Add script to get SNOs for two platforms#51
Conversation
…ager satellite Signed-off-by: Adam Dybbroe <Adam.Dybbroe@smhi.se>
Signed-off-by: Adam Dybbroe <Adam.Dybbroe@smhi.se>
Signed-off-by: Adam Dybbroe <Adam.Dybbroe@smhi.se>
|
I have added the SNO script, it works on my laptop, I have a bit of unittests (though they are kind of system-tests) and some documentation. I was considering if we should/could merge it and then I could continue doing refactoring and making the tests more like real unittests, etc? |
Signed-off-by: Adam Dybbroe <Adam.Dybbroe@smhi.se>
pnuu
left a comment
There was a problem hiding this comment.
Few comments, and one change request (the addition of main()). Otherwise looks good to me.
bin/get_snos.py
Outdated
| return args | ||
|
|
||
|
|
||
| if __name__ == "__main__": |
There was a problem hiding this comment.
Encapsulate all this in main() function, and in if __name__ == "__main__": block call that. Can save some time later on when hunting a bug that is caused by polluted namespace.
| TLE_SATNAME = {'Suomi-NPP': 'SUOMI NPP', | ||
| 'NOAA-20': 'NOAA 20', | ||
| 'EOS-Aqua': 'AQUA', | ||
| 'Metop-B': 'METOP-B', | ||
| 'NOAA-19': 'NOAA 19', | ||
| 'NOAA-18': 'NOAA 18'} |
There was a problem hiding this comment.
How about finding the platforms by satellite ID number? The platform names might not always be in the TLE data, but the ID is.
There was a problem hiding this comment.
@pnuu Does pyorbital know how to do that, thought it required a name?
Signed-off-by: Adam Dybbroe <Adam.Dybbroe@smhi.se>
…ml template Signed-off-by: Adam Dybbroe <Adam.Dybbroe@smhi.se>
Signed-off-by: Adam Dybbroe <Adam.Dybbroe@smhi.se>
Signed-off-by: Adam Dybbroe <Adam.Dybbroe@smhi.se>
|
@adybbroe where are we at with this PR ? |
|
@adybbroe I tested this with one day's TLE file which contains CALIPSO (29108) and Sentinel-5P (42969). TLE filetle-20200102.txtsnos.yamlTerminal runAs the intersect time is around 2020-01-02 13:46, I downloaded two data files: However, I plotted the Sentinel-5P swath and CALIPSO lines: Datetime in both datasets:CALIPSO: Sentinel-5P: 13:46 is in the time period of the CALIPSO data, but not for the Sentinel-5P data. DebugI tried to output some variables for debugging: Here's the (coord_start, coord_end) in the It seems the first one (ref_pltfrm: CALIPSO) runs so fast. Could that be the problem? |
|
We hade som problems that this script is not accurate enough as the approximation of the path is not exact enought and done with only one arc. I made a more accurate version with several smaller arcs (see below). It takes more time but better finds the SNOs. |
|
Hi! |
|
@zxdawn I am doing good progress, but not yet ready to commit. Think that will be done in the morning (Thursday) CET. |
|
@adybbroe Nice to hear you're working on that! I don't have data at my hand now. Maybe you can test the CALIPSO and TROPOMI cases above? If it makes sense, then I can try to fetch the data later and find the exact location. |

...typically Calipso and an Imager satellite like Metop/NOAA
Signed-off-by: Adam Dybbroe Adam.Dybbroe@smhi.se
flake8 pyorbital