search

sunpy / radiospectra

Provide support for some type of radio spectra on solar physics
https://docs.sunpy.org/projects/radiospectra
BSD 2-Clause "Simplified" License
13 stars 21 forks source link

eCallisto archive frequency, time and other information #60

Open samaloney opened 3 years ago

samaloney commented 3 years ago

Description

Currently it's not clear what the duration of all the files returned by the eCallisto client so no end time is returned. most of the real eCalliso file seem to be 15min but not sure. Check with Christian Monstein to see if there any way to know from the file names alone.

nabobalis commented 3 years ago

Do we know the variation on average per file?

samaloney commented 3 years ago

From email correspondence

For e-Callisto spectrometers are the generated fits files always 15 minutes duration?

900 seconds is the default size in time. But observer can stop onservation any time. For e-Callisto spectrometers is there a way to know their frequency range without opening the fits file? There is an indication here: http://www.e-callisto.org/Callisto_DataStatus_Vwww.pdf But this just tells the startup-configuration whihc may change any time unpredictable. There is another indication here: http://soleil.i4ds.ch/solarradio/data/readme.txt Again no guarantee. I'm often NOT informed about changes. Finally, the onyly way is to read out frequency axis from FIT-file. This is the current and true value. For the other spectrometers in the archive e.g. Phoenix is there a way to tell the duration from the filename alone? No, you need to read time-axis or the header-information: DATE-OBS= '2021/11/01' / Date observation starts TIME-OBS= '01:15:01.840' / Time observation starts DATE-END= '2021/11/01' / date observation ends TIME-END= '01:30:01' / time observation ends Also be aware that time-resolution can be different. This is part of configuration (number of spectra per second 1, 2, 4, 8, ...)

I’ve seen the description file from here http://soleil.i4ds.ch/solarradio/data/readme.txt which could help with some of these but not all.

Any thoughts or suggestion welcome

It would be great if one could append different frequency ranges from different stations. E.g. low frequency from BIR and higher band from TRIEST. I know this is not straight forward, it requires several assumptions at which frequency to split the spectra. And often time needs adjustment when PC is out of sync (shift in time). It would also be convenient to offer different methods of background subtraction, e.g.:

  • a pixel-value
  • a single spectrum at time-slice T
  • average of the whole 2D-spectrum
  • Individual value per frequency
  • and there might be many more methods. We also miss kind of edit function to wipe out rfi manually. Finally kind of editor similar to our JavaViewer but more functions and more colour-tables.
nabobalis commented 3 years ago

Thanks for the context.