Remaining Data Sources #10
Description
This is an umbrella issue to keep track of all the data sources we are still missing.
The following data sources are currently not supported yet in the data-loader but are listed in Table 1 of the Overleaf document:
- GOES-16 Outgoing Longwave Radiation (OLR) data.
- OLR data is not a readily available output band from the public GOES-16 data. Lilli from our research group has access to a Fortran script developed by researchers NOAA which computes the OLR from the GOES-16 channels. Doing this computation isn't too expensive (OLR is computed is a simple linear transformation of 3-4 GOES-16 channels) but at the end of the day, this is a dataset we need to create ourselves. Once we have created the dataset we could try to make it accessible through the CEDA archive (Bryan Lawrence might be able to help), then we don't personally have to deal with issues of hosting the dataset.
- DYAMOND OLR and Precipitation data.
- The DYAMOND data is currently not directly publicly accessible. The official guidance on the website seems to be to email the ESiWACE coordination team and request access. This requires gaining access to the DKRZ compute infrastructure. Even if someone gets access to the DYAMOND data, transferring the data to another server is going to be non-trivial. I think our best bet will be to ask some of the DYAMOND members and be precise in what subsets of the data we want for our benchmark and ask them whether they would be open to making it accessible through some public API (e.g. host the data on the CEDA Archive).
- CAMS Nitrogen Dioxide
- As far as I can tell the CAMS data needs to be accessed through the CDS API Client. This requires a CAMS account and API key which is a bit cumbersome.
- ESA Biomass
- Addressed in Add ESA biomass cci dataset #7 but might need some re-design of the
Datasetfirst Changing theDatasetClass #9
- Addressed in Add ESA biomass cci dataset #7 but might need some re-design of the