Back to Bulletins index Back to INAG-62 Contents page

OBLIQUE SOUNDING IN AUSTRALIA

K.J.W. Lynn
Ionospheric Systems Research, 34 Gallery Rd, Highbury, Australia SA5089
email:
kenlynn@senet.com.au

Introduction

While oblique ionosondes have operated for a long time, scientific study and routine monitoring of the ionosphere has mainly been conducted with vertical ionosondes.

Improvements in the personal computer, the advent of GPS timing, stable synthesizers capable of rapid frequency changes and stable frequency sources has resulted in a new breed of flexible ionosonde where the format of the sounding is under computer control and both pulse and frequency-swept (FMCW) operation is available. The use of oblique sounding has become an option to an extent not previously seen. This article explores some of the characteristics and applications of oblique ionospheric sounding now made available as illustrated by oblique sounding in Australia.

Equipment

Results discussed here were initially derived from a type of oblique ionosonde built within the Australian Defence Science and Technology Organisation (DSTO) for research purposes. This equipment has been effectively deployed since the early 1990s (Clarke etal, 1993). These ionosondes trade a comparatively long duration frequency sweep for low power of operation (<50 W) and were originally developed because of the high cost and limited flexibility of the commercial oblique FMCW (chirp) sounders then available. However these ionosondes use the standard commercially available waveform and can operate in conjunction with such equipment. FMCW rather than pulsed operation is preferred for long-range oblique operation because of the low power required and high sensitivity obtained. Conventional pulsed oblique ionosondes of a past generation typically operated with peak powers of 40 KW or more with all the problems and costs such high powers generated..

A frequency range from 2 MHz up to 64 MHz was needed to cope with the high maximum frequencies seen on long, one-hop low latitude paths at sunspot maxima. Frequency sweep rates from 100 kHz/s to 500 kHz/s (or higher) were made available and successively employed for reasons described later. Transmitters and receivers were built in separate racks and could be operated separately or together at the one location.

Rubidium frequency standards were used to eliminate relative oscillator drift and GPS timing successively added to allow absolute timing synchronisation of ionosondes many thousands of kilometres apart and thus the group propagation time delay over the path. Each receiver could be programmed to receive a number of transmitters with a variety of sweep rates and frequency ranges tailored to the path in question.

The Australian ionosonde manufacturer KEL Aerospace has developed a number of additional options to their commercially available IPS-71 ionosonde finally arriving at a system which can pulse vertically for conventional or fast Doppler ionograms while also able to both transmit and receive the conventional oblique wave-form for complete compatibility with the DSTO ionosondes. The IPS-71 is now operating in conjunction with existing DSTO oblique sounding equipment.

While results discussed here were obtained specifically with the above equipment they should be regarded as indicative of what can be obtained with modern ionosondes although specific results will depend on the quality of the equipment (eg receiver sensitivity and antennas)

Antennas

At a distance less than 1500 km, the normal crossed-delta antenna commonly used by existing vertical ionosondes has been found to be effective in oblique operation despite the poor gain at low elevation angles inherent in such a system. Figure 1 shows an oblique ionogram obtained over a path of 2105 km by an IPS-71 in conjunction with a DSTO transmitter operating on a routine basis at the high sweep rate of 500 kHz/s and 10-20w . Existing crossed delta antennas are used at both sites. Such operation is at the limit of scientifically useful reception over a 24hr period as indicated by the high noise background and presence of the horizontal line generated by internal ionosonde noise which becomes apparent at the limit of receiver sensitivity. The extent of Spread F, for example, becomes very difficult to recognise under such conditions.

Figure 1. Oblique ionogram, 2105 km path, 500kHz/s sweep rate

At ranges greater than 1500 km, low launch antennas are increasingly essential. The choice here is between directional and non-directional antennas which is best determined by the application. For omni-directional reception at long ranges, some form of vertical monopole antenna is required in a simple system. Broad band antennas capable of operating over a 4-64 MHz range are not available and recourse must be made to a suite of antennas. DSTO has used a three or four antenna combination of a low frequency and mid-frequency vertical monopole and one or two high frequency yagiis for frequencies above some 25 MHz. The ionosonde equipment (both transmit and receive) operates with an internal programmed switching unit which switches the transmitter to the desired antenna at different stages of the sweep.

The minimum range an oblique FMCW ionosonde has been routinely operated by DSTO over flat terrain has been 11 km between a vertical transmit delta and a horizontally polarised dipole with both aligned to minimise reception of the ground wave so as not to overload the receiver. This quasi-vertical system provides the equivalent of a vertical ionosonde. Such expedients are no longer required when using a single ionosonde which can handle both vertical and oblique operation.

Ionosonde Programming

The basic limitation in running a sounder with a program of vertical and oblique sounding/reception is one of time. At the original rate of 100 kHz/s, an oblique ionosonde sweep over a 2-30 MHz range takes nearly 5 minutes. This is barely acceptable in taking a "snap-shot" of the ionosphere. At a 15 minute sounding rate this would allow the monitoring of up to three oblique paths and another 4 vertical soundings just might be squeezed in. To cover a 4-64 MHz range at this rate would take a prohibitive 10 minutes/ionogram.

The solution is to go to a higher sweep rate. In this case the ionogram is completed in a shorter time but with some possible degradation in quality. Only experience with a specific ionosonde and antenna system will provide the optimum trade-off between ionogram quality and sweep rate. With the equipment discussed above, 250 kHz/s sweeps were found to provide acceptable quality if antennas were suitable to the path length (up to 3000 km). However to increase the number of paths monitored it was sometimes necessary to go to 500 kHz/sec (as shown in Figure 1 ) with the proviso that path lengths at such rates be preferably kept to less than 1500 km. At this fast rate, a 28 MHz range is completed in less than 1 minutes and a 60 MHz range in 2 minutes. Some additional time may be taken in processing the ionogram but this should be small compared to the sweep time.

Oblique Ionosonde Applications

All applications listed below require accurate measurement of group time delay over the propagation path except for (a).

(a) Propagation Prediction Testing. This was the original requirement met by the DSTO oblique sounding program (Clarke etal, 1993). Many codes are available today to provide propagation predictions for HF users. Such codes are usually based on a database of measurements of foF2 and M(3000) or related phenomena derived from vertical sounding. Such programs achieve their greatest accuracy at shorter ranges. However little validation has been carried out (or is indeed possible) over long ranges and in all geographic circumstance. For this purpose, oblique sounding is imperative.

Whereas the upper frequency limit for short-range propagation (<500 km) is set by the maximum density of the ionospheric profile, the maximum frequency available for the F region over an oblique path is sensitive to both ionospheric critical frequency and height and in consequence suffers a greater variation than is familiar to operators of vertical ionosondes. The absence of direct comparisons may leave engineers and other users with entirely unrealistic views as to the variability and predicability of long range reception.

(b) Direct Range Conversion. This is a technique long known (Davies,1965) but little used which becomes of particular value as a direct method of converting oblique ionograms to an equivalent ionogram for any chosen range. The author considers this simple technique as the key to the full realisation of the potential of networked oblique ionosondes (Lynn, 1995). It has the great virtue of using all the experimental data available in an ionogram while dealing with both height and critical frequency changes without resort to modelling assumptions inherent in prediction codes. A limitation is that it is only effective for the o-ray and has a small but defined error (largely removed by the k factor).

(c) Ionogram Inversion. Methods for converting a vertical ionogram into an electron density profile have been developed over the last 30-40 years and have reached a high degree of accuracy and acceptance with programs such as POLAN in common use. Methods for inverting oblique ionograms have received less attention and have yet to achieve widespread acceptance. In Australia, a program called OBLINV (Phanivong, etal, 1995 ) has recently been developed. This has proved capable of producing excellent results on test oblique ionograms for paths less than some 1500 km but has yet to be tested over a wide range of oblique ionograms.

An alternative inversion technique is to convert the oblique ionogram into an equivalent vertical ionogram and then process this ionogram using a vertical inversion code. Again not enough testing has been done to demonstrate a wide-ranging capability. Poole (1995) has suggested other methods and there are many possible variants.

Whatever technique is used runs into the problem of increasing time delay compression inherent in an oblique ionogram. A vertical ionogram which extends over 4 ms will be reduced at some range to less than 2 ms with the greatest compression at the lowest heights. The oblique ionosonde does not have the time delay resolution to enable these lower levels to be expanded to equivalent vertical values without a major degradation in measuremental accuracy. Beyond some 2000 km the E region will be cut-off altogether losing all hope of accurate direct reconstruction. At long ranges a hybrid system employing model values for the lower layers is required to obtain an electron density profile estimation.

On the positive side, the delay compression means that an oblique ionogram covers a greater effective height range than the usual vertical ionosonde setting which becomes of particular value during ionospheric storms or when other large increases in virtual height occur. Such events may lift the virtual F region trace beyond the normal range of a vertical ionosonde.

(d) Networking. The recent rapid improvement in data communications makes the networking of ionosondes an important step in providing ionosonde data to those who need it in real-time or near real-time. The constant battle to prove the ongoing relevance of ionosondes can be greatly assisted by developing techniques to meet the needs of a potentially new generation of users. In this regard, a network of ionosondes linked directly to a central location is an important step in the immediate distribution of raw or processed data to both civil and defence users of HF as well as for scientific study.

Such a system should be regarded as an ionospheric weather service with as much relevance to HF and satellite communications (as well as GPS and OTH Radar) as a tropospheric weather service to surface applications. (For a general overview of ionosonde networking and where it has currently been achieved see UAG-104, 1995 in References.) Digital data transmission over phone lines is continually decreasing in cost but the INTERNET may provide the ultimate system for the cheap collection and distribution of ionosonde data. Some initial examples are already on the World Wide Web and can be considered pioneers in this field..

