is that at these altitudes the density
of the Earth’s neutral atmosphere is sufficiently thin that collisions between particles happen far
less frequently than in the lower atmosphere. However even at these altitudes the particles of the
neutral atmosphere within the ionosphere still out number the free electrons by about 10000:1.
So the free electrons live much longer before they are recombined.




   Figure 1. The location of the layers of the ionosphere in altitude and a sketch showing the peaks in the
    electron concentration that are referred to as the D, E and F layers of the ionosphere.

A generalised profile of the electron density with altitude in the ionosphere would look similar to
that shown on the right hand panel in Figure 1. The peaks in the profile are labelled the D, E and
F layers for historical reasons.

The density of the electrons in the ionosphere varies with altitude in accordance with the balance
between varying wavelengths or energies of electromagnetic and particle radiation producing the



                                                     8
D42-1                          O BLIQUE IONOSPHERIC SOUNDING                           F INAL R EPORT


ionisation, originating from the Sun and the near Earth environment, and the requisite neutral
atmospheric chemistry and density. The motions of the neutral atmosphere further complicate
the environment both by turbulent motion and via tides and waves. Magnetic and electric
currents also influence the ionised particles.


GENERAL PRINCIPLES OF PROPAGATION IN THE IONOSPHERE

The presence of the free electrons in the ionosphere is what effects radio signals across several
bands of the spectrum from 3kHz to 30GHz. The alternating electric field of the radio wave
causes the motion in the free electrons. Depending upon the density of the electrons and their
range of movement and the frequency and amplitude of the radio wave, the effects range from
total absorption of the radio waves to selective reflection and phase delays. Leading to distortions
in the communications or complete failures.

The proximity of the transmitted radio wave to the local resonant frequency, or plasma
frequency, in the ionosphere determines many aspects of how the electromagnetic wave
propagates, i.e. what fraction of the energy is refracted, reflected, transmitted or absorbed.

The ionosphere is a plasma of freely moving electrons and ions. Plasmas are not rigid structures,
the electromagnetic force creates organisation upon the collection of charges particles. Once an
electron is displaced from the fairly uniform background of ions, by say the alternating field of a
passing electromagnetic wave, then the displacement of the electron creates an electric field
which attempts to restore the neutrality of the plasma by returning the electron to its original
position. However, because of inertia the electron overshoots and then proceeds to oscillate
about its equilibrium position with a characteristic frequency known as the plasma frequency, wp.

The relationship between the density of electrons and the plasma frequency can be shown to be:

                                                 Ne 2
                                  w p = 2pf p =                                           Equation 1
                                                e o me

Here wp is the angular plasma frequency and fp the plasma frequency, N is the density of electrons
and e the charge on the electron with me equal to the electron mass and eo is the permittivity of
free space. Since the electron density varies with altitude, N(h), then so does the plasma
frequency, wp(h). So the electron density profile in Figure 1 could be shown as either density of
electrons or plasma frequency.

For typical ionospheric electron densities of between 1x1011 to 1x1012 electrons per square meter
the plasma frequency, fp , is between 1 to 10MHz, which is in the HF band. This is why HF is the
band most significantly effected by the ionosphere.

The plasma frequency plays a fundamental role in the refractive index and hence propagation in
the ionosphere.

Without going into the derivation, which is covered in many text books (Davis 1990, Chen 1984,
Budden 1985), the phase refractive index of the ionosphere, mp, in its simplest case is given by:

                                               c      w2p
                                          mp =    = 1- 2                                  Equation 2
                                               vp     w




                                                  9
D42-1                          O BLIQUE IONOSPHERIC SOUNDING                           F INAL R EPORT


Where w is the angular frequency of the transmitted radio wave, c , the speed of light in a vacuum
and vp the phase speed of the radio wave.

This equation serves to illustrate some of the unique features of propagation in ionised media
compared to non-ionised media such as in the lower atmosphere.

When the transmitted radio frequency equals the local plasma frequency, w = wp , the refractive
index is then 0. This means that the radio wave is evanescent and does not propagate. The energy
of the wave is absorbed because it is in resonance with the local natural vibrational frequency of
the free electrons at the plasma frequency.

If the frequency of the radio wave is greater than the plasma frequency, w > wp , then the
refractive index is less than 1 and phase velocity, vp, of the radio wave would be greater than the
speed of light, c. This is allowed because the group velocity of a wave packet, vg , would be
correspondingly less than c, and it is the group velocity that carries the information and energy in
an electromagnetic wave (Budden 1985). This aspect is illustrated in practical terms in the next
section.

When the frequency of the radio wave is less than the plasma frequency, w < wp then the
refractive index is complex and radio wave experiences absorption which increases as w
approaches wp. But this is only one form of signal loss that the radio wave can experience.

The simple expression in Equation 2 for the phase refractive index assumes that there is no
effect of the Earth’s magnetic field and no effect from the presence of the neutral atmosphere
and the ions in which the free electrons are embedded. In reality both these factors significantly
effect the propagation of radio waves in the ionosphere and are some of the causes of signal
distortion.

With magnetic field but no collisional absorption the expression for the phase refractive index
for radio waves perpendicular to the magnetic field direction, becomes (Budden 1985):




                                                                                          Equation 3

Where wc is the electron cyclotron frequency. This is the frequency a free electron would naturally
gyrate about a magnetic field line.

The consequences of this are to illustrate how radio propagation in the ionosphere is
birefringent, which is to say that different polarisations of radio waves propagate differently. This
is because the Earth’s magnetic field in the ionosphere has a definite direction, which varies
according to latitude and radio signals with components parallel and perpendicular to the Earth’s
magnetic field propagate differently. This is a common source of signal distortion and fading.
Because the Earth’s magnetic field emerges and ends at the magnetic poles, at high latitudes the
field is almost vertical. Here the third polarisation of radio waves relative to the magnetic field
offers another propagation route to radio signals that shall not be gone into here.

With collisional absorption but no magnetic field the phase refractive index of Equation 2
becomes (Budden 1985):




                                                 10
D42-1                         O BLIQUE IONOSPHERIC SOUNDING                         F INAL R EPORT


                                                                                       Equation 4


Where i is the √-1 and γ is a mean collisional frequency between the free electrons and the
neutral particles of the upper atmosphere and ions.

The consequence of the inclusion of these collisions is again to make the refractive index
complex. It is possible to rearrange Equation 4 to separate the real and imaginary terms:

                                 m p = m p real - im p imaginary                       Equation 5

And the term µ p imaginary relates to the reduction in the amplitude or signal strength of the
electromagnetic wave through collisions. However because µ p imaginary relies on γ and some
knowledge of the density and composition of the neutral and ion populations throughout the
ionosphere it is very difficult to determine. The atmospheric density and composition at altitudes
of 50 to 400km is virtually never known. Only occasional rocket missions can be used to
determine this as no radio probing technique can be used. So modelling has to be used to make
the best guess as to the loss of signal strength due to this aspect of the propagation.

Most ionospheric ray traces either neglect this aspect of the propagation or use only crude
estimates for the signal loss due to the ionosphere. One of the contributions made during this
project has been a serious attempt to produce a combined ray trace and absorption model that
can be tested against real measurements.


DETERMINING HOW RADIO WAVES PROPAGATE IN THE
IONOSPHERE BY MEASUREMENT

How the radio signal travels overall from transmitter to receiver depends upon the accumulated
effects gathered by the radio wave as it travels through different parts of the electron density
profile of the ionosphere. The variation is not just limited to the vertical propagation as the
ionosphere is far from horizontally uniform. For instance effects like the passing of dawn and
dusk terminator each day produce strong tilts and horizontal gradients in the ionosphere.

The way the electron density profile of the ionosphere is determined experimentally is either by
incoherent scatter radar techniques (Davis 1990), which are large expensive facilities, or more
commonly by vertical ionospheric sounding.

It is important to mention in passing that the ionosphere intrusion on communications systems is
not just limited to propagation effects. Many communication and navigation satellites orbit at
altitudes that mean they are actually flying within the ionosphere or in the case of geostationary
satellites, the neighbouring region just beyond the plasmasphere. The presence of the fluctuating
levels of ionisation and electric and magnetic fields of this harsh environment frequently moves
satellites out of position, causes electronic failures and even destroys satellites.


VERTICAL IONOSPHERIC SOUNDING
Both vertical and oblique ionospheric sounders operate on the same principles. However the
geometry of the vertical case is more straightforward to interpret and have been used in this
project to compare with the more complicated oblique measurements.




                                                 11
D42-1                            O BLIQUE IONOSPHERIC SOUNDING                                F INAL R EPORT


Vertical ionosondes are radars that transmit HF (1- 30MHz) radio waves vertical up to the
ionosphere. Receive antennas co-located with the transmitter detect the return echoes from the
ionosphere (see the sketch in Figure 2). The time-of-flight of the radio signals at a particular
frequency gives an indication of the height of the reflecting layer, which is generally between 70
and 400km.




                               hv

                                                                           ht



   Figure 2. A sketch of the geometry of vertical ionospheric sounding. N(h) represents the profile of the
    electron density variation with altitude.

The raw measurement of the ionosphere from vertical or oblique ionosondes is shown
graphically in the form of an ionogram. Figure 3 shows a typical vertical ionogram. In the
ionogram the virtual height or range in km to the reflecting layer in the ionosphere is plotted
against the receiver frequency, which has been swept from 1 to 10MHz in this case as the
transmitter signal sweeps through the same frequencies.




   Figure 3. A typical vertical ionogram. The virtual range or altitude is plotted against the receiver
    frequency. (Picture thanks to the Ionosonde monitoring group at RAL).


The asymptotes in the plot correspond to the critical frequencies of the ionospheric layers. In the
ionosphere there are three main layers, or peaks in the electron density with altitude the D, E and
F layers. It is quite common for the upper most F layer to sub-divide into two layers depending




                                                     12
D42-1                          O BLIQUE IONOSPHERIC SOUNDING                            F INAL R EPORT


upon the latitude of the ionosphere and the time of year. The layers are then called the F1 and
F2.

The critical frequencies of each of the layers are identified by the asymptotes and the heights, or
more accurately the virtual heights of the ionospheric layers are scaled at the lowest points on
each curve between the critical frequencies. The asymptotes occur as a consequence of the group
delay of the radio waves progressively slowing down by the increasing depth of underlying
ionisation. At the asymptote or critical frequency the incident radio wave from the ionosonde has
reached maximum resonant frequency of the layer and the radio wave does not propagate. Once
the critical frequency of a layer is exceeded then the radio wave is no longer reflected by the layer
and is transmitted until it encounters an ionospheric layer of a higher electron density with
correspondingly higher critical frequencies. Eventually the transmitter frequency exceeds the
maximum ionospheric critical frequency then the radio wave continues into space.

The red trace shown in Figure 3 is not an additional ionospheric layer it is where the transmitted
radio signal has been split into two polarisations by the interaction with the Earth’s magnetic
field. The green trace, or ordinary ray, in the ionogram is the returned echoes from the
component of the transmitted radio signal that is parallel with the direction of the Earth’s
magnetic field. The red trace is the extraordinary ray, which is the component of the incident
radio wave perpendicular to the magnetic field direction. This illustrates the how the added
complication of the ionosphere being a birefringent medium (as shown in Equation 3).


TRUE HEIGHT AND DETERMINING THE ELECTRON DENSITY
PROFILE

In order to calculate the features of a received radio signal through the ionosphere, such as its
amplitude, polarisation, relative phase, time of flight, dispersive spread, it is necessary to know
the path the radio rays took from transmitter to receiver. To do this requires details of the
electron density profile to be extracted from the ionograms.

The vertical ionosonde actually measures time of flight (∆t ) between transmitting a signal and
receiving the return echo, which is usually less than 5ms. This is generally converted to a ‘virtual’
height in kilometres for convenience using an assumed constant speed of c, the speed of light in
vacuum, for the radio pulse.

