Currently, GCM timeseries with lots of years, ssps, and runs are painfully slow to download because each layers has to be queried separately and there can be > 100,000 raster layers. I way to improve this could be to store the gcm_ts values in arrays in a table column, linked to a single reference raster.
cmahony
commented
6 months ago
Currently, GCM timeseries with lots of years, ssps, and runs are painfully slow to download because each layers has to be queried separately and there can be > 100,000 raster layers. I way to improve this could be to store the gcm_ts values in arrays in a table column, linked to a single reference raster.