The advantages that fielding combined vertical and oblique ionosondes can provide when networked are illustrated in Figures 2(a) and (b). Figure 2(a) shows a hypothetical distribution of eight vertical ionosondes deployed so as to provide an ionospheric weather service over the

Figure 2. (a) Coverage achieved with vertical sounders

Figure 2. (b) Coverage achieved with vertical/oblique sounders

continent of Australia. (If such a number seems large it should be noted that there are more than eight ionosondes operated in Australia by a variety of independent users but with as yet no major effort to rationalise and integrate them. One suspects that similar conditions exist in other countries.) The circles with a radius of 300 km drawn around the vertical ionosonde reflection points indicate the area of high F region correlation. It is evident that the continental area is very much undersampled by such a distribution if an accurate mapping of ionospheric conditions is desired.


Figure 3. Spatially interpolated maps of MUF normalised to a path length of 2000 km
(a) 0600 UT

Figure 2(b) shows the coverage achieved if these same eight ionosondes are able to operate obliquely as well as vertically. By allowing each ionosonde to receive and transmit to its closest three neighbours, the number of reflection points being monitored has increased to 24 and the continental area is now well covered.
A major objection to the use of oblique sounding is that the information contained cannot be readily be integrated with that obtained by a traditional vertical sounder. This is not so. The use of direct range conversion as mentioned in (d) can convert ionograms from a mixed distribution of vertical and oblique paths of differing length to a common effective range (as can ionogram inversion combined with ray-tracing, if such can be made reliable). Moreover this removes the last vestige of undersampling by allowing the development of an integrated picture within which spatial interpolation can be carried out to counter the effect of large-scale gradients across the area of coverage.

When used for real-time frequency management it is thus possible to convert all ionograms to the range of the communication link and spatially interpolate the picture to the reflection point of that circuit.

By way of illustration, consider the two pictures shown in Figure 3(a) and 3 (b). Here a group of oblique ionograms (and vertical ionograms from


Figure 3: (b) 1200 UT.

Darwin), representing sparsely sampled snapshots of the low latitude ionosphere, have been range-converted (in this case to 2000km) and combined by the Kriging spatial interpolation technique (Samardjiev etal 1993) to show the effective maximum frequency for a path of this length reflecting anywhere in the sampled region at the time in question. Figure 3 (a) shows the development of the equatorial anomaly whereas Figure 3(b) shows the major north-south gradient which occurs at a later time of day. The round spots within the picture show the apparent location of the oblique reflection points (oblique path mid-points) used in the interpolation.

It is obvious that such a mapping technique can have both scientific and practical application. In the former case, it can be used to compare the actual ionosphere with predictions generated by computer codes. If all sites are connected to a central computer then the system can be used to provide real-time frequency management of HF communications.

The definition of Sporadic E over a given area remains a more difficult problem. Measurements made by backscatter sounding in northern Australia indicate that when sporadic E is significant, the geographic area of coverage is often large (if variable in strength). This is one application where a backscatter sounder may provide the best solution (Houminer etal, 1996). Failing that, the highest possible density of reflection point coverage is needed and is most readily provided by networked oblique sounders.

In some parts of the world (notably Europe and some parts of the USA) the density of vertical ionosondes is such that an integrated system of vertical ionosondes alone would be sufficient to define the ionosphere over very large geographic areas if networking could be achieved. That would also appear to have been a possible goal arising from the PRIME project initiated by Peter Bradley. However on past performance, inter-country cooperation appears to be very difficult to achieve. For areas less well endowed, a network of combined vertical/oblique ionosondes will remain the most cost-effective solution. For ocean areas, the oblique sounding technique has no peer. Topside sounding provides an effective ionospheric slice along the satellite track as does tomography but neither can achieve a sufficiently high revisit rate without an increase in the number of relevant satellites.

References:

Clarke R.H., Fyfe D.F., Kettler D.I., Lynn K.J.W., Malcolm W.P., Sprey B.M., Taylor D.P. and Wright C.S., TENERP Conf. Proceedings, June 22-24, NPS, Monterey, CA., 1993.

Davies K., "Ionospheric Radio Propagation", NBS Monograph 80, 1965

Houminer Z., Russell C.J., Dyson P.L. & Bennett J.A., Ann.Geophys., V14, p1060, 1996.

Lynn K.J.W, World Data Centre Report UAG -104, p59, 1995.

Phanivong B, Chen, J., P.L. Dyson & Bennett J.A., J. Atmosp. Terr. Phys., V 57, p1715, 1995.

Poole A.W.V. & Mercer C.C., World Data Centre Report UAG -104, p59, 1995.

Samardjiev T, Bradley P.A., Cander L.R. & Dick M.I., PRIME COST 238 Workshop, p257, Graz, Austria, 1993.

Back to Bulletins index Back to INAG-62 Contents page