                                                1                                        Equation 6
                                         hv =     cDt
                                                2

However the true altitude of the reflecting layer has to allow for the true speed at which the radio
wave has travelled:



                                                                                           Equation 7

Where hr is the real height of the reflection and vg is the group velocity of the radio signal.

Unlike radars in the troposphere, the real height in the ionosphere is always less than the observed
virtual height because the group velocity, vg , is less than c. The amount of the discrepancy
between the virtual height and the true altitude of the reflecting layer is dependent upon the
quantity of ionisation below the reflecting layer.




                                                  13
D42-1                         O BLIQUE IONOSPHERIC SOUNDING                          F INAL R EPORT


As mentioned in Equation 2, the reduced group velocity is due to the phase refractive index of
the ionosphere being less than 1. The group refractive index, mg , is given by:
                                        c     dk d ( wm p )
                                 mg =      =c    =                                      Equation 8
                                        vg    dw     dw

Where k is the wave number.

The general relationship between the group refractive index, mg , and the phase refractive index,
mp , for a in a non-isotropic medium, is given by (Budden 1985):
                                                 dm p
                                   mg = mp + w                                          Equation 9
                                                  dw
The distinction between the group and phase velocity (or index) of an electromagnetic wave in a
medium is most important when the velocity of the wave depends upon frequency, as is the case
in the ionosphere. In a vacuum mp mg = c 2.


In terms of group refractive index Equation 6 for the virtual height measured by an ionosonde
becomes:



                                                                                       Equation 10


Which can be expressed in terms of the phase refractive index using Equation 9 as:




                                                                                       Equation 11

If the simplest expression for the phase refractive index were used, which is Equation 2, in this
equation, then this will give an example of a connection between the measured virtual height at a
particular frequency as observed by an ionosonde and the electron density profile.

Using Equation 2 in Equation 11 provides:




                                                                                       Equation 12

Which, by using the relationship between the plasma frequency and the electron density given in
Equation 1, becomes:




                                                14
D42-1                            O BLIQUE IONOSPHERIC SOUNDING                                F INAL R EPORT


                                                                                                Equation 13

The virtual height, hv, or equally the time-of-flight of the radio pulse, is what is measured and
N(h) is what is required in order to analyse how a radio signal would propagate through an
ionosphere for communications purposes. Even in this most simplest case, which excludes both
magnetic field and absorption, the problem arises that the form of the electron density profile
function N(h) is not known. The sketch of a ‘typical’ ionospheric electron density profile in
Figure 4 below is an idealised profile and would be constantly varying in reality. Even this profile
however cannot be expressed fully analytically making it impossible to invert the integral in
Equation 13 to obtain a true solution for N(h) from the measured data. Despite this some
techniques have been developed which come close to achieving this goal.




   Figure 4. A sketch representing the form of a typical day time ionospheric electron density profile.


VERTICAL IONOGRAM PROFILE INVERSION

The process of converting an ionogram into a true-height electron density profile, which can be
used by prediction models and ray tracers, is called profile inversion.

For vertical ionosondes one of the most common programs to do this is POLynomial Analysis
or POLAN, which is a FORTRAN program developed by Titheridge (1985). This solves the
inversion problem by breaking the profile up into simpler sections for which physically
exceptable solutions can be found and using extrapolation and interpolation for the rest. POLAN
inversions of vertical ionograms have been used extensively in this project. An example of the
true-height profile derived from an ionogram using POLAN is shown in Figure 5.




                                                    15
D42-1                          O BLIQUE IONOSPHERIC SOUNDING                             F INAL R EPORT




   Figure 5. A vertical ionogram with a POLAN true height profile. (Picture thanks to the Ionosonde
   monitoring group at RAL).




OBLIQUE IONOSPHERIC SOUNDING
Much of the physical principles outline in the previous section for the vertical ionosonde case is
the same for the oblique ionosonde. However the geometry of an oblique ionosonde makes the
analysis even more complicated.

Obliquely propagating radio waves are refracted and not just reflected and are much more prone
to the effects of horizontal gradients and variations in the ionosphere. The situation is further
complicated by the fact that the radio signals from a transmitter can take a variety of different
routes to reach the receiver, adding to the possible sources of signal distortion and loss.

The geometry of oblique sounding is illustrated in the sketch in Figure 6 along with two
alternative routes for the radio signals.




Figure 6. A sketch showing the geometry for oblique propagation with multi-hop propagation routes
 between transmitter and receiver.

Despite the additional complications in understanding oblique propagation, the oblique
ionospheric sounder offers several important advantages for the understanding of radio systems
propagation over vertical sounding. Firstly it is a technique than enables the ionosphere to be
monitored across large otherwise inexcessible distances such as across oceans. Secondly one



                                                  16
D42-1                         O BLIQUE IONOSPHERIC SOUNDING                            F INAL R EPORT


receiving station can detect a network of remote transmitting stations using a pre-arranged
transmitting schedule. And perhaps most importantly it is a technique that enables investigation
of precisely how radio signals in communications systems actually travel by re-creating real life
propagation scenarios.

As a simple operational field instrument, oblique ionograms provide high-resolution images that
permit communications users to quickly identify frequencies that are propagating between a
selected transmitter station and the receiver station. This is why they are used so much by the
military who need real time channel evaluation. A typical oblique ionogram is show in Figure 7
for propagation over a distance of more just over 3000km from Svalbard in Norway to Chilton
in the UK.




   Figure 7. A typical oblique ionogram for propagation from Svalbard in Norway to Chilton in the UK.
    The effect of propagation paths that involved multiple reflections from the ionosphere is shown.


The effect of propagation paths that involved multiple reflections from the ionosphere is
apparent by the separate curves. By looking at the frequency axis and noting where there is a
receive echo an operator can determine the available comminations bands and the likelihood of
experiencing interference. In this example frequencies between 19-24MHz using a single
reflection from the ionosphere would be available to a HF radio user to establish
communications between Svalbard and Chilton. Between 13 to 15MHz would also available
utilising a propagation route that involves two reflections from the ionosphere and one reflection
from the ground. In principle between 7-11MHz would also be available using either two, three
and four ionospheric reflection propagation. However because of the multiple routes available
the signals would suffer distortion from multipath interference and fading. This could also
contaminate the reception of signals between 19-24MHz because there is more than one route
within the single hop, a high altitude propagation path (high-ray) and low altitude route (low ray).

An oblique ionogram such as this clearly shows the available communications bands and the gaps
where no link could be established. However since the ionosphere changes on time scales which
can be less than a minute, the information available to a communications user on the available
frequencies needs constant updating. Much of the purpose of the oblique sounding project is to
attempt to understand the factors effecting the propagation that will eventually reduce the link
unreliability produced by natural ionospheric variability.




                                                 17
D42-1                           O BLIQUE IONOSPHERIC SOUNDING                               F INAL R EPORT


COMPARING OBLIQUE AND VERTICAL IONOGRAMS
Oblique ionograms look different to vertical ionograms. However if a simple horizontally
uniform, monotonically increasing density ionosphere is considered, a geometrical relationship
between the time of flight and frequencies of vertical and oblique echoes can be found that
illustrates some of the important aspects of oblique propagation. This is illustrated in Figure 8.


            AA                                    A
                                                                  Uniformly increasing
                                                                  density ionosphere
               BB                                  B


                                                                                     Real path
                                        f0
                  Vertical                                                           Virtual path
                                                Oblique


                                               Earth

   Figure 8. A sketch showing the equivalent paths for a vertical and obliquely propagating waves reflected
    at the same real height (B and BB) and at the same virtual heights (A and AA).


With a uniformly increasing electron density ionosphere the real ray path eventually reaches total
internal reflection at the apex of the path. At this point applying Snell’s Law relating the angle of
incidence (i) and refraction (r) to the refractive index, mp = sin(i)/sin(r) to the virtual path, then
because sin(r ) = 0 at the turning point A, then using Equation 2:




              Þ                    wo = w p sec(f 0 )                                         Equation 14

Here f0 is the angle the oblique ray makes with the base of the ionosphere (see Figure 8).
The transmitted frequency normally just w, is here labelled wo, to explicitly identify it as an
obliquely propagating radio wave. This often referred to as Breit & Tuve’s theorem (Davis 1990).


This equation shows that oblique propagation supports much higher frequencies than the vertical
by a factor of sec(f0) . How much greater wo is to the vertical equivalent depends upon the angle
of incidence of the oblique ray with the base of the ionosphere. Equation 14 would have the
effect of extending the x-axis or frequency axis of an ionograms such as Figure 3 by a factor of
sec(f0) if the vertical ionogram were to used to extrapolate to an oblique ionogram. A similar
relationship can be used on the y-axis of an ionogram.

If the apex of the oblique virtual ray path (A) is at the same height as the apex of the vertical
virtual ray (AA) then it can be shown that a very similar relationship exist between the time of
flights (Dt) of the two rays. So:
                                    Dt o = Dt v sec(f 0 )                                Equation 15

This because for the virtual height the speed of the ray is taken to be a constant = c.



                                                   18
D42-1                            O BLIQUE IONOSPHERIC SOUNDING                               F INAL R EPORT



However, this simple geometrical approach could never be expected to hold except for very
short ground ranges where none of the complications of a real ionosphere or a curved Earth
intrude. Nevertheless the validity of these simple projections was tested in real propagation
conditions during the course of this project using independent oblique and vertical
measurements. The performance was then compared with more complicated modelling
approaches.

Maximum Useable Frequency

                               Vertical                           Oblique




Virtual height
Or
time of flight




                           Frequency                              Frequency

    Figure 9. A sketch illustrating how vertical ionograms project to oblique ionograms. The virtual height,
     h’ is equal to the time-of-flight of the radio signal times c.


The sketch in Figure 9 shows the relationship between vertical and oblique ionogram. The points
A to F on the idealised single layer vertical profile, shown on the left, have been mapped to the
points A’ to F ’ on the right hand panel using Equation 14 and Equation 15 to simulate the
equivalent oblique ionogram. Because the projection from one to the other uses sec(f0) there is a
competition between the increasing plasma frequency with altitude and the reduction in the angle
f0 as the altitude increases. Hence the maximum frequency or critical frequency of the vertical
ionogram, fc, does not relate to the maximum frequency on the oblique ionogram. In fact it
corresponds to a much lower part in the vertical profile (labelled C in the figure) in both plasma
frequency terms and altitude terms.

The use of virtual height and time of flight have been used here interchangeably. Over the large
distances of most oblique propagation (> 3000km) virtual height has little meaning. However the
time of flight, which is usually in milliseconds, does have meaning. It is also simple to convert to
virtual height since, by definition, it is just the time of flight times the speed of light in a vacuum.
Virtual height is still useful for vertical ionograms.

In summary the maximum frequency of propagation obliquely occurs at much higher frequencies
than the critical or maximum frequency of the ionospheric layer that relates to the maximum
electron density. The Maximum Usable Frequency obliquely, or MUF, also occurs at much lower
equivalent vertical heights than the critical frequency vertically. This makes interpreting measured
oblique ionograms into electron density profiles more difficult than the vertical case.




                                                    19
D42-1                           O BLIQUE IONOSPHERIC SOUNDING                                F INAL R EPORT




                                        SECTION 2:
                    THE INSTRUMENTATION

THE OBLIQUE IONOSPHERIC SOUNDER

The oblique ionosonde used in these experiments was a FMCW (Frequency-Modulated
Continuous Wave) or ‘chirp’ (Fenwick and Barry, 1965). The particular oblique sounder receiver
is called "IRIS" (Improved Radio Ionospheric Sounder) and was developed by the Radio Science
and Propagation Group at the UK Defense Evaluation and Research Agency (DERA) (Arthur et
al. 1997). The ‘chirp’ compatible transmitter was built by the Department of Engineering at the
University of Leicester in the UK. The nature of a chirp type ionosonde signal is illustrated in
Figure 10. A ‘chirp’ type ionosonde has better height resolution than a typical digitally pulsed
ionosonde. However a digital pulsed ionosonde has better frequency resolution than a chirp
ionosonde.




                                                    Transmitted signal


                                                         ∆t

                                                                echoes




   Figure 10. A sketch showing the nature of a chirp ionosonde signal. The transmitted signal is a phase
    continuous with a ramping frequency at a specified rate. The return echoes from the ionosphere track
    the transmitted frequencies but with a time delay (Dt) corresponding to the flight time to and from the
    reflecting ionospheric layer. Multiple reflections from the ground back up to the ionosphere and down
    again are also seen with time delays that are multiples of the single reflection time Dt.


The broadcast signal is a 10W phase continuous 100kHz/s chirp extending from 2 to 30MHz.
The sensitivity of the receiver is up to ±1dB. GPS (Global Positioning System) time
synchronisation between transmitter and receiver provides a measurement of absolute group
delay or time of flight of the HF signals. The frequency sweep rate of 100 kHz/s provides a
theoretical range resolution of 10ms corresponding to 1.5 km in the group height. Table 1 lists
the operational parameters of the IRIS receiver and Table 2 the operation parameters of the
transmitter.




                                                    20
D42-1                         O BLIQUE IONOSPHERIC SOUNDING                              F INAL R EPORT


            Table 1. A summary of the IRIS receiver operational parameters and specifications.

           Parameter                              Value
           Frequency range                        2- 30 MHz
           Sweep Rate                             100kHz/s
           Selectable signal strength range       10dBm to –170 dBm
           Horizontal resolution                  33 kHz (836 pixels)
           Vertical resolution (group delay)      7.3 ms (682 pixels)
           Baseband signal bandwidth              500 Hz
           Processing Gain                        105 dB
           Signal acquisition window              2 sec
           Sweep duration 2-30MHz                 280 sec
           Sweep interval                         5 minutes
           Operating system                       MS DOS
           Control PC                             486
           IRIS serial number                     IRIS 1965
           Date of manufacture                    Oct 1996 (DERA)
           HF receiver                            Racal RA3711
           Synchronising                          GPS
           Control software (DERA)                IRIS v1.1 to 4.11
           Manufacturer                           UK Defence Evaluation and
                                                  Research Agency (DERA)


             Table 2. A summary of the transmitter operational parameters and specifications.

           Parameter                               Value
           Frequency range                         2- 30 MHz
           Sweep Rate                              100kHz/s (variable)
           Output power                            10W (100 W max)
           Operating system                        UNIX (QNX)
           Frequency synthesiser                   PTS040
           Synchronising                           GPS
           PC                                      486
           Sweep interval                          Programmable to 1 s
           Manufacturer                            Uni. Leicester


Photographs of the oblique sounder receiver and transmitter equipment are show in Figure 11
and Figure 12 respectfully. At each site the equipment used a sloping ‘Vee’ type wire, HF
antenna. Figure 13 shows a sketch of the antenna layout and the connections to the electronics.
The antenna wires consisted of two 40 m long, 5mm diameter copper wires each attached to a
50W to 600 W balum transformer mounted at the top of the 15 m high mast. The other ends of
the antennas were terminated with 600W resistors. In the Scandinavian sites the ground end of
the antennas were raised 2m above the ground level in order to be above the snow level during
winter.

The oblique sounder transmitter and receiver were placed in several locations about the UK and
Scandinavia during the course of this project. Figure 14 shows a map of Europe with the
different locations of the receiver and transmitter marked. The latitudes and longitudes plus the
dates spent at each location are summarised in Table 3.




                                                 21
D42-1             O BLIQUE IONOSPHERIC SOUNDING                              F INAL R EPORT




        Figure 11. A photograph of the IRIS oblique sounder receiver unit.




         Figure 12. A photograph of the oblique sounder transmitter unit.




                                     22
D42-1                         O BLIQUE IONOSPHERIC SOUNDING                               F INAL R EPORT




   A simple 8 cm                                                    40m
   dia Al pole mast                                                Antenna wires
   mounted
   on a 30cm
   metal            15 m
   baseplate



                                                                                                   40m
                                                         37m

                   8m

Antenna feed                         The HF radio antenna and mast
co-axial cable
50 to 200 m
                                                                            Cable ends and mast
                                                                            support guyes
                                                                            anchored into the
                               GPS                Clear view of             ground with Al delta
                               antenna            the sky                   wedges


                                Modem
                                connection
                                to internet         Some suitable housing for
                                                    the equipment
            IRIS


Mains power



  Figure 13. A sketch of the antenna arrangement in this case for the receiver. A similar antenna was
   used for the transmitter sites.




                                                 23
D42-1                         O BLIQUE IONOSPHERIC SOUNDING                              F INAL R EPORT




  Figure 14. A map of Europe showing the RAL oblique sounder receiver sites and the transmitter sites
   throughout the course of this project (see key in Table 3). Signals from non-RAL transmitters were
   also received during the course of the project. The locations of those transmitters are not shown.

  Table 3. Details of the oblique sounder receiver (Rx) and transmitter (Tx) locations and dates spent
  at the sites.

                                                                    Location
           Site name            Owners of site                 [Longitude, East,         Dates at site
                                                               latitude North ]           start - end

    1       Chilton                   RAL                Rx     [-1.33, 51.7]           25/09/1997 –
             (UK)                                                                       10/11/1997
    2     Chilbolton                  RAL                Tx     [-1.44, 51.14]          07/03/1997 –
             (UK)                                                                       20/05/1997
    3   Great_Baddow             GEC-Marconi             Tx     [0.502, 51.71]          13/08/1997 –
             (UK)                                                                       27/09/1997
    4   Hartland Point         British Geological        Rx     [-4.48, 50.99]          13/08/1997 –
             (UK)                    Survey                                             27/09/1997
    5        York                Uni. of York            Rx     [-0.06, 53.94]          09/12/1997 –
             (UK)                                                                       07/07/1998
    6      Lancaster             Uni. Lancaster          Tx     [-02.8, 54.01]          09/12/1997 –
             (UK)                                                                       07/07/1998
    7       Tromsø           Auroral Observatory         Rx     [19.0,    69.42]        27/08/1998 –
         (NORWAY)                                                                       30/09/2000
    8    Karesuvanto        Sodankyla Geophysical        Tx     [22.54, 68.46]          27/08/1998 –
         (FINLAND)              Observatory                                             30/09/2000




                                                  24
D42-1                        O BLIQUE IONOSPHERIC SOUNDING                         F INAL R EPORT



                                    SECTION 3:
         RESULTS FROM THE 1997-1998 UK
        OBLIQUE SOUNDING CAMPAIGNES

                               INTRODUCTION
Rutherford Appleton Laboratory already operates several GPS locked pulsed-amplitude
ionosondes that provide accurate vertical ionograms every 30 minutes. These DigisondeTM
Portable Sounders (DPS) (Reinisch, 1986) are located at Chilton (51.7° N, 1.3° W) and Lerwick
(60.1° N, 1.2° W) in the UK and at Port Stanley (51.7° S, 302.2° E) in the South Atlantic.

The results and analysis presented here are from a series of observational campaigns conducted
in the UK during the period of spring 1997 to autumn 1997 using simultaneous measurements
from GPS locked oblique and vertical ionospheric sounders (Bamford and Levy, 1998). The
propagation over three different ground ranges were recorded, pseudo-vertical (51 km), short
(360km) mid-latitude and medium (3100 km) trans-auroral paths.


    1. Chilbolton to Chilton

The first brief campaign was conducted between the Chilbolton and the Chilton sites, which is
only a distance of 51km, so the soundings were more vertical than oblique. The purpose of this
campaign was primarily an equipment check. It allowed the new IRIS receiver to be assessed and
close comparisons to be made with the digital vertical monitoring ionosonde located at Chilton.
The two ionospheric sounders operate on slightly different techniques. The IRIS sounder uses a
phase continuous chirp transmission whereas the monitoring vertical ionosonde is digital and
therefore transmits a series of coded pulses. The chirp transmission in principle should have
better height resolution than the digital pulse, but not as good precision in frequency. This
equipment test allowed the first direct comparison between the two types of instrument
observing essentially the same part of the ionosphere at more or less the same time.

    2. Great Baddow to Hartland Point

 For this campaign the oblique soundings were made East to West across the southern part of
the UK. The distance between the transmitter and receiver (which was limited by the east-west
extent of the UK) was 363 km with the Chilton monitoring vertical ionosonde located at the
mid-point. This geometry allowed the first direct comparison to be made between oblique
sounding and vertical soundings made at the approximate part of the ionosphere that the oblique
propagation was being reflected from (Levy and Bamford 1997). A direct comparison between
the co-incident measurements and the vertical to oblique simple and complex theoretical
mappings and could be made to test the validity of these mappings commonly in use.

    3. Svalbard (Norway) to Chilton

One of the advantages of an oblique sounder is its ability to make measurements from more than
one transmitter. Whilst at Chilton making the IRIS receiver was also able to receive the
transmissions from the University of Leicester’s chirp transmitter at Ny-Ålysund on the Svalbard
archipelago within the auroral oral (78.91° N, 11.93°E). The distance between the transmitter and
the receiver was 3100km.



                                               25
D42-1                           O BLIQUE IONOSPHERIC SOUNDING                               F INAL R EPORT


A map showing the location of the oblique sounder transmitter (Tx) and receiver (Rx) sites in the
UK is shown in Figure 15.




                                                 Britain

                                          363 km                        Tx
                                        51 km                           Great
                                                VI
                                 Chilbolton                             Baddow
                                   Tx          Chilton
                        Rx
                        Hartland Point



   Figure 15. A map showing the location of the oblique and vertical sounder transmitter and receiver sites
    in the UK during 1997 to 1998. The monitoring vertical incidence (VI) ionosonde was located at
    Chilton.



RESULTS FROM THE CHILBOLTON – CHILTON
CAMPAIGN

CHIRP AND DIGITAL INSTRUMENTAL COMPARISONS
The chirp sounder transmitter was located at the Rutherford Appleton Laboratory owned
Chilbolton Observatory (51.14° N, 1.44° W) which is only 51 km away from the main laboratory site
at Chilton (51.7° N, 1.3° W), where the IRIS receiver was located. The ionospheric monitoring
Digisonde at Chilton, being a pulsed amplitude ionosonde rather than a chirp, could then be used as
a independent source of data. Interference between the ionosondes was avoided by separating the
start times of the frequency sweeps by 50 seconds.

Figure 16 shows nearly simultaneous measurements of the same region of ionosphere by the chirp
and digital ionosondes. The blue and yellow traces are the ordinary and extraordinary components
from the Chilton Digisonde. The black trace is the ionogram from the chirp ‘oblique’ sounder
operating in near vertical path with a ground range of 51km. The IRIS oblique ionospheric
sounder does not distinguish between ordinary and extraordinary echoes explicitly, however both
clearly present. The agreement is excellent between the chirp IRIS results operating over a near-
vertical path and the vertical digital DigisondeÔ DPS-1.




                                                   26
D42-1                            O BLIQUE IONOSPHERIC SOUNDING                                F INAL R EPORT




   Figure 16. An overlay of the almost simultaneous ionograms recorded by a digital (blue and yellow
    trace) and chirp (black trace) ionosondes. There is a 50 second difference in the sweep start times of
    the two ionograms.



RESULTS FROM THE GREAT BADDOW –
HARTAND POINT CAMPAIGN

RESULTS FROM SHORT RANGE MID-LATITUDE PATH
IRIS was then deployed on a 363 km path between, what was at the time, the GEC-Marconi
Research Centre at Great Baddow (51.71° N, 0.5° E) and the British Geological Survey site at
Hartland Point (50.99 °N, 4.5° W). This was an east-west path across the southern part of the UK,
which placed Chilton reasonably close to the mid-point. The locations of the sites are illustrated in
the map in Figure 15 and schematically in Figure 17.

