problems with kompsat plots

[jameswilburlewis]

From Vassilis:

In sifting through KOMPSAT/SOSMAG plots I see some positions were not converted to proper coordinates. See below:

https://themis.ssl.berkeley.edu/summary.php?year=2022&month=08&day=06&hour=0024&sumType=kompsat&type=sosmag

versus two days later:

https://themis.ssl.berkeley.edu/summary.php?year=2022&month=08&day=08&hour=0024&sumType=kompsat&type=sosmag

There is a clear jump in position.

[nickssl]

SOSMAG has two data sets, real-time and calibrated. It turns out both sets have problems for some dates, and their team is currently trying to fix these problems.

The real time data had coordinates system problems (as in the example above, date 2022-08-06), and that problem was fixed in the calibrated data, so it is better to use the calibrated data only. However, the calibrated data has the problem that some days contain spurious NaN data entries interleaved with correct data with shifted timestamps, and data for these dates cannot be used. Also, the calibrated data contains gaps for many dates, because the ESA server has a script cutoff while receiving data from the SOSMAG team.

Both these problems will be resolved at some point. When the problems have been resolved, we will run a full reprocessing of the summary plots using the original logic: always use the calibrated data, except for days when no calibrated data exists.

[nickssl]

We are waiting for the server fix, so that the recalibrated data do not have any gaps.

For the time being, we use the real-time data.