CDO

Hi,

I'm trying to remap 0.1 x 0.1 degree global EDGAR fluxes to a WRF grid (about the same resolution: 10 x 10 km, Lambert projection). The domain covers the European continent. I'm using cdo remapcon to conserve tracer mass. For large parts of the domain the results look good. But unfortunately I get negative weights, which will result in unphysical loss of mass. There are also fewer weights greater 1.
I tried several combinations of settings environment variable NORMALIZE_OPT differently, setting 0 fluxes to missing value (cdo setmisstoc,0) and using remapcon2. But the problem persists. The problem doesn't occur when using other non-conserving operators (remapnn, remapbil, remapdis,). The same issue exists with remaplaf and remapbic (for the latter no negative weight warning, but negative fluxes).

Can you tell me if I'm doing something wrong? And if not, is there a best practice how to deal with the negative weights?

Thank you and regards,
Roberto