Figure 18 shows a typical vertical ionogram from Chilton and the simultaneously recorded oblique
ionogram for the Great-Baddow to Hartland Point path. This illustrates very well the differences
between vertical and oblique ionograms as outlined theoretically in Figure 9. Even in this very short
path the extension of the vertical ionogram to higher frequencies is clear resulting in a maximum
useable frequency greater than the maximum critical frequency or foF2. It also illustrates how E layer
propagation can compete or overlap with the F layer frequencies depending upon the geometry of
the oblique path and the relative heights and layer critical frequencies.




                                                    27
D42-1                            O BLIQUE IONOSPHERIC SOUNDING                                F INAL R EPORT



                                         N(h) profile



        oblique paths


                           vertical paths




    Oblique Tx                                                                           Oblique Rx
    GPS locked                                   Vertical sounder                        GPS locked
                                                    GPS locked
   Figure 17. A sketch of the relative position of the oblique transmitter and receiver with the vertical
    ionosonde. The electron density profile of the ionosphere is represented as N(h).




   Figure 18 A vertical and oblique ionogram taken within 1 minute each other, both observing
    approximately the same region of the ionosphere.


Oblique ionogram reconstruction using Breit & Tuve’s theorem

In general vertical and simultaneous oblique measurements are not available and oblique propagation
needed for communications assessment is extrapolated from the more common vertical
measurements. This can be done using either the simple arguments outlined on page 18 using
Equation 14 and Equation 15 (Breit & Tuve’s theorem) or using ray tracing, which shall be
covered in the next section. The unique arrangement of vertical and oblique sounders for this
southern England campaign allows these prediction methods to be tested on real data. Thanks to
GPS-locked operation on both instruments, absolute group delays can be measured accurately and
directly compared.




                                                     28
D42-1                          O BLIQUE IONOSPHERIC SOUNDING                         F INAL R EPORT


The theoretical expressions of Equation 14 and Equation 15 have been interpreted in terms of
the measured parameters of virtual height (equivalent to time of flight) and radio frequency. So:


                                       ( D 2) 2 + (hv '( f v )) 2
               f o = f v sec j = f v .                                      Equation 16
                                            hv '( f v )


In this expression fo is th equivalent oblique frequency, fv is the vertical frequency and hv’ the
virtual height or time-of-flight. D is the ground range between the transmitter and receiver (great
circle). Similarly the equivalent height is given by:


                     hv ' ( f v ) = ( D 2) 2 + (hv ' ( f v )) 2      Equation 17


These mappings have been applied in Figure 19. Figure 19 shows the same vertical ionogram
data from the Chilton Digisonde, as was shown in Figure 18, but the frequencies and heights of
the vertical ionogram data have been converted to equivalent oblique frequencies and heights
using the simple geometric conversions shown above.




                                                     29
D42-1                          O BLIQUE IONOSPHERIC SOUNDING                             F INAL R EPORT




   Figure 19. A comparison between the measured oblique sounding data and geometrically scaled vertical
    ionogram data ( fo = fv sec(f)).

The good agreement in form, group delay and frequency illustrated in the individual example
shown in Figure 19 confirms the validity of this simple geometric ‘secant law’ equivalent vertical
to oblique mappings conversion for short path lengths (below ~ 400 km).

The consistency of this simple scaling over many ionograms can be illustrated by representing the
sequence of ionograms in a spectrogram, where the median intense echo at each frequency, for each
ionogram is plotted against time. This is show in Figure 6 for the raw IRIS oblique ionogram data.
Displayed like this, Figure 20 clearly shows the diurnal variation of LOF and MOF
(Lowest/Maximum Observed Frequency) and the effect of sporadic E on the bandwidth and
signal strength over this short ground range link. Figure 21 shows the predicted equivalent
oblique bandwidth based on secant law mapping from the Chilton Digisonde vertical ionogram
data over the same few days.




                                                  30
D42-1                         O BLIQUE IONOSPHERIC SOUNDING                            F INAL R EPORT




   Figure 20. The observed diurnal variation of bandwidth for a short hop (363km) between Great-
    Baddow and Hartland Point in the UK.




   Figure 21. The extrapolated bandwidth for a short hop oblique path (360km) based on the Chilton
    Digisonde DPS-1 ionogram data and a simple fo = fv sec(f) vertical-to-oblique conversion.


The agreement between the observed (Figure 20) and the predicted (Figure 21) bandwidth for
this short hop shown is good. This illustrates how the LOF as well as the MOF is scaled well
over this short distance and also the observation of the occurrence of sporadic E is moderately
consistent.




                                                 31
D42-1                               O BLIQUE IONOSPHERIC SOUNDING                                     F INAL R EPORT


Oblique ionogram reconstruction using an analytical ray tracing

The process of converting a vertical ionogram into a true-height electron density profile N(h)
using a program called POLAN was discussed on page 15. The POLAN program was run on the
Chilton vertical incidence data to provide a vertical electron density profile which is shown in Figure
22. An analytic ray-trace was then carried out to reconstruct oblique incidence group delays, and the
reconstructed ionograms were compared to those measured by IRIS oblique sounder.



         250
                           POLAN profile points


         200               QPS boundaries points
True
Height
(km)
                                                                          Chilton digisonde,
         150                                                              30/4/1997, 17:00 UT
                                                                          error tolerance: 2%
                                                                          8 QP segments
         100
            0                 1             2               3             4              5              6
                                                Plasma Frequency (MHz)
   Figure 22. A POLAN electron density profile (expressed as plasma frequency see Equation 1) for a vertical
    ionogram recorded at Chilton on 30/4/1997 at 17:00 UT. POLAN produces discrete points which are shown
    as red circles. The blue crosses mark the QPS boundary points used in the analytical reconstruction of the profile.
    The solid black curve shows the QPS fit.

POLAN was run in a “no-valley” mode, meaning a simple cusp was used to merge the E and F1
layers as recommended in (Krashenninikov et al, 1996). In order to use an analytical ray trace on the
electron density profile (which can be interpreted as refractive index profile), the discrete points of
the POLAN profile must be replaced by an analytical expression, which can describe the sections of
the profile. Since no single simple expression can adequately be fitted to the experimentally derived
electron density profiles, the profile has to be broken up into smaller sections to which an analytical
expression can be fitted. The approach propose by Dyson et al (1988) was to use a series of quasi-
parabolic equations fitted to each segment of the true height profile.

In each quasi-parabolic segment has the form:
                                        Ai Bi
                         f Ni ( r ) =
                           2
                                           +   + Ci                               Equation 18
                                        r2   r

Where f Ni is the plasma frequency and r the distance from the centre of the Earth. Following the
                                                           (             )
work of Chen and Bennett (1990), the coefficients Ai , Bi , Ci are determined from a least squares
fit to true height profiles, ensuring continuity of the multiple quasi-parabolic (MQP) profile and
profile gradient at joining points between segments. The number of segments is not known in
advance, but is determined by the fitting algorithm. In this case 8 segments were obtained. The
boundary points between each the quasi-parabolic segments are shown as blue crosses on the true
height profile shown in Figure 22.



                                                         32
D42-1                             O BLIQUE IONOSPHERIC SOUNDING                                 F INAL R EPORT


The advantage of the multiple quasi-parabolic (MQP) profile is that the ray-trace integrals giving
ground range and group delay are given in closed form (Dyson and Bennett 1988), allowing very fast
computations. This is particularly useful in oblique incidence ionogram reconstruction, where a
homing-in procedure has to be carried out at each frequency to find the rays linking transmitter and
receiver, meaning thousands of rays have to be traced to reconstruct the full ionogram. Typical run-
time for ionogram reconstruction is a few seconds on an average personal computer when written in
either C or FORTRAN.

Once the analytical expression for the electron density profile is formed, established techniques
(described in Dyson and Bennett 1988) can be employed to determine ray group path using a
homing-in facility at each frequency and with adjustable ground ranges.

In Figure 23, an example is shown of the calculated path of that radio waves at 4 and 8 MHz would
take, when launched into an ionosphere whose electron density profile has been obtained via a
multiple quasi-parabolic fit to a POLAN profile derived from vertical sounding ionogram following
the ray tracing techniques outlined in Dyson and Bennett 1988.



                       250                                         Ray visualisation
                                                                   Using a MQP profile fitted to
                                                                   Chilton 30/4/97 16.59 sounding
                       200
                                                                   Elevation angles 30 to 80 deg. in
                                                                   5 deg. steps
                       150
       Height (km)

                       100                                                 8 MHz
                                                     4 MHz
                         50


                                        200       400        600        800       100
                                           Ground range (km)
   Figure 23. Ray paths calculated using the multiple quasi-parabolic (MQP) profile fitting analytical ray
    trace.

Figure 24 shows a comparison between actual oblique sounder measurement and the
MQP/POLAN ray trace model results for an oblique propagation path of 363 km. Shown in blue is
the IRIS group delays measured on the Great Baddow to Hartland Point propagation path. In red
the model results based on independent measurement of the vertical electron density profile at the
mid-point of the path obtained from a virtually simultaneously recorded vertical incidence ionogram
at Chilton.




                                                      33
D42-1                           O BLIQUE IONOSPHERIC SOUNDING                            F INAL R EPORT




   Figure 24. A comparison between measured and MQP ray-trace reconstructed oblique ionograms. The
    blue trace is the measured oblique ionogram between Great Baddow and Hartland Point (363km) on
    16/08/1997 at 14:59UT. The red trace is a MQP ray analytical ray trace reconstruction of the
    ionogram based upon a POLAN profile obtained from a mid-point vertical incidence ionogram
    recorded at Chilton on 16/08/1997 at 15:00UT.


Agreement is generally good between the reconstructed oblique ionogram also shown in Figure 24
and the ordinary IRIS trace, with a very good reconstruction of the F2 mode, and weaknesses as
expected in the transition region between the E and F1 layers. It is quite encouraging that despite all
the approximations involved, a reasonably accurate simulation of link behaviour can be made.

A simple way of identifying the main source of error in the model is to go full circle with the
POLAN/MQP/ray trace model and use it to re-create a simulation of the vertical ionogram
from which the model input was derived, and then compare this with the original. This has been
done in Figure 25.

The comparison shows the vertical incidence ionogram measured by the Chilton Digisonde (in blue
and orange) on 30/04/1997 at 16:59 Universal Time (UT) and the model re-construction in red. The
orange trace is the measured extraordinary ray propagation that has not been included in the ray
trace model. A strong E-layer is clearly visible as well as the F1 and F2 layers in the original
ionogram. The model reconstruction re-creates the measured ionospheric layer heights and critical
frequencies well except for the F1 in inflection. As can be seen in Figure 22 the MQP fitting to the
POLAN profile is accurate to within 2%. The difficulty lies with POLANs ability to calculate the F1
characteristics from the ionogram. This then has knock-on effects when this is used as a model input
for the ray trace. This problem would arise no matter which ray tracing technique is applied. It
highlights the extreme difficulty in ionogram profile inversion mentioned on page 15.




                                                  34
D42-1                         O BLIQUE IONOSPHERIC SOUNDING                            F INAL R EPORT




   Figure 25. A comparison between a measured vertical incidence ionogram recorded at Chilton and the
    MQP ray trace reconstruction of the vertical ionogram based upon a POLAN profile created from the
    original measurement.



How do the two approaches compare?

The plot in Figure 26 the same data is used to generate a similar comparison between
measurement and modelled data to that shown in Figure 24. Here though the modelled or
predicted ionogram is generated using the simple geometrical approach outline on page 18 and
28 of Breit & Tuve theorem.

By comparing Figure 26 to Figure 24 it can be seen that on this occasion the simple geometric
works much better than the complicated POLAN/ray trace approach over these short distances
of less than 400km. However most HF communication is over much greater distances 1000s of
km. In these conditions the simple Breit and Tuve theorem approach is not expected, nor
intended, to give good estimates. Further it is unable to deal with the multiple bounce
propagation necessary. Under these conditions only a ray tracing model, either analytical or
numerical, could be expected to provide adequate predictions of the main propagation
parameters.




