[rabernat]

True, but in fact, the Google ERA5 public data suffers from the exact chunking problem described in the post: it's optimized for spatial queries, not timeseries queries. I just ran a benchmark, and it took me 20 minutes to pull a timeseries of a single variable at a single point!

This highlights the needs for timeseries-optimized chunking if that is your anticipated usage pattern.

[bitschubser_]

a good source for ERA5 historical data is https://open-meteo.com/en/docs/historical-weather-api (not affiliated, just a happy user) you can also run open-meteo locally, its quite fast for spatial and timeseries queries

[open-meteo]

Creator of Open-Meteo here. There is small tutorial to setup ERA5 locally: https://github.com/open-meteo/open-data/tree/main/tutorial_d...

Under the hood Open-Meteo is using a custom file format with time-series chunking and specialised compression for low-frequency weather data. General purpose time-series databases do not even get close to this setup.