Back to Bulletins index Back to INAG-61 Contents page


INAG MEETING: BOULDER, COLORADO 10 JULY 1995


The INAG (Ionosonde Network Advisory Group) meeting was held on July 10 and ran from 7:30 PM until around 10:45 PM. It was a very successful meeting attended by 17 people all of whom had some interest in ionosonde data.

Attendees

Terence BullettPhillips Lab, Geophysics, 29 Randolph Road, Hanscom AFB, MA 01731, USA
bullett@plh.af.mil
Interest: Phillips Lab ionosondes
Ray ConkrightNGDC, 325 Broadway, E/GC2, Boulder, CO 8303, USA
rconkright@ngdc.noaa.gov
interest: NGDC, ionograms and ionosondes
Ivan GalkinUMLCAR, 450 Aiken Street, Lowell, MA 01854, USA
reinisch@caedc.uml.edu
interest : Digisondes
Per HoegSolar Terrestrial Physics Division, Danish Meteorological Institute, Lyngbgvej 100, DK-2100 Copenhagen, DENMARK
hoeg@dmi.min.dk
Interest: Global circulation patterns, Greenland
Vinita Ruth HobsonNGDC, 325 Broadway, E/GC2, Boulder, CO 8303, USA
rhobson@ngdc.noaa.gov
Interest: NGDC
Kiyoshi IgarashiCommunications Research Laboratory, 4-2-1 Nukui-kita, Koganei-shi, Tokyo 184, JAPAN
igararshi@crl.go.jp
Interest: 1 Terabyte Japanese data, Japanese ionosondes
Karen Fay O'LoughlinNGDC, 325 Broadway, E/GC2, Boulder, CO 8303, USA
kfo@ngdc.noaa.gov
NGDC, Boulder Ionosonde
Murray L. ParkinsonSchool of Physics, La Trobe University, Bundoora, Victoria 3083, AUSTRALIA
tspmlp@lure.latrobe.edu.au
Interest: Beveridge Digisonde 256, Casey DPS
Adolf K. Paul7455 Brockway. Boulder, CO 80303, USA
akp@ngdc.noaa.gov
Interest: San Diego Ionosonde
Gerd W. ProlssInstitut fur Astrophysik, Univ Bonn, Auf dem Hugel 71, 53121 Bonn, GERMANY
Interest: research
Edward ShiffmacherNOAA/NGDC (retired), 2155 Emerald Rd, Boulder, CO 80304, USA
Interest:. Ionosonde net
George TalarskiNGDC, 325 Broadway, E/GC2, Boulder, CO 8303, USA
gtalarski@ngdc.noaa.gov
Interest: NGDC, Processing and quality control
Lucia Villanueva
Phil WilliamsPhysics Department, University of Wales, Aberystwyth, WALES
pjw@aber.ac.uk
Interest: Swedish Network
Phil Wilkinson (Chair)IPS Radio and Space Services, Ionosonde network, (Paul Alekna, coordinator)
phil@ips.gov.au
ionosondes, Australian ionosonde network, space weather
Bill WrightSvavelgatan 6, S-59338 Vastervik, SWEDEN
wright@plasma.kth.se
Interest: Dynasonde
Dr. Alexander ZaitzevIZMIRAN, Troitsk, Moscow Region, 142 092 RUSSIA
zaitzev@charley.izmiran.rssi.ru
Interest: IZMIRAN ground-based geomagnetometers

Ionosonde Networks

Status of the US Air Force DISS network:

All DISS (Digital Ionospheric Sounding System) are undergoing a process called baselineing, in order to obtain a common hardware and software platform (Replacing 9 track tape drives with QICs and using uniform software). Changes to the DISS include a re-worked Final Power Amplifier, a more reliable and higher capacity UPS, and a new computer for the ARTIST, which also includes a 150 MB QIC tape cartridge to replace the expensive and unreliable 9-track tape drive. One downside to the new tape drive is that fewer people can handle the new media, but the advantages were sufficiently great that this still seems like a good idea. The status of the sounders in the DISS net are as follows:

