Add GEOS5 and MSG solar radiation to downloader

[Bich Tran]

It would be nice to add GEOS5 (tavg1_2d_rad_Nx: swgdn) and MSG (MDIDSSF: DSSF) for solar radiation input as MSG is used in WaPOR version 2 all levels and version 3 level 1, GEOS5 is used in version 3 NRT

[bert.coerver]

GEOS5 should be quite easy to add I think (need to look into it).

MSG i’m not sure, do you have any links to where to download it? What I seem to remember about this dataset is that its not available as a long term archive (i.e. only the last couple of days are available online, hence I never added it to pywapor).

‌

[Bich Tran]

I downloaded MSG data through LSA SAF data server: https://datalsasaf.lsasvcs.ipma.pt/PRODUCTS/MSG/MDIDSSF/ (the temporal coverage is from 2004, latency is 1 day)

They also provide access through OpenDAP: https://gitlab.com/helpdesk.landsaf/lsasaf_data_access/-/wikis/home

[Bich Tran]

I’m not 100% sure which MDSSF (MSG Downward Surface Shortwave Flux) product is used in WaPOR pipeline, because there are 2 operational products: MDSSFTD (15min) and MDIDSSF (daily): https://lsa-saf.eumetsat.int/en/data/products/radiation/

It’s mentioned that both instantaneous and daily solar radiation is needed in the WaPOR wiki: https://bitbucket.org/cioapps/wapor-et-look/wiki/Intermediate_Data_Components/SolarRadiation. So for instantaneous, I guess the 15min MDSSFTD product is used. However, it only provides 2 variables: Total and
Diffuse DSSF. It’s not super clear from the wiki how solar radiation is calculated from these two variables in WaPOR pipeline. I guess it’s something like this:

• transmissivity = diffuse DSSF/total DSSF.
• solar radiation = shortwave radiation at the top of the atmosphere (ra_flat) * transmissivity (the inverse of transmissivity function)

‌

[bert.coerver]

GEOS5 (tavg1_2d_rad_Nx: swgdn) is actually already implemented, just need to double check the units and add something to rename the variable to “ra”. This can also be done through the config file.

[bert.coerver]

✅ Added LSASAF Solar Radiation download module (https://bitbucket.org/cioapps/pywapor/commits/7a3b0fe4e4e86d61254f329d1ca20ee36fd96958)

✅ Added GEOS5 Solar Radiation variable (https://bitbucket.org/cioapps/pywapor/commits/242b056a246cf0f99d18031d024ffc2b1c1331b5)

Still to do:

• Rename ra variable to ra_flat to emphasise that it refers to a flat surface.
• Adjust et_look.py to correctly apply methodology of Allen et al (2006) to convert ra_flat to ra if required inputs are given (i.e. slope and aspect). And if not given, use ra == ra_flat.
• Add some more variables to the LSASAF module (e.g. lst, r0, ndvi)
• Add code to setup LSASAF account (un/pw)
• Update data-table in documentation.

Comparison between different Solar Radiation prods:

‌

[bert.coerver]

✅ Rename ra variable to ra_flat to emphasise that it refers to a flat surface.

✅ Adjust et_look.py to correctly apply methodology of Allen et al (2006) to convert ra_flatto ra if required inputs are given (i.e. slope and aspect). And if not given, use ra == ra_flat.

✅Add some more variables to the LSASAF module (e.g. lst, r0, ndvi)

(https://bitbucket.org/cioapps/pywapor/commits/7ae1d8732eaa225587dfe8079b8ffea49d62d089)

✅ Add code to setup LSASAF account (un/pw)

✅Update data-table in documentation.

(https://bitbucket.org/cioapps/pywapor/commits/01475b5d523c9b164ce315a7ccf8180fe00bf1f7)

Comparison between ra_24 and ra_flat_24
et_look will now apply Allen et al. 2006 if aspect and slope are found in the et_look_in.nc file.

‌