Make the behavior previously enabled by setting -DSP_MCICA_RAD the default for SP simulations. Essentially, this just guarantees that the cloud fraction diagnosed within the CRM is always used by the radiative transfer parameterization, rather than relying an a somewhat arbitrary condensed water amount threshold. This then enables a non-trivial subcolumn sampling of clear/cloudy "pixels" within the radiative transfer MCICA scheme. The old behavior (which we had already went away from before the Early Science campaign) set each radiative column to be either entirely clear or cloudy at a given level based on a threshold of condensed cloud water amount. Since our compsets had included the -DSP_MCICA_RAD flag, this PR should be BFB, and just clarifies the code somewhat, and makes it harder for someone to accidentally override the intended behavior.
Make the behavior previously enabled by setting
-DSP_MCICA_RADthe default for SP simulations. Essentially, this just guarantees that the cloud fraction diagnosed within the CRM is always used by the radiative transfer parameterization, rather than relying an a somewhat arbitrary condensed water amount threshold. This then enables a non-trivial subcolumn sampling of clear/cloudy "pixels" within the radiative transfer MCICA scheme. The old behavior (which we had already went away from before the Early Science campaign) set each radiative column to be either entirely clear or cloudy at a given level based on a threshold of condensed cloud water amount. Since our compsets had included the-DSP_MCICA_RADflag, this PR should be BFB, and just clarifies the code somewhat, and makes it harder for someone to accidentally override the intended behavior.