You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
While checking your data and building our benchmarking dataset, I found the problem with gs://weatherbench2/datasets/hres_t0/
It seems that all values for total_precipitation_6hr parameter are zero.
To evaluate IFS forecasts, we use the IFS analysis as the ground truth. Note that here we use the initial conditions of the HRES forecasts, i.e. the forecasts at lead time zero as analysis. This is not exactly the same as the analysis dataset provided by ECMWF (see paper for details).
Accumulated parameters at lead time zero are always zero.
I guess, the only way to fix this is to get this parameter from the IFS analysis dataset, which has the 6-hourly data.
The text was updated successfully, but these errors were encountered:
Greetings to the weatherbench2 team!
While checking your data and building our benchmarking dataset, I found the problem with gs://weatherbench2/datasets/hres_t0/
It seems that all values for total_precipitation_6hr parameter are zero.
This is explainable, quote from the docs:
Accumulated parameters at lead time zero are always zero.
I guess, the only way to fix this is to get this parameter from the IFS analysis dataset, which has the 6-hourly data.
The text was updated successfully, but these errors were encountered: