This PR prevents a divide by 0 case with the SAM microphysics. Namely, the