ERA5 format conversions

Summary

We want to use ERA5 data downloaded as GRIB files in stack_creation.py, so we convert to GEOTIFF. Since we are moving from 300mt to 10mt resolutions, we might encounter some memory limitations, to solve this issues it can be useful to have a NETCDF version of the files.

Description

1. propose a solution to convert GRIB to GEOTIFF in python.
2. propose a solution to convert GRIB to NetCDF in python. Since downloading in netCDF is currently marked as an experimental feature, we will install ecCodes in the docker image and use it to convert the GRIB files.
3. add a method to process data into daily accumulators, averages etc... taking into account multiple variables
4. to feed processed meteo data to the stack creator we have developed a script which splits the netcdf to one geotif per day

Note:

It might be needed to investigate further ERROR: Ran out of file reading SECT0 This does not appear to be a critical error, the output files seem well formatted. This error is reported to be generated in the low level GRIB loading library used by GDAL.