Dear all,
I am running WRFV3.9.1 and I am trying to assimilate reflectivity observations using the EnKF. The main problem I have, is that in areas where it is not observed reflectivity values (clear air), the model produce a default value of -35 dBZ. This produces a problem when I try to assimilate obs, because the data assimilation algorithm try to correct this -35 dBz to 0 dBz, generating in this way values close to 0 dBz (but not exactly 0 dBz), so generating reflectivity values in regions where we should have 0 dBz. In other words, if the minimum default value of the model was 0 dBz, then during the assimilation, the model would remain the same and no spurious reflectivity would be generated.
On the figure attached you could observed what happen when I try to assimilate 0 dBz values using the WRF3.9.1 version. In older versions, I did not experience this behavious because the model gave 0 dBz values.

Any suggestion on how I can solve this problem (e.g., modifying the default minimum value for the reflectivity)?
Best regards,
Diego
I am running WRFV3.9.1 and I am trying to assimilate reflectivity observations using the EnKF. The main problem I have, is that in areas where it is not observed reflectivity values (clear air), the model produce a default value of -35 dBZ. This produces a problem when I try to assimilate obs, because the data assimilation algorithm try to correct this -35 dBz to 0 dBz, generating in this way values close to 0 dBz (but not exactly 0 dBz), so generating reflectivity values in regions where we should have 0 dBz. In other words, if the minimum default value of the model was 0 dBz, then during the assimilation, the model would remain the same and no spurious reflectivity would be generated.
On the figure attached you could observed what happen when I try to assimilate 0 dBz values using the WRF3.9.1 version. In older versions, I did not experience this behavious because the model gave 0 dBz values.

Any suggestion on how I can solve this problem (e.g., modifying the default minimum value for the reflectivity)?
Best regards,
Diego