I want to estimate the heating profile, so I add up "RTHCUTEN" (from cumulus), "RTHRATLW" (from long-wave radiation), RTHRATSW" (from short-wave radiation), RTHBLTEN (from PBL parameterization), and H_DIABATIC (from microphysics) (RTHNDGDTEN, nudging is not applied in my em_real case), but I found the outputs of H_DIABATIC contain many negative values, I want to confirm that whether it is normal for a real case running? Thanks a lot.