search

meteostat / meteostat-python

Access and analyze historical weather and climate data with Python.
https://dev.meteostat.net/python/
MIT License
415 stars 60 forks source link

Model data when no station nearby #88

Closed sramirez closed 2 years ago

sramirez commented 2 years ago

Hi!

Congratulations for this amazing repository. I think meteostat could be greatly improved if model data is provided by default if no nearby station is found for a given data point. This happens to me trying to find data in southern Spain as stations are really scarce in some areas there. As this scenario may appear frequently in certain areas, this new feature could be really cool to add.

clampr commented 2 years ago

That would certainly be nice.

However, we'd need to change the way Meteostat works to achieve that. Currently, model data is specifically stored and retrieved only for the locations where a weather station is installed.

Maybe we can set up a grid (e.g. 5 * 5 km) and provide model-based time series for each grid cell. Then, if there are no stations nearby or gaps in the series, we fall back to the closest grid cell.

Biggest challenge would certainly be the amount of data. If we want to have a reasonable resolution, we'll end up with millions of grid cells. I doubt this is feasible without bigger funding. But I agree it would improve data availability in remote areas (or those with few stations).

sramirez commented 2 years ago

Just curious, what is/are the source/s for model data?