We provide datalibs for accessing pressure-level ERA5 and HRES data. We should implement similar interfaces to model-level ERA5 and HRES data (likely in two separate datalibs).
[ ] Consider caching retrieved GRIB files using a DiskCacheStore, particularly if this allows more efficient MARS requests.
[ ] Retrieve model-level data at the target horizontal resolution, and interpolate to target pressure levels using metview.
[ ] Consider including conveniences (parallel workers, cleanup of metview temporary files) similar to the ARCO ERA5 datalib provides access to model-level ERA5 data, but only through mid-2023 and not for ensemble members).
[ ] For ERA5, include interfaces to nominal analyses (example) and ensemble members (example).
[ ] For HRES, implement an interface to nominal forecasts (example), and confirm that individual ensemble members are not available in the operational archive.
Alternatives
The ARCO ERA5 datalib provides access to model-level ERA5 data, but only through mid-2023 and not for ensemble members.
Description
We provide datalibs for accessing pressure-level ERA5 and HRES data. We should implement similar interfaces to model-level ERA5 and HRES data (likely in two separate datalibs).
Recommended approach:
DiskCacheStore, particularly if this allows more efficient MARS requests.metview.metviewtemporary files) similar to the ARCO ERA5 datalib provides access to model-level ERA5 data, but only through mid-2023 and not for ensemble members).Alternatives