                                                35
D42-1                          O BLIQUE IONOSPHERIC SOUNDING                             F INAL R EPORT




   Figure 26. A comparison between measured and Breit & Truve extrapolated vertical to oblique
    ionogram. The blue trace is the measured oblique ionogram between Great Baddow and Hartland
    Point (363km) on 16/08/1997 at 14:59UT. The red trace is the measured Chilton vertical
    ionogram data recorded at the path mid point at the same time but converted into an oblique ionogram
    of the same distance using the simple sec(ø) mapping.



     RESULTS FROM THE SVALBARD – CHILTON
               MEASUREMENTS

RESULTS FROM 3000KM RANGE TRANS-AURORAL PATH
Figure 27 shows a typical oblique ionogram recorded at Chilton (51.7°N, 1.33°W) from the chirp
transmitter on Svalbard (78.91°N, 11.93°E), which is operated by Leicester University. The map
in Figure 28 shows the location of the transmitter and receiver and an instrument in Northern
Finland that measures the ionospheric absorption (riometer) which is located at Kilpisjärvi
(69.03°N, 20.5°E).




                                                  36
D42-1                          O BLIQUE IONOSPHERIC SOUNDING                              F INAL R EPORT




                       3F MOF 2F MOF                     1F MOF
  Figure 27. A typical oblique ionogram for propagation from Svalbard in Norway to Chilton in the
   UK.



                                          Svalbard
                                          transmitter




                                 3100 km                     Kilpisjarvi
                                                             Riometer




                           Chilton
                           receiver

  Figure 28. A map of Northern Europe showing the location of the IRIS oblique sounder receiver at
   Chilton in the UK and the location of the transmitter on Svalbard. The location of the Kilpisjärvi
   riometer (69.03N, 20.5E), which measures the ionospheric absorption, is also indicated.




                                                  37
D42-1                          O BLIQUE IONOSPHERIC SOUNDING                             F INAL R EPORT


Figure 29 shows many Svalbard-Chilton ionograms recorded over several days in 1997. Here
each ionogram has been plotted as a single time point to produce a plot of frequency against
time. In this way the observed spectrum occupancy for this propagation path can be seen very
easily.




   Figure 29. An spectrogram of the IRIS data recorded at Chilton from the chirp transmitter located on
    Svalbard .

The diurnal variation of the available bandwidth can be clearly seen in the first 2 days of the four
days plotted. However, ionospheric absorption disrupts this pattern for the latter two days.

The bandwidth of each of the multiple hops can be clearly seen for this trans-auroral oval path.
Between 20 and 25 MHz propagation is only via a single ionospheric F-layer reflection, the
1F MOF (1 reflection F layer, Maximum Observed Frequency). Yet below 15 MHz the multiple
F-layer reflections can be seen to extend the available bandwidth. The 2F MOF, 3F MOF and
even 4F MOF are all available during certain conditions on the 22nd and 23rd. There is evidence
of scatter propagation above 25 MHz at noon on the 23rd. Increased auroral absorption
prevented propagation on 25th October 1997 which is coincident with an increase in geomagnetic
activity on the 24th and 25th .




                                                  38
D42-1                        O BLIQUE IONOSPHERIC SOUNDING                           F INAL R EPORT




                           Lerwick vertical ionosonde




  Figure 30. The Polar Cap radio Absorption (PCA) event of November 1997. Top: The 30 MeV
   protons and 0.1-0.8 nm X-ray flux as recorded by the GOES-9 satellite on the 5th-11th November
   1997. Middle: The oblique HF propagation between Svalbard and Chilton during this event.
   Bottom: The Lerwick (1.2W, 60.1N) vertical ionosonde ionogram records.




                                               39
D42-1                         O BLIQUE IONOSPHERIC SOUNDING                           F INAL R EPORT


In the Svalbard-Chilton spectrogram presented in Figure 30 another period reception blackout
can be seen to occur on the 7th November. But on this occasion the origin of the loss of
propagation was identified from satellite observations. The 30 MeV proton flux data from the
GOES-9 satellite is shown in the top panel of in Figure 30, reveals the solar terrestrial origins of
the increased auroral absorption for this particular disruption in propagation. The GOES
satellites (Geostationary Operational Environmental Satellites) are operated by NOAA, the
National Oceanic and Atmospheric Administration and monitor the emissions from the sun
coming towards the Earth (GOES MISSION 1998). The top panel in Figure 30 shows an
increase by a factor of nearly 1000 in the 30 MeV protons flux passing the GOES-9 satellite
location on 3rd - 9th November 1997 (shown in black). In red the 0.1-0.8 nm X-ray flux recorded
by GOES-9 is also shown. This shows that the burst of X-rays occurred at the same time as the
rise in proton numbers. However the increase in X-ray emissions was very short-lived. So it is the
arrival of the large number of energetic protons some hours later at the northern auroral region is
the probable cause of the sustained interruption in the Svalbard-Chilton trans-auroral link,
producing a Polar Cap radio Absorption (PCA) event. The increase in ionospheric absorption
can be seen by the trace in black overlaid on the oblique sounder data in the lower panel of
Figure 30. This is data from the riometer instrument located at Kilpisjärvi in Finland. The
riometer passively measures the ionospheric absorption by monitoring the strength of radio
signals from astronomical sources and is therefore an independent measure of the ionospheric
absorption. This instrument shall be discussed in more detail in the next section on high latitude
observations. The increase in absorption observed on the riometer coincides with the loss of
signal over the propagation path between Svalbard and Chilton.

The lowermost panel of Figure 30 shows the ionogram data from the Lerwick vertical ionosonde
located in the Shetlands (1.2W, 60.1N). This shows how during this November 1997 PCA event
which prevented high latitude propagation, there was a knock-on effect at lower latitudes. The
daily maximum frequency shown in the lower panel of Figure 30, which relates to the foF2, can
be seen to decrease from 7 to 8MHz to below 6MHz for the days following the proton pulse
arrival. This clearly demonstrates how events at high latitude also effect lower latitude
ionosphere.

The significant aspect about this observation of this November 1997 event is that the proton
pulse passed the GOES satellite nearly 24 hours ahead of the HF blackout occurring offering the
potential for early warning.

                    CONCLUSIONS TO SECTION 3
A number of conclusions can be drawn from the results of the UK oblique sounding
measurement and analysis campaign between 1997 and 1998. The analysis demonstrates the
validity quantifiably of simple fo = fv sec(f) vertical-to-oblique conversion between oblique and
vertical ionogram measurements for distances of less than 400km. This was only possible because
of the accurate GPS timing on two sets of independent observations made where the vertical
reference was located at the oblique path mid-point. This result validates the use of this approach
over short distances as very quick and accurate way to produce propagation predictions in
computer models. This could now be applied with confidence to a point to point HF web
prediction facility, at least for shorter propagation paths.

The two sets of independent ionosonde data were also used to test more sophisticated models
using ionogram inversion and ray tracing. The comparisons demonstrated the validity of this
technique for determining the propagation critical parameters of maximum observable frequency
MOF and virtual height of the reflecting layers. However the comparison did highlighted the
difficulties in ionogram inversion of certain aspects of the ionospheric profile such as F1 ledge.
This would effect any ray tracing technique used.




                                                40
D42-1                         O BLIQUE IONOSPHERIC SOUNDING                         F INAL R EPORT



The extensive database of GPS oblique sounding measurements gathered during this project has
the potential to be used much more extensively. There is an opportunity to model more of the
long propagation paths recorded using vertical ionosonde comparisons. This could be a statistical
study of the key parameters both obliquely and vertically. The analysis would involve further
development of analytical ray trace to cope better with long and multiple hop paths.

The use of satellite data in the analysis of individual storm events, such as the November 1997
event, clearly shows the potential for developing early warning facilities of disturbances to
communication and navigation systems for the UK. In the example shown the solar proton pulse
that was identified as the cause of the propagation disturbance passed the GOES-9 satellite
nearly 24 hours ahead of the HF blackout occurring at mid latitudes.

There is clearly scope for more analysis of the origins of the propagation disturbances using
satellite observations. A better understanding of the underlying links between cause and effect
would clearly be beneficial to future communication system early warning facilities, be that at HF
or higher frequencies, or indeed satellite protection systems.




                                               41
D42-1                         O BLIQUE IONOSPHERIC SOUNDING                          F INAL R EPORT



                                    SECTION 4:
      RESULTS FROM HIGH LATITUDE
    OBLIQUE SOUNDER MEASUREMENTS

                                     RATIONALE

One of the most significant omissions currently in the prediction of a HF channel is the inability
to accurately predict the received signal strength over a given HF link. Without accurate
information on the amount of signal loss to be expected a broadcaster will tend to over
compensate rather than risk loosing a connection. The wide spread use of excessive transmitter
powers by broadcasters results in unnecessary interference and frequency congestion.

The amplitude of propagating radio waves can be attenuated by a host of reasons. These include
multipath interference, focusing due to large-scale inhomogeneities and polarization fading. Ray
tracing has had appreciable success in accommodating most of these characteristics when given
accurate information on the state of the ionosphere. However, the calculations of signal
amplitude and phase produced by a ray trace program cannot allow for the main source of
absorption which is due to the collisional exchange between the free electrons and the other
particles of the atmosphere. The D layer of the ionosphere is the major culprit here. To include
the collisional absorption requires an accurate assessment of the complicated chemistry of the
upper atmosphere. This is not easy to evaluate. The Lancaster-Sodankylä ion chemistry model is
available to provide these estimates for the steady state (Burns et al, 1991, del Pozo 1997). But
before this model can be used to estimate the accumulative, non-deviate attenuation of a HF ray
making its way through the ionosphere two important facets need to be included. Firstly the non-
deviate absorption needs to be evaluated and secondly the path the ray takes through the
ionosphere has to be determined. For this latter part a ray tracing program, such as the one
produced by Västberg et al (1997), is ideally suited. The "RaTS" numerical ray-tracing program is
written in C++ in an object orientated way. This makes it highly adaptable to be combined with
the Lancaster-Sodankylä model.

The high latitude oblique sounding campaign conducted between 1998 and 2000, involved a new
joint instrument experiment to make for the first time direct measurements of both ionospheric
absorption and received signal strength of HF channel, for the same region of the ionosphere
(Bamford et al 1999).

The ionospheric absorption was monitored using an instrument that is co-incidentally named
“IRIS”. The Imaging Riometer for Ionospheric Studies (IRIS) (Browne et al 1995) at Kilpisjärvi
in Northern Finland has a field of view that spans an area of 240 square kilometres. This is
sufficiently large for the radio waves from a suitably placed oblique ionosonde to only propagate
within this region.

The geographical location in Northern Finland is close to the EISCAT incoherent radar and
associated facilities. This means that accurate independent information on the ionospheric
profiles would be available as inputs for both ray tracing and absorption models in any analysis.




                                                42
D42-1                          O BLIQUE IONOSPHERIC SOUNDING                         F INAL R EPORT


GRAPHICAL INTERFACE ADDED TO THE RAY TRACE
During the course of this project a graphical, more user friendly, front end was created for the
otherwise command line operated “RaTs” numerical ray trace. The graphical front end was
created in IDL 5.0, which in principle would allow operation on any platform, provided
adaptations were made to the ray trace to operate on the particular operating system. Currently
the ray trace runs in Unix. However the C++ code is standard ANSI so it ought to be suitable to
be compiled on other operating systems such as Windows NT. Figure 31 shows a typical plot
produced from the calculations of the "RaTS" ray trace. A picture of the graphical interface and a
multiple plot output from the "RaTS" ray trace is shown in Figure 32.




   Figure 31. An example of the calculations performed by the ray tracing program.




