search

FPS-URB-RCC / CORDEX-CORE-WG

This repository is dedicated to the CORDEX-CORE data analysis WG
0 stars 0 forks source link

Land area fraction and orography in different grids #7

Closed jesusff closed 1 year ago

jesusff commented 1 year ago

Current orography and land-sea mask under data/ lie on different grids for REMO EAS-22:

ds = xr.open_dataset('data/REMO/orography/orog_EAS-22_MPI-M-MPI-ESM-LR_historical_r0i0p0_GERICS-REMO2015_v1_fx_r0i0p0.nc')
mask = xr.open_dataset('data/REMO/land-sea-mask/BLA_EAS-22.nc')

operands could not be broadcast together with shapes (255,417) (273,435)

The land-sea mask from ESGF is currently not available (seems a temporary problem) http://esgf1.dkrz.de/thredds/fileServer/cordex/cordex/output/EAS-22/GERICS/ECMWF-ERAINT/evaluation/r0i0p0/GERICS-REMO2015/v1/fx/sftlf/v20191030/sftlf_EAS-22_ECMWF-ERAINT_evaluation_r0i0p0_GERICS-REMO2015_v1_fx_r0i0p0.nc

jesusff commented 1 year ago

I added to the nextcloud folder the landsea mask from our ESGF replica at unican. It aligns now with the orography file.

However, the urban fraction is still on a different grid. It seems to be due to a relaxation zone of 9 gridpoints all around the domain. I'll test it on EAS-22

jesusff commented 1 year ago

Done. I added a sample file for EAS-22 and seems to align well (see static_fields.ipynb, also for code used to crop the original UF file)