4. INAG MEETING: CHRISTCHURCH, NEW ZEALAND, 31 DECEMBER 1993

4.1. Attendance:

Prof. Jack Baggaley

Dr Bob Bennett

Dr Harvey Cummack

Christer Juren

Dr John Titheridge

Lester Tomlinson

Dr Phil Wilkinson

Since Christer was visiting New Zealand and wanted to meet with some of the people in New Zealand and Australia who were concerned with ionospheric monitoring, a short meeting was held at the University of Canterbury. There was no agenda and people expressed views and outlined their expectations for their respective networks and stations. The meeting started at 0900 and ended at 1300.

4.2. Comments from Christer

Christer opened the meeting by briefly describing the Swedish Institute of Space Physics before giving a more detailed explanation of their present ionosonde network (see INAG-59 for more details). Computerising the network started roughly three years ago, when there were suggestions that the network may have to close. The system is based around several transputers. Currently the three ionosondes of the Swedish network (Kiruna in the north, Lycksele and Uppsala) can be accessed from the Internet, provided the codes are known. A later demonstration of this was frustrated by a couple of unhelpful computers placed in the network line somewhere. At each station the sampling speed is 1MHz and is achieved using five transputers. A further two transputers are used for GPS measurements so that there is absolute time synchronisation between the three sites. Phase and amplitude are measured at each site but ordinary and extraordinary separation has not yet been implemented, making automatic computer scaling difficult.

The upper level data access system is programmed in Smalltalk. Currently, it is set up in a PC environment and can also be accessed in an X-windows environment on UNIX. Manual scaling, using the computer, is also written in Smalltalk. Several other languages are used at different stages in the network: C and pascal for passing information to the transputers, occam for control in the transputers.

Data from the Kiruna station is accessed in real time to support the nearby rocket station at Esrange. These data, together with other sources, are used to select periods when rockets are fired. Data are also used to support HF systems as well as other scientific groups such as EISCAT. One interesting scientific purpose for collecting the data is to track gravity waves formed at high latitudes. It has been suggested that these processes may be important in global warming. Finally, the ionograms (images) and scaled data form the basis for a moderate sized database. This is a good prototype system capable of being expanded to include the collection of other solar terrestrial data sets.

4.3. Phil's Comments

IPS Radio and Space Services is a service based organisation within a service oriented Government Department. The main theme of IPS operations is to collect ionospheric data and return it, in one form or another, to customers. Although there are no definite customers for real time ionospheric data, near real time data is of interest and the IPS network will be able to supply both types of data on request in the near future. The 5A ionosonde development, described in an earlier INAG Bulletin has not been wholly successful. While the software development has succeeded and the system can be managed from a central location, scaled and raw data being stored in separate directory files, the hardware has been disappointing. This phase of the development is now under review. In the interim period, the DIGION board, developed by John Titheridge, at the University of Auckland, is being installed at all the IPS stations. When fully implemented, the system will include a DAT tape system to record five minute raw ionograms together with autoscaled data. These scaled data, together with raw hourly ionograms will be sent by telephone to the central computer.

4.4. John's comments

John gave a brief review of the growth of his department. He built up a chain of total electron content stations which has now been reduced to two stations at Auckland and Invercargill. He pointed out that TEC is the most accurate measurement available of the ionosphere being, potentially, capable of 0.1% error with good calibrations. This accuracy can be maintained objectively for periods as short as minutes to several decades.

The Auckland ionosonde station supports these TEC measurements as well as offering bottomside information directly. In 1987 a student started developing ways of recording data digitally from the IPS-42. The present DIGION board has been developed from this work. The IPS-42 stores data with 0.8km resolution in virtual height for 576 frequency channels. Three soundings are made per frequency channel and for an echo to be judged real it must appear in all three soundings. This information is passed back to the analogue display and is the information intercepted by the DIGION. While 0.8km resolution could be stored, in fact this is divided by two-thirds giving a final resolution of 1.6km which is adequate for inverting ionograms using polan. More details on the DIGION are given in an accompanying article in this issue of INAG.

Accompanying the DIGION is display software for scaling the ionograms. For research purposes, John pointed out there is not much value in a scaling system that is right 80% of the time and can be completely wrong the rest of the time. To improve scaling, there is considerable value in using digital ionograms with a scaling system designed to ease manual labour. He stressed how easy it is to compare ionograms digitally. In scaling foF2 for gravity wave studies, it is very easy to scale large numbers of ionograms quickly. His scaling system has several features to aid scaling. When heights are displayed, the primary height is accompanied by a series of height markers at 2*height and 3*height so multiples can be lined up and quickly recognised. For mid latitudes, by far the majority of returns come from overhead, as can be shown using the multiples. Frequencies are shown with an accompanying pair of ordinary and extraordinary cursors scaled mathematically correctly. This makes scaling easier as sometimes one or other trace is indistinct and using the pair of traces makes feature identification simpler. Currently, John is working on an improved manual input to polan based on the present DIGION. There is obviously much work yet to be done in developing user friendly, and effective, computer aided scaling systems for manually processing ionogram data.

John pointed out that there are some errors in the display of the IPS-42. The frequency marker display around the border of the ionogram is incremented one frequency channel high and the heights are in error by a constant 9km. This last point was discussed as the Christchurch ionosonde did not appear to have this error. All people using any data must be constantly aware of the possible errors inherent in their data due to the methods used to record it.

4.5. Comments from Jack

When New Zealand ceased to support an ionosonde network the Campbell Island (and many years earlier, the Rarotonga) ionosonde was closed and the Scott Base and Christchurch ionosondes were passed over to the University of Canterbury. Operation of the ionosondes has been patchy over the last few years and some of the data is still to be processed by IPS. However, the Scott Base ionosonde will now be converted, using a DIGION, to collect digital ionograms early in 1994. The hardware for this has been funded by the US Navy. Later this year, the Christchurch ionosonde will also be converted.

4.6. Summary

After discussions, the DIGION was demonstrated and an attempt was made to contact the Swedish network through Internet. Unfortunately, this was unsuccessful. However, aside from that minor setback, this was a useful and interesting half day meeting. It demonstrated very clearly the wide range of computer based techniques that are now being used with, in some cases, moderately old equipment. Provided funding is available, there is likely to be more progress in this area over the years to come.