                                                  43
D42-1   O BLIQUE IONOSPHERIC SOUNDING   F INAL R EPORT




                    44
D42-1                           O BLIQUE IONOSPHERIC SOUNDING                               F INAL R EPORT


   Figure 32. Multiple outputs from the modified ray tracer program. The graphical interface to the ray
    tracer is written in IDL, the ray tracer itself is in C++.



THE IMAGING RIOMETER
The IRIS system at Kilpisjärvi (69.05°N, 20.79°E) is operated by the Lancaster University (UK)
in conjunction with Sodankylä Geophysical Observatory, Finland. The system operates at 38.2
MHz within a low noise band protected for radio astronomy. IRIS monitors the radio signals
from extra-terrestrial sources such as stars and galaxies that are virtually invariant in time.

The ionospheric absorption is calculated from the difference between the received signal power
and the measured "quiet day" level. The quiet-day curve shows the background cosmic noise
level variation with sidereal time as the background cosmic noise is not constant in all directions
(Browne 1995).




   Figure 33. The projection onto the ionosphere at 90 km altitude of the Kilpisjärvi IRIS imaging
    riometer beams. The locations of the oblique sounder transmitter and receiver below the riometer's
    observational region are also indicated.

An array of 49 narrow monitoring beams with widths between 13° and 16° is produced by the
phasing of 64 crossed-dipole antennas. The basic scanning interval of the array is one second.
Figure 33 shows the projection of the 49 IRIS beams onto the ionosphere.

The spatial resolution provided is approximately 20 km at an altitude of 90 km from most of the
49 beams. The horizontal distribution of radio absorption events is imaged over a total area of
240 square kilometers of the ionosphere.


THE OBLIQUE IONOSPHERIC SOUNDER
The propagation mode and signal strength variation of HF signals traversing the riometer region
has been recorded using the frequency swept oblique ionosonde.

The variations in the automatic gain and the received HF signal are both recorded for each
frequency between 2 and 30MHz. Only the relative signal strength is needed to study temporal
variations in the absorption at fixed frequencies.




                                                   45
D42-1                          O BLIQUE IONOSPHERIC SOUNDING                            F INAL R EPORT


For the inter-instrument comparisons the ground range between the RAL transmitter and
receiver of the oblique ionospheric sounder is deliberately very short at less than 200 km. This
was so that the HF radio signals would pass through the D layer of the ionosphere entirely within
the central field of view of the riometer beams on both the upward and downward legs of their
journey.

The ionosonde receiver was located at the Auroral Observatory in Tromsø, Norway (69.42 °N ,
19.0°E). The RAL transmitter is in a remote site in Northern Finland near the town of
Karesvanto (68.46°N, 22.54°E).

However during the observations campaign the oblique sounder also made observations of
remote transmitters between the short-hop observations. Oblique propagation paths between
Svalbard (NOR) (78.91°N, 11.93°E) and Tromsø, Farnborough (UK) (51.27°N, 0.63°E) and
Tromsø and between Angle (UK) (51.68°N, 5.07°W) and Tromsø were also recorded.

The locations of the two sites using in the riometer comparisons are shown relative to the region
imaged by the riometer is included in Figure 33. This geometry is illustrated in the sketch shown
in Figure 34.




   Figure 34 A vertical projection of the imaging riometer beams at 90km (as shown in Figure 33.) and
    the path of an oblique incidence HF radio wave between Karesuvanto and Tromsø.



  RESULTS FROM THE KARESUVANTO – TROMSØ
                CAMPAIGN
A typical oblique ionogram from the RAL chirp receiver is shown in Figure 35. Because the
oblique sounder measures the absolute group delay of the signals, it is possible to isolate the
echoes returned from the different layers. For instance, in Figure 35, echoes that took less than
1.3ms were reflected only by the E layer. In this way any biasing of the signal strength due to
multiple reflections between the ground and the ionosphere is also excluded. However
polarization discrimination is not available with this instrument.




                                                 46
D42-1                         O BLIQUE IONOSPHERIC SOUNDING                          F INAL R EPORT




   Figure 35 A typical ionogram from the Karesuvanto-Tromsø short-hop oblique HF sounder. An
    absolute group delay of 1ms over the 200km ground range corresponds to a virtual height of
    approximately 110km vertically.


The change in the relative received signal strength of 2.5MHz (3kHz bandwidth) HF radio wave
is shown in Figure 36(a) for the day of 5th December 1998 (Bamford et al 1999). At this
frequency the absorption will be predominately non-deviative in origin. The absolute group
delays or time-of-flights of the received HF signals were restricted to echoes from the E layer.
During approximately 04:00UT to 12:00 UT the ionospheric absorption reached levels such that
propagation was prevented at this frequency. This period corresponds to enhanced levels of D
layer absorption as indicated by the data from the imaging riometer for this same period. This is
shown in Figure 36(b). The estimated daily mean Ap was 12 for this day.

The absorption in Figure 36(b) is an average value for the three central riometer beams that lie
within the path of the propagating oblique HF rays. On this occasion, the HF propagation at
2.5MHz was inhibited when the D-layer absorption measured by the imaging riometer exceeded
a threshold of 0.13dB. This value was approximately the same for all the frequencies below
5MHz on this day. The maximum observed frequency for non-sporadic, non-auroral E-layer
reflections was 5MHz for this path (EMOF). This corresponds to a foE of about 2.5MHz over
Kilpisjärvi.




                                                47
D42-1                           O BLIQUE IONOSPHERIC SOUNDING                                F INAL R EPORT




                                                                                       (a)




                                                                                       (b)




   Figure 36 (a) Relative median signal strength variation of a 2.5MHz HF signal propagating from
    Karesuvanto to Tromsø reflected from heights below 150km (E-layer). (b) The measured amount of
    ionospheric absorption recorded by the imaging riometer. The baseline is derived from the quite day
    reference data (see (Browne et al 1995)).



  RESULTS FROM THE COMBINED ABSORPTION
          AND RAY TRACING MODEL

The first step in attempting a comparison between the measured HF signal strength and the
ionospheric absorption with the calculated values for each from a combined ray trace and
absorption model, the electron density of the ionosphere has to be determined. This is the
primary input into both the ray trace and the absorption model.

Figure 37 shows a POLAN (see page 15) electron density profile obtained form the Dynasonde
(Wright 1998) digital vertical ionosonde based at the EISCAT facility in Tromsø. The POLAN



                                                   48
D42-1                          O BLIQUE IONOSPHERIC SOUNDING                           F INAL R EPORT


profiles could have been obtained using the oblique ionosonde data on this occasion, as the
propagation path is near vertical. However by using the Dynasonde vertical sounder data then
the model input parameter are derived from independent measurements.




   Figure 37. A POLAN true height electron density profile obtained from a Dynasonde (Wright 1998)
    vertical ionosonde measurement made at Tromsø.

When a radio wave enters the environment of the free electrons of the ionosphere, there are
several factors that determine how much of the radio wave’s energy is absorbed all of which
change with altitude. Primary amongst these is the concentration of electrons, also the
temperature of the electrons and the temperature and density of the surrounding neutral and
ionised atmosphere. The interactions between the electrons, ions and various species of neutral
atoms and molecules are very complex and shall not be gone into here. For further information
the reader is referred to the work by Burns et al (1991) and del Pozo et al (1997) as these relate to
the specific model whose results shall be quoted here.

The Lancaster-Sodankylä ion chemistry model provided estimates for the steady state electron
temperature and mean electron collisional rate with altitude. The plots of these two are shown in
Figure 38(a) and (b) over laid on the same POLAN profile previously shown in Figure 37.
The resulting model estimate of the ionospheric absorption per km for radio waves at 2.5 MHz is
shown in Figure 39.

The jagged part lower part of the absorption profile is caused by the discrete nature of the
POLAN profile input data and the sparcity of data points in the lower most part of the
ionospheric profile. The consequences of this shall be considered later.

Any calculation of the total loss of signal strength of a radio wave propagating between
transmitter and receiver through an absorbing medium requires knowledge of the path the ray
took through the medium.




                                                 49
D42-1                            O BLIQUE IONOSPHERIC SOUNDING                                F INAL R EPORT




                                Ne(h)                                                      Ne(h)


          Electron
          temperature




                                                                                    Collisional rate




   Figure 38. (a) The electron temperature profile, (b) mean electron collisional rate calculated using the
    Lancaster-Sodankylä ion chemistry model overlaid on the POLAN electron density profile (dashed
    curves).



                                                            Ne(h)




                                          Absorption per km




   Figure 39. A plot of the ionospheric absorption per km for 2.5MHz radio waves determined using the
    Lancaster-Sodankylä ion chemistry model overlaid on the POLAN electron density profile (dashed
    curve).


The most established ray trace for ionospheric propagation is a numerical ray trace program
written in FORTRAN developed by Jones and Stephson (1975). This program is based upon the
equations of Haselgrove (1954) that describe the propagation of radio wave through a general
ionosphere. The ray trace used here is a development of this earlier work where the ray trace
program has been re-written in C++. This object-orientated language offers a flexibility to
interchange and add model components such as different electron density models and different
absorption models easily. The ray trace is called "RaTS" (Ray Tracing System) and was developed
by Västberg (1997).




                                                     50
D42-1                         O BLIQUE IONOSPHERIC SOUNDING                            F INAL R EPORT


An example of the output of the RaTS ray trace is shown in Figure 40 using the POLAN profile
for the electron density previously shown in Figure 37 and used with the absorption model plots
of Figure 38 and Figure 39. Here only a single 2.5MHz ray is shown. However all rays between
the transmitter and the receiver at all frequencies were calculated by the ray tracing program.




   Figure 40. A single 2.5 MHz ray path calculated using the RaTs numerical ray trace and the
    POLAN electron density profile based on Tromsø Dynasonde observation of the 5th December 1998
    at 17:42UT.

When the results of Figure 39 and Figure 40 are combined and the incremental radio absorption
along the path shown in Figure 40 is plotted, the result is Figure 41. This plot suggests that the
total ionospheric absorption of a 2.5MHz HF ray travelling between the transmitter and the
receiver was 6dB. This can then be compared with the total ionospheric absorption recorded by
the riometer. Since the 24 beams of the imaging riometer allow for some spatial discrimination, it
is necessary to determine using projections of the riometer beams and the ray trace path, which
of the beams the HF travelled through. This is illustrated in Figure 42. Here a 6.5MHz ray has
been used for clarity of the image.




   Figure 41. A plot of the accumulative absorption experienced by a 2.5 MHz based on the model
    results of the Lancaster-Sodankylä ion chemistry model and the RaTs ray trace.




                                                 51
D42-1                         O BLIQUE IONOSPHERIC SOUNDING                          F INAL R EPORT




   Figure 42. A projection of the path of a 6.5MHz HF ray on the IRIS imaging riometer beams.


Because the riometer operates at 38MHz, a much higher frequency than the HF propagation, a
correction is necessary. If the Earth's magnetic field is not included it is possible to relate the
auroral ionospheric absorption A(dB) with the radio frequency, f, and the zenith angle of the ray,
thus (Browne et al 1995):

                                     sec(q )
                          A(dB ) µ                                Equation 19
                                       f2

When this correction is applied to the model prediction of 6 dB for a 2.5 MHz ray with a zenith
angle 33° then the result is a predicted non-deviate auroral absorption of 0.3 dB at 38MHz.

This agrees very well with the measured absorption of 0.3 dB measured by the imaging riometer
for this time. The beam averaged riometer ionospheric absorption is shown in Figure 43. The
vertical line indicates the time of the electron density profile used in the models.



                                                          Time of the Ne(h)
                                                          profile




   Figure 43. The observed ionospheric absorption by the riometer at 38 MHz.




                                                 52
D42-1                        O BLIQUE IONOSPHERIC SOUNDING                         F INAL R EPORT


                   CONCLUSIONS TO SECTION 4

The quantitative comparison between measured D layer absorption and the variations on
received signal strength of a obliquely propagation HF signal have never been carried out before
using independent, co-located instruments.