Baselined and Operational:

Vandenburg Air Force Base (AFB), California, USA

Dyess AFB, Texas, USA

College, Alaska, USA

Wallops Island, Virginia, USA

Eglin AFB, Florida, USA

Ramey, PR

Bermuda

Goose Bay, Newfoundland, Canada

Learmonth, Australia

Baselined and non-operational:

San Vito, Italy

Argentia, Newfoundland

King Salmon; Alaska, USA (awaiting operator)

Non-baselined and operational:

Qaanaaq, Greenland

Sondrestrom, Greenland

Narsarsuaq, Greenland

Awaiting Installation:

Fairford, UK

Osan, Korea

Hawaii

South Pacific (possibly Wake Island)

Russian Network

IZMIRAN currently operates 15 Topside sounders and a network of vertical sounders including two at Salekhard, vintage 1957 and 1963.

EISCAT Ionosonde Network

Phil Williams supplied a list of the proposed north/south chain of digital ionosondes in association with EISCAT and EISCAT Svalbard Radar.

NyDigisondeExists at present (long term uncertain)
Longyear-byenDynasonde Under active discussion
BjomoyaDynasonde (or Digisonde) Under preliminary discussion
TromsoDynasonde and Digisonde Exist
KirunaSwedish Digital Ionosonde Exists
LyckseleSwedish Digital Ionosonde Exists
UppsalaSwedish Digital Ionosonde Exists

Faraday Station now Vernadsky Station

John Dudeney updated the present status of Faraday Station (65S 64W), in the Argentine Islands. The station was formally transferred from Great Britain to the Republic of the Ukraine in February 1996, together with all the scientific equipment. Under a Memorandum of Understanding, the routine geophysical monitoring program will continue. Data will be available both from the Ukraine and the World Data Centres. The observatory has been renamed Vernadsky Station.

Australian Network

The IPS Network is running a continuous five minute sounding program where possible. At some sites, such as Norfolk Island, the RFI from by the 4B ionosonde may possibly cause inconvenience to local people, so a 15 minute sounding program is used. Port Moresby was reopened in Nov 1994. All ionosondes in the Australian network now collect digital ionograms. These are scaled using UNIX based software developed at IPS. However, for ionogram exchange purposes, ionograms can be supplied in a format that is compatible with Titheridge's DIGION program. This ionogram viewer with built in scaling tools, and movie viewer for well sampled sequences, is most convenient for this purpose. The Australian ionosonde network will be significantly increased in size when several ionosondes commence regular operation in Northern Australia in support of the Australian Defence over-the-horizon radar.

New Zealand - Scott Base is now fully operational again with a digion and computers partly funded by the US Navy and IPS. Christchurch is currently installing a radio link so near real time data will be retrieved at least every six hours. All data from both Australia and New Zealand are manually scaled, automatic scaling not being considered trustworthy enough for archive purposes. Data are scaled automatically for real time purposes.

Meetings of Interest

URSI General Assembly meeting in Lille, France. Papers have been requested concerning computer scaling of digital ionograms and Quality control. (See later note on this). If possible, a UAG report containing the papers from meeting will be published.

Note: copies are still available of UAG-104, Ionosonde Networks and Stations. If you don't have a copy contact either myself (Phil Wilkinson) or Ray Conkright (NGDC, WDC-A for STP, 325 Broadway, Boulder, Co 80303, USA. Some copies are still available.

Codes for Ionospheric Data:

All present agreed that there is a need for new codes and that probably some of the old codes are not useful.

Some new codes that have been introduced

IPS added some local codes about 25 years ago for scaling spread F, although then, as now, there was no serious intent to distribute these data widely. These are: Frequency Spread (code = 55), and Range Spread (56). The code numbers were chosen according to the URSI convention.

The URSI IIWG (Ionospheric Informatics Working Group), has been introducing new parameters into the computer codes. INAG has been advised of some of these, but until recently was unaware of all of them. None have been recognised as official URSI codes . It is expected this will be discussed at the URSI INAG Meeting at Lille, 1997. A list of IIWG and ARTIST parameters with exchange codes is given later in this Bulletin.

Revising the current codes.

The possibility of deleting old codes for use with new parameters was discussed. Do we want to make room in the 10x10 matrix of scaled ionospheric parameters for new parameters or old parameters derived by new or different techniques? There may be further problems if the new derived parameter invalidates the original parameter.

There was much discussion on the issue of hmF2, it's storage and the method by which it was obtained. Terry Bullett offered some good comments on this.

"First, I believe that hmF2 should be a part of the distributed parameters. Our internal disputes as to the accuracy of one hmF2 method versus another, while valid, are secondary in importance to the issue of the use of these data by the scientific community. In the absence of ionosonde-derived hmF2 values, many scientists will use a model, and the differences between model and 'actual' hmF2 are typically several times larger than the uncertainty in the data. Also, being able to provide hmF2 will increase the overall value of ionosonde data, and we will all benefit from this.

Second, relating to our disagreements as to one hmF2 method or another, it may be possible to have an additional parameter which 'qualifies' or indicates the methodology used to determine hmF2. This parameter could also allow for 'version numbers' of methods, such as ARTIST, which change with time.

Third, to maintain scientific integrity, I believe we should make an attempt at putting 'error bars' on our hmF2 data due to the variety of methods used. Now I am personally biased in this realm because I work with the ARTIST method and have seen several comparisons with incoherent scatter radar. Anyway, if we were to make a list of the hmF2 methods in use, and use each on an identical set of data (not too large or small) we could use the variability in the results to get some idea of the uncertainty due to the different methods."

This is probably a typical example of some of the problems likely to be encountered changing the present codes. If hmF2 has been obtained by non-standard methods, should a new, more accurate method get its own code? Is part of this issue caused by INAG opinions and should INAG's internal disputes, such as to the accuracy of one hmF2 method versus another, be more important than providing fairly accurate measurements to model users? To some extent, this is an issue of standards versus opinion and is one of the reasons why changes are not made quickly. There should be plenty of time for others to participate in the debate.

While it will be possible to remove some parameters it may be hard to re-use them in the near future. Current usage of parameters based on the NGDC CD-ROM data may give some feel for the past use of parameters and has been used as an indicator in the article later in this Bulletin.

Another suggestion floated at the meeting was to create a pool of codes that could be used freely for experimental purposes. This would allow people to use codes locally, demonstrate both their validity and utility, before proposing a specific code be set aside by URSI for the parameter under consideration.

There is a problem obtaining programs for putting data into exchange formats. The IIWG data format, while compact, is difficult to program. A well documented version should be made available through NGDC.

After the meeting I spoke with Dave Anderson, Chair of IIWG, about the decisions on data formats. Informally, he is happy for INAG to take over responsibility for this work. It is an issue that will be resolved at the Lille General Assembly.

Ionogram Data Formats

The meeting felt that ionogram formats are not freely distributed, and are often considered proprietary information by their creators, for commercial reasons. Why can't we make formats public domain? Is it really the format that is proprietary, or is it the source code that would read this format?

Phil Wilkinson pointed out that while IPS has access to the format of many of the digital ionograms available, IPS has obtained the formats in commercial-in-confidence circumstances. IPS feels that ionogram formats should be freely available and where the format is not freely available, the community should be aware of it. However, in some circumstances, the format may be difficult to release. The IPS raw ionogram format for series 5 ionograms is relatively easy to supply while the cleaned format is not.

Ionograms come in many formats and some data users cannot read some formats. Discussion was held on what an appropriate data exchange format would be. Ideally the rawest form should be available. Formats should be universally available. With this goal in mind, the INAG Home Page (see later) will contain descriptions of formats, programs to read different formats, and the source code when available. Spaces will be left for formats that do not have descriptions or read programs. Where possible, the programmer's name will be attached when available and consent given.

Terry Bullett made several good points on this. He felt there were several difficulties with the ideas, particularly when it comes to source code.

Multiple versions, user support, negative criticism, unlimited distribution are all valid issues. It is also not trivial to create a subroutine to read ionogram data that is generally useful to a scientific user. Even the routines he has, which he feels are very good, well written and fairly general, would require some work to make distributable.

He thought the key to getting this kind of capability is to attempt to address the concerns of the people who are concerned with the distribution of data formats. We should have some 'rules of the road' for the use of these programs, data and formats. Maybe a copyright that would allow reading of data in a given format but not the writing of or conversion to that format without permission of the copyright owner? Is there an unethical use for someone else's data format? What are the concerns of the major players and who are they?

After discussion, all agreed there is a definite need for this information to be more easily obtained and there needs to be much more discussion on the topic. Meanwhile, formats for all the various ionosondes will be sought by those present at the meeting.

INAG Home Page

IPS will set up an INAG Home Page.

The INAG Home Page will contain the following: description of INAG affairs, links to other locations of interest, ionogram format descriptions, and programs for reading ionograms.

Data Availability

"With all these new digital sounders around the world, why can't I get any data from when and where I want it?" - the cry of the modern-day ionospheric scientist.

It is difficult to obtain real-time ionospheric data. Currently, worldwide, 80 stations report data of which 40 are real-time. It was agreed that if a screening program is used, incoming computer-scaled data can be filtered to obtain reliable estimates of ionospheric conditions.

The digital ionosonde has held great promise for promoting ionospheric science, but much of this promise has been unfulfilled, due to lack of data or poor data quality. Terry pointed out that these may be the side effects of a paradigm shift in the world of ionosondes from the manual, film based world to the automatic, digital world. One can lament the fact that few people look at ionograms anymore, but we cannot deny that things have changed. We cannot, for budget reasons and others as well, go back to the days of large numbers of accurately hand-scaled ionograms.

In addition, the issues of data formats, the sheer volume of digital data, the fact that it is generally unwieldy, and the frighteningly short technology cycle between birth and death of a recording medium has truly fragmented our ionosonde database. To attempt to recover from this, Terry is trying to get all the Digisonde data on CD-ROM, but this is far from a complete solution. Baselineing the DISS with a new computer and tape drive is another piece to this puzzle. There are still many issues here, and many must be dealt with to meet the realities of this new world of ionosondes in a new way.

Ray Conkright spoke about attempting to bridge the gap between real-time and archive data with near real-time data being supplied as provisional data.

Miscellaneous

Ray asked for ideas on how to display parameters beyond the usual stack and overlay plots. IDL programs will be used in future on-line NGDC Ionosphere data displays. The code for generating these plots will be shared. The present CD will be on-line soon.

So far, 65 CD-ROMs of ionospheric data have been distributed. During the IUGG and CEDAR meetings, in excess of 20 papers using ionosonde data were presented compared with none at the Vancouver IUGG, eight years ago.

Ray brought up the possibility of using Artificial Intelligence (AI) to improve automatic scaling statistics. Bill Wright suggested an auto-scaling program fed automatically into POLAN may be worthwhile, the autoscale program providing continuity. Ivan Galkin said that second generation ARTIST is already using AI.

Ruth pointed out that NGDC has changed station codes containing an explanation mark ( ! ) to an underscore ( _ ). Both are ASCII, but the ! is reserved as a special operator in some programming languages.

This was a particularly good Meeting thanks to the open and frank opinions expressed by all present. My special thanks to Ruth, Karen and Terry who jointly produced all the correct parts of these minutes - Phil.
Back to Bulletins index Back to INAG-61 Contents page