Frequently Asked Questions for the STARE Radar

1. What does STARE measure? STARE is a Doppler radar which detects the speed of meter-scale ionospheric irregularities in the ionospheric E-region. In the E-region these irregularities propagate approximately with the local ExB velocity, thus providing a possibility to indirectly measure the electric field (or, equivalently, the plasma flow pattern) in a two-dimensional grid of about 400 x 400 km size over northern Scandinavia with 20 second time resolution. The measurement principle is something similar as detecting the wind speed from the reflection of the moon from the surface of water.

2. Give me more technical information Some technical information is available here, but you can also look at the original paper by Greenwald et al. (Radio Science 13, p. 1021-1039, 1978).

3. Is it any good? Since STARE is dependent on the existence of meter-scale ionospheric irregularities, it is not a reliable workhorse instrument as, for instance, is a magnetometer (see the IMAGE magnetometer network).

4. What is the main strength of STARE? The main strength is that in suitable conditions, it produces a two-dimensional electric field distribution with relatively good (20 s) time resolution. No other instrument can do this.

5. What is the main weakness? The main drawback of STARE is that most of the time it sees no echoes at all, because the electric field is too weak to form irregularities or because there are too few electrons in the critical altitude range.

6. Can it be used for other purposes? Yes, one can study the physics of the irregularities themselves with it.

7. What are the angles of the radar beams? For Midtsandan, the beam angles are 13.6, 17.2, 20.8, 24.4, 28, 31.6, 35.2, 38.8 degrees from local geographic north towards east. For Hankasalmi they are -31.7, -28.1, -24.5, -20.9, -17.3, -13.7, -10.1, -6.5, where negative signs mean that the beams are pointing slightly westward (see map).

Getting and using data

1. What kinds of data files are there?

   The raw data files are binary files consisting of 20-second dumps. The size of a raw data file for each 24-hour UT-day is about 73 megabytes. The raw data files are not compressed.

   Analyzed data files are in the HDF format, which can be read by Tela and other programs. Analyzed files contain the echo intensity and the mean Doppler velocity (zeroth and first moments of the autocorrelation function, respectively) as 3D arrays. The dimensions are the temporal index, the beam number (1..8) and the range index (1..50).

2. How do I get the spectral width? Currently, you must use the raw data files to get the spectral width. You have to write your own routines to read and the files. This is not difficult, but it involves low-level programming.

3. How do I access the raw data files? The files are stored on CD-Rs (readable by a normal CD-ROM drive) at FMI/GEO. There is about one CD-R for each week for each radar. We can write these files to DAT tapes for you, for instance, on request, or make them available on FTP. We may have to charge for the medium.

4. How do I use the raw data files? You have to write your own program for doing it using the file format description.

5. How do I access the analyzed files? Once you are in our list of allowed WWW users, you can obtain these directly from the Web, from the same place where you get the RTI plots.

6. How do I use the analyzed files? These are HDF files which contain the following variables: entime <real[4320]> range <real[50]> snr <real[8,50,4320]> sttime <real[4320]> vel <real[8,50,4320]> The sttime,entime vectors define the start and end time for each 20 second dump. In practice one can just use sttime to get the time axis, or one can use the average of sttime and entime. These times are in seconds from the beginning of the UT day. The range vector gives the range in kilometers. It should be the same in every file. The snr array is the signal to noise ratio in dB for each beam, range and time. Similarly, vel is the Doppler velocity in meters per second. The sign of vel should be positive towards the radar. For some strange reason however, one has to change the sign for Norwegian data currently. (Equivalently, one could say that vel is positive away from the radar and for some strange reason one has to change the sign for the Finnish radar.) It is best to use common sense and one's geophysical intuition to check the sign of vel, however.

   The cleanest way to use these data files is to follow the instructions for using stare.t. Another way is to install Tela and simply load the HDF files directly to Tela. If one wants to use Matlab, one can use export_matlab from Tela. If one wants to use IDL, one can read the HDF file directly (Paul Eglities has done this).

7. What are those RTI plots? Range-Time-Intensity plots.

8. Where are the merged vector data? There are no such data. The merging is done by the stare.t, IHMETYS or GetSTAREdata program each time. If you don't want to or can't use one of these programs, you have to write your own merging code.

Interpreting data

1. What is the interpretation of the echo strength (SNR, S/N ratio)? The SNR is literally a product of many factors and thus very difficult to interpret. See RTI description for more.

2. Is the velocity measurement reliable? I would say that it is usually rather good. During very disturbed periods, however, the changes in the magnetic field may be large enough to cause distortion.

3. The vectors produced by stare.t must be wrong since when I plot them separately for each radar, the sum does not correspond to their vector sum, and at least approximately, it should (!) You are probably looking at a case where one of the radars sees only a small Doppler velocity. In such a case, the resulting vector must be almost orthogonal to the beam direction of the corresponding radar. If the beams were always orthogonal, the vector sum would be the correct answer but since they are not, in cases like this the vector sum gives an incorrect idea.