Results and analysis have been conducted using a combined observational campaign using the
Imaging Riometer for Ionospheric Studies (IRIS) at Kilpisjärvi in Northern Finland, and the
oblique HF sounder spanning the same region of the ionosphere.

These experimental observations form the basis for validating the predictions in the changes in
absorption based on the predictions of the Lancaster-Sodankylä ion chemistry model combined
with the ray path calculations of the RaTS numerical ray tracing program.

The comparisons between model predictions and measurement show excellent agreement in the
case studies made.

These observations, combined with the modelling, represents progress towards being able to
include ionospheric absorption in HF channel simulations.

The oblique sounding measurements made at the Tromsø site were made in the same region as
so many ionospheric and geophysical instruments. These instruments include EISCAT tri-static
incoherent radar, CUTLASS HF coherent radar, IMAGE magnetometer chain and the
ionospheric tomography receivers amongst others. The oblqiue sounding observations database
offers a large potential for inter-instrument analysis. A fact that has been recogised by other
researchers, such as the University of Leicester, who have recently deployed oblqiue sounders in
this region.

The data between Svalbard and Tromsø, and between the UK and Tromsø, shows some
interesting disturbances that warrant further study.




                                               53
D42-1                          O BLIQUE IONOSPHERIC SOUNDING                           F INAL R EPORT




                            REPORT SUMMARY
An extensive database of oblique sounding ionospheric propagation measurements has been
acquired both at mid and high latitudes. This database has the advantage of GPS timing the lack
of which seriously hampered earlier propagation studies. The choice of transmitter and receiver
sites in the UK and abroad offered the opportunities for inter-instrument comparisons. In the
UK, the mid-point of the oblique propagation path was observed by an independent vertical
ionosonde. This allowed comparisons to be made between vertical and oblique mapping
methods over short distances using actual independent measurements. This analysis confirmed
the applicability of simple fo = fv sec(f0) over short (less than 400km) mid-latitude paths. Although
the major parameters were adequately predicted, the electron density inversion and ray tracing
analysis highlighted the sensitivity of such techniques to the accuracy of the electron density
profile.

The results of the high latitude oblique sounding measurement campaign allowed the first
comparisons between the measured ionospheric absorption and HF signal strength at the same
time, for the same region of ionosphere. During the course of this project, both an analytical and
a numerical ray trace were acquired. The particular analytical ray trace offered the potential for
high-speed ray tracing, but lacked the precision and versatility of the numerical ray trace.
The numerical ray traced was used in the high latitude oblique sounding analysis combined with
an ionospheric absorption model. The direct comparisons between two independent
measurements and the modelling proved very successful.

All the oblique sounding data collected during the project has been assembled onto CD ROMS
along with file format information and IDL programs to read and plot the data. The full list of
the dates and paths recorded are listed in Appendix

An interactive web facility also exists at:
"http::/www.rcru.rl.ac.uk/iono/irisweb/select_many_data.html ".
At this web site the hourly oblique ionograms can be displayed and automatically generated daily
plots estimating the minimum frequency and variation in signal strength at 3.0 MHz. The data is
selected by date. Data from other ionospheric and relevant geophysical instruments, such as the
magnetometers, can also be displayed for the same time periods.

The experience of propagation and ionospheric absorption effects gained during this project lead
to another highly successful Radio communication Agency project. The project titled "Radio and
the 1999 UK Total Solar Eclipse" (Bamford, 2000) relied upon an understanding of the effect of
the ionosphere on more than just the HF band, but the VLF, LF and MF bands. The eclipse
project would not have been possible if it had not been for the foundations laid by the Oblique
Ionospheric Sounding project.




                                                 54
D42-1                         O BLIQUE IONOSPHERIC SOUNDING                          F INAL R EPORT



                                  FUTURE WORK
The database of GPS oblique sounding measurements gathered during this duration of this
project has not been fully utilised and has potential to be used much more extensively particularly
for statistical comparisons and further case studies. Outlined below are some specifics areas of
development that might be considered.

At mid latitudes, there is the opportunity to model more of the long propagation paths recorded
using vertical ionosonde comparisons. This could be a statistical study of the key parameters both
obliquely and vertically. The analysis would involve further development of analytical ray trace to
cope better with long and multiple hop paths.

Within this project, a single example was presented of the solar origins of a HF blackout. The
pulse of protons passed the monitoring satellite nearly 24hours before the disruption occurred to
the comminations system. There is clearly scope for more analysis of the origins of the
propagation disturbances using satellite observations. A better understanding of the underlying
links between cause and effect would clearly be beneficial to future communication system early
warning facilities, be that at HF or higher frequencies, or indeed satellite protection systems.

The combined ray trace and absorption model could be made into a generally available robust
ionospheric propagation tool with some development. Currently the analysis is very time
consuming and user intensive to do one instant in time. What is needed is the ability to easily do
a series of time intervals so that the calculated and measured absorption can be compared. This
way the evolution of disturbances can be investigated. This would also allow an investigation of
the HF signal cut off levels variation with frequency correlation with the riometer observed
absorption level.

There is the potential to use the unique arrangement of the high latitude campaign and the ray
trace and absorption models to examine deviate and non-deviate components of absorption. This
could then be extended to mid latitudes and longer propagation paths.

An offer of collaboration as been received from the Sodankylä Geophysical Observatory in
Finland who has a more sophisticated multiple element ionospheric absorption model. The
location of the oblique sounder with the riometer in Northern Scandinavia meant that the future
analysis of the existing data set could be extended to use EISCAT in-coherent radar
measurements of the ionospheric profiles. EISCAT provides a much more extensive set of
measured parameters than a vertical ionosonde. These could be used in place of the electron
density and temperature as model inputs. In principle EISCAT is also able to make accurate
measurements of the D layer of the ionosphere, that is primarily responsible for the HF
absorption, which is difficult to accurately observe with ionosondes. EISCAT is also able to
determine ionospheric profiles during high absorption events, which is when the HF ionosondes
cease to work.

Simplifications and comparison using an analytic ray trace that would speed the analysis with the
ultimate goal of producing a quick web facility. This could be incorporated with pre-existing real
time and ionospheric forecasting facilities.

The data between Svalbard and Tromsø, and between the UK and Tromsø, shows some
interesting disturbances that warrant further study. These could be combined with satellite as well
as ground based observations. The oblqiue sounding database offers a large potential for many
other inter-instrument analysis than those just outlined with EISCAT. The oblique sounding
measurements spaned the same region as many other ionospheric and geophysical instruments




                                                55
D42-1                        O BLIQUE IONOSPHERIC SOUNDING                        F INAL R EPORT


such as CUTLASS HF coherent radar, IMAGE magnetometer chain, all sky cameras, DERA
DAMSON HF Dopper studies and the ionospheric tomography receivers, amongst others.

Such an analysis offers the potential to better determine cause and effect for the HF channel.
This could then add to the growing awarness of the consequences of Space Weather on
communications and navigation systems, which when combined with modelling and computing
facilities will hoefully lead to better warning and protection schemes.




                                              56
D42-1                          O BLIQUE IONOSPHERIC SOUNDING                            F INAL R EPORT



                        LIST OF REFERENCES

1. Arthur, P.C., Lissimore, M., Cannon, P.S., and Davies, N.C. 1997 "Application of a high
   quality ionosonde to ionospheric research", Seventh Int. Conf. on HF RadioSystems and
   Techniques, IEE Conf. Pub., 441, pp. 135-139.
2. Bamford R.A. & Davis C. J., "UK radio experiments for the 1999 eclipse", COST 251, Side,
   Turkey, 7110, 28th March - 3rd April 1998.
3. Bamford R.A. and Levy M.F., "Results from the 1997-98 UK oblique sounding campaigns",
   Proceedings of PIERS (Progress in Electromagnetics Research Symposium), Nantes, France,
   pg 257, 13-17 July 1998.
4. Bamford R.A., Honary F., Tao H. and Västberg A., "First results from a combined Riometer
   and oblique sounder campaign in the auroral region", Proceedings of the National
   Conference on Antennas & Propagation (NCAP’99), Uni. York, UK, Pub. No. 461, pg 356 -
   358, 31st March - 1st April 1999.
5. Bamford, RA., "Radio and the 1999 UK Total Solar Eclipse", Final Report, Radio
   communications Agency, D48-1, May 2000.
6. Browne S., Hargreaves J. K. , and Honary B., Oct, 1995. "An imaging riometer for
   ionospheric studies". Electronics and Communication Engineering Journal, 7(5):209-217.
7. Budden K.G., "The propagation of radio waves: The theory of radio waves of low power in
   the ionosphere and magnetosphere," Cambridge Uni. Press, Cambridge, 1985.
8. Burns C.J. , Turunen E., Matveinen H., Ranta H., and Hargreaves J.K., "Chemical modelling
   of the quiet summer D- and E-regions using EISCAT electron density profiles", JATP, 53,
   115-134, 1991.
9. Chen F.F., "Introduction to plasma physics and controlled fusion", Second Ed. Vol.1 Plasma
   Physics, Plenum Press, New York and London, 1984.
10. Chen, J. and Bennett, J.A.: "Automatic fitting of quasi-parabolic segments to ionospheric
    profiles with application to ground range estimation for single station location" J. of
    Atmospheric and Terrestrial Physics, 52, (4), pp. 277-288, 1990.
11. Davis K., ‘Ionospheric Radio’, Peter Peregrinus Ltd. London, 1990.
12. del Pozo, C.F., Hargreaves J. K. and Aylward, A. D., "Ion composition and effective ion
    Recombination rate in the nighttime auroral lower ionosphere", JATP, 59, 1919-1943,1997.
13. Dyson, P.L., and Bennett, J.A.: "A model of the vertical distribution of the election
    concentration in the ionosphere and its application to oblique propagation studies", J. of
    Atmospheric and Terrestrial Physics, 50, (3), pp. 252-262., 1988.
14. Fenwick, R.B. and Barry, G.H., 1965, "Step by step to a linear frequency swee", Electronics,
    38, 66-7.
15. Haselgrove , J., "Ray theory and a new method of ray tracing", in Conference on the Physics
    of the ionosphere, pp. 355-364, London Phys. Soc. , London, 1954.
16. Jones R.M. and Stephson J.J, "A versatile three-dimensional ray tracing computer program
    for radio waves in the ionosphere", OT Rep. 75-76, U.S. Dep. Of Commerce Office of
    Telecommunications, 1975.
17. Krashenninikov, I.V., Jodogne, J.-C., And Alberca, L.F.: "Compatible analysis of vertical and
    oblique sounding data", Annali di Geofisica, 39, (4), pp. 763-768, 1996.



                                                 57
D42-1                         O BLIQUE IONOSPHERIC SOUNDING                         F INAL R EPORT


18. Levy M.F. and Bamford R.A., "First results from Essex-Devon oblique sounding campaign",
    COST 251, Linköping, Sweden, 6101, 8-11 October 1997.
19. Reinisch, B.W., "New techniques in ground based ionospheric sounding studies" Radio Sci., 21,
    p.331, 1986.
20. Titheridge,, J.E., "Ionogram analysis with the generalised program POLAN", U.S. Dept of
    Commerce, Nat. Oceanic and Atmo. Admin (NOAA), Report UAG-93, December 1985.
21. Västberg A., 1997, "Investigations of the Ionospheric HF Channel by Ray Tracing", IRF
    Scientific Report 241, April 1997, ISSN 0284-1703, Swedish Inititute of Space Physics,
    Uppsala, Sweden.
22. Wright, J. W. and M. L. V. Pitteway, "Data acquisition and analysis for research ionosondes",
    in "Computer aided processing of ionograms and ionosonde records", Report UAG-105,
    Word Data Centre A, Boulder, Colorado, pp 1-11, 1998.




