This repo includes three separate processes:
request.py retrieves into L2_data/ and reduces into L3_data/ the resolution of all .nc files covering a certain area (aoi, as .geojson) in a certain
timeframe (--date yyyyMMdd yyyyMMdd). With the -l/--links flag the program quits after printing the URLs. The -s/--skip flag starts reducing already
downloaded files without downloading missing files. Part of this code is courtesy of Bilel Omrani's s5p-analysis, under an MIT license found in this file.
After that, process.py takes in those files as argument and averages them over time, saving them in processed/.
analyse.py clips a processed file (raster) to a certain area's shape (shapefile, as .shp, see shp/ folder), averaging them over the surface and registering the result
and other information about the file in analysis.csv (no flag) or fnanalysis.csv (-s).
harp
numpy
pandas
sentinelsat
tqdm
geopandas
xarray
rioxarray
netCDF4
shapely