                                               58
D42-1                               O BLIQUE IONOSPHERIC SOUNDING              F INAL R EPORT



                    APPENDIX A: LIST OF DATA
Key:
/Transmitter location/Receiver location/   year/month
Day

======
1997
======

/Hartland/Great_Baddow/1997/08
130897 150897 170897 190897 210897 230897 250897 270897 290897 310897
140897 160897 180897 200897 220897 240897 260897 280897 300897

/Hartland/Great_Baddow/1997/09
010997 020997 030997 040997 050997 070997
080997 090997 100997 170997 230997 240997

/Chilton/Svalbard/1997/09 /Chilton/Great_Baddow/1997/09
250997 260997 270997 280997 290997 300997

/Chilton/Svalbard/1997/10 /Chilton/Great_Baddow/1997/10
011097 041097 071097 101097 131097 161097 191097 221097 251097 281097 311097
021097 051097 081097 111097 141097 171097 201097 231097 261097 291097
031097 061097 091097 121097 151097 181097 211097 241097 271097 301097

/Chilton/Svalbard/1997/11
011197 021197 031197 041197 051197 061197 071197 081197 091197 101197

/Lancaster/York/1997/12 /Farnborough/York/1997/12
091297 121297 151297 181297 211297 241297 271297 301297
101297 131297 161297 191297 221297 251297 281297 311297
111297 141297 171297 201297 231297 261297 291297


======
1998
======

/Lancaster/York /Farnborough/York/ /1998/01:
010198 070198 090198 190198 210198 230198 250198 270198 290198 310198
060198 080198 100198 200198 220198 240198 260198 280198 300198

/Lancaster/York /Farnborough/York/ /1998/02:
010298 030298 050298 180298 200298 220298 240298 260298 280298
020298 040298 060298 190298 210298 230298 250298 270298

/Lancaster/York /Farnborough/York/ /1998/03:
010398 040398 070398 100398 130398 170398 200398 230398 260398 290398
020398 050398 080398 110398 150398 180398 210398 240398 270398
030398 060398 090398 120398 160398 190398 220398 250398 280398

/Lancaster/York /Farnborough/York/ /1998/04:
010498 040498 070498 100498 130498 160498 190498 230498 260498 290498
020498 050498 080498 110498 140498 170498 200498 240498 270498
030498 060498 090498 120498 150498 180498 220498 250498 280498

/Unst/Farnborough/1998/04
010498 020498 030498 040498 050498 060498 070498

/Unst/Lancaster/1998/04
010498 020498 030498 040498 050498 060498 070498

/Lancaster/York/ /Farnborough/York/ 1998/05:
010598 040598 090598 120598 150598 180598 220598 260598 290598
020598 070598 100598 130598 160598 190598 230598 270598
030598 080598 110598 140598 170598 200598 240598 280598

/Lancaster/York /Farnborough/York/ /1998/06:
010698 040698 070698 100698 130698 160698 190698 220698 250698




                                                        59
D42-1                           O BLIQUE IONOSPHERIC SOUNDING             F INAL R EPORT


020698 050698 080698 110698 140698 170698 200698 230698
030698 060698 090698 120698 150698 180698 210698 240698

/Lancaster/York /Farnborough/York/ /1998/07
010798 060798 070798


/Karesuvanto/Tromso/ Svalbard/Tromso/ /Farnborough/Tromso/ 1998/08:
270898 280898 290898 300898 310898

/Karesuvanto/Tromso/1998/09:
010998      080998     150998        220998       290998
020998      090998     160998        230998       300998
030998      100998     170998        240998       301098
040998      110998     180998        250998
050998      120998     190998        260998
060998      130998     200998        270998
070998      140998     210998        280998

/Karesuvanto/Tromso/ Svalbard/Tromso/ /Farnborough/Tromso/ 1998/10:
011098 041098 071098 101098 131098 161098 191098 221098 251098 281098
021098 051098 081098 111098 141098 171098 201098 231098 261098 291098
031098 061098 091098 121098 151098 181098 211098 241098 271098 301098

/Karesuvanto/Tromso/ Svalbard/Tromso/ /Farnborough/Tromso/ 1998/11:
011198      081198    151198      221198     291198
021198      091198    161198      231198     301198
031198      101198    171198      241198
041198      111198    181198      251198
051198      121198    191198      261198
061198      131198    201198      271198
071198      141198    211198      281198

/Karesuvanto/Tromso/ Svalbard/Tromso/ /Farnborough/Tromso/ 1998/12:
011298 041298 071298 101298 131298 161298 191298 221298 251298
021298 051298 081298 111298 141298 171298 201298 231298 261298
031298 061298 091298 121298 151298 181298 211298 241298

======
1999
======

Karesuvanto/Tromso/ Svalbard/Tromso/ /Farnborough/Tromso/ /1999/01/:
040199, 050199, 060199, 070199, 080199, 090199, 100199, 110199, 120199,
130199, 140199, 150199, 160199, 170199, 180199, 190199, 200199, 210199,
220199, 230199, 240199, 250199, 260199, 270199, 280199, 290199, 300199,
310199,

/Karesuvanto/Tromso/ Svalbard/Tromso/ 1999/02/:
010299, 020299, 030299, 040299, 050299, 120299, 130299, 220299, 230299,
240299, 250299, 260299, 270299, 280299,

/Karesuvanto/Tromso/ Svalbard/Tromso/ 1999/03/:
010399, 020399, 030399, 040399, 050399, 080399, 090399, 100399, 110399,
120399, 130399, 150399, 160399, 170399, 180399, 190399, 200399, 210399,
220399, 230399, 240399, 250399, 260399, 270399, 280399, 290399, 300399,
310399,

/Karesuvanto/Tromso/1999/04/:
010499, 020499, 030499, 040499, 050499, 060499, 070499, 080499, 090499,
100499, 110499, 120499, 130499, 140499, 150499, 160499, 170499, 180499,
190499, 200499, 210499, 220499, 230499, 240499, 250499, 260499, 270499,
280499, 290499, 300499,

/Karesuvanto/Tromso/1999/05/:
010599, 020599, 030599, 040599, 050599, 060599, 070599, 080599, 090599,
100599, 110599, 120599, 130599, 140599, 150599, 160599, 170599, 180599,
190599, 200599, 210599, 220599, 230599, 240599, 250599, 260599, 270599,
280599, 290599, 300599, 310599,

/Karesuvanto/Tromso/1999/06/:
010699, 020699, 030699, 040699, 050699, 060699, 070699, 080699, 090699,
100699, 250699, 260699, 270699, 280699, 290699, 300699




                                                   60
D42-1                           O BLIQUE IONOSPHERIC SOUNDING             F INAL R EPORT


/Karesuvanto/Tromso/1999/07/:
010799, 020799, 030799, 040799, 050799, 060799, 070799, 080799, 090799,
100799, 110799, 120799, 130799, 140799, 150799, 160799, 170799, 180799,
190799, 200799, 210799, 220799, 230799, 240799, 250799, 260799, 270799,
280799, 290799, 300799, 310799

/Karesuvanto/Tromso/1999/08/:
010899, 020899, 030899, 040899, 050899, 060899, 070899, 080899, 090899,
100899, 110899, 120899, 130899, 140899, 150899, 160899, 170899, 180899,
190899, 200899, 210899, 220899, 240899, 250899, 260899, 270899, 280899,
290899, 300899, 310899,

/Karesuvanto/Tromso/1999/09/:
010999, 020999, 030999, 040999, 050999, 060999, 070999, 080999, 090999,
100999, 110999, 120999, 130999, 140999, 150999, 160999, 170999, 180999,
190999, 200999, 210999, 220999, 230999, 240999, 250999, 260999, 270999,
280999, 290999, 300999,

/Karesuvanto/Tromso/1999/10/:
011099, 021099, 031099, 041099, 051099, 081099, 091099, 101099, 111099,
121099, 131099, 141099, 151099, 161099, 171099, 181099, 191099, 201099,
211099, 221099, 231099, 241099, 251099, 261099,

/Karesuvanto/Tromso/1999/11/:
011199, 021199, 031199, 041199, 051199, 061199, 071199, 081199, 091199,
101199, 111199, 121199, 131199, 141199, 151199, 161199, 171199, 181199,
191199, 201199, 211199, 221199, 231199, 241199, 251199, 261199, 271199,


/Karesuvanto/Tromso/1999/12/:
011299, 021299, 031299, 041299, 051299, 061299, 081299, 091299, 101299,
111299, 121299, 131299, 141299, 151299, 161299, 171299, 181299, 191299,
201299, 211299, 221299, 241299, 251299, 261299, 271299, 281299, 291299,
301299, 311299,


=====
2000
=====

/Karesuvanto/Tromso/2000/01/:
010100, 020100, 030100, 040100, 050100, 060100, 070100, 080100, 090100,
100100, 110100, 120100, 130100, 140100, 150100, 160100, 170100, 180100,
200100, 210100, 220100, 230100, 240100, 250100, 260100, 270100, 280100,
290100

/Karesuvanto/Tromso/2000/02/:
010200, 020200, 030200, 040200, 050200, 070200, 080200, 090200, 100200,
110200, 120200, 130200, 140200, 150200, 160200, 170200, 180200, 190200,
200200, 210200, 220200, 280200, 290200

/Karesuvanto/Tromso/2000/03/:
010300, 020300, 030300, 040300, 050300, 060300, 070300, 080300, 090300,
100300, 110300, 120300, 130300, 140300, 150300, 160300, 170300, 180300,
190300, 200300, 210300, 220300, 230300, 240300, 250300, 260300, 270300,
280300, 290300, 300300, 310300

/Karesuvanto/Tromso/ /Angle/Tromso/ 2000/04/:
010400, 020400, 030400, 040400, 050400, 060400, 070400, 100400, 110400,
120400, 130400, 140400, 150400, 160400, 170400, 180400, 190400, 200400,
210400, 220400, 230400, 240400, 250400, 260400, 270400, 280400, 290400,
300400

/Karesuvanto/Tromso/ /Angle/Tromso/ 2000/05/:
010500, 020500, 030500, 040500, 050500, 060500, 070500, 080500, 090500,
100500, 110500, 120500, 130500, 140500, 150500, 160500, 170500, 180500,
190500, 200500, 210500, 220500, 240500, 250500, 260500, 290500, 300500,
310500

/Karesuvanto/Tromso/ /Angle/Tromso/ 2000/06/:
010600, 020600, 030600, 040600, 050600, 060600, 070600, 080600, 090600,
100600, 110600, 120600, 130600, 140600, 150600, 160600, 170600, 180600,




                                                   61
D42-1                           O BLIQUE IONOSPHERIC SOUNDING             F INAL R EPORT


190600, 200600, 210600, 220600, 230600, 240600, 250600, 260600, 270600,
280600, 290600, 300600

/Karesuvanto/Tromso/ /Angle/Tromso/ 2000/07/:
010700, 020700, 030700, 040700, 050700, 060700, 070700, 080700, 090700,
100700, 110700, 120700, 130700, 140700, 150700, 160700, 170700, 180700,
190700, 200700, 210700, 220700, 230700, 240700, 250700, 260700, 270700,
280700, 290700, 300700, 310700

/Karesuvanto/Tromso/ /Angle/Tromso/ 2000/08:
010800, 020800, 030800, 040800, 050800, 060800, 070800, 080800, 090800,
100800, 110800, 120800, 130800, 140800, 150800, 160800, 170800, 180800,
190800, 200800, 210800, 220800, 230800, 240800, 250800, 260800, 270800,
280800, 290800, 300800, 310800

/Karesuvanto/Tromso/ /Angle/Tromso/ 2000/09:
010900, 020900, 030900, 040900, 050900, 060900, 070900, 080900, 090900,
100900, 110900, 120900, 130900, 140900, 150900, 160900, 170900, 180900,
190900, 200900, 210900, 220900, 230900, 240900, 250900, 260900, 270900,
280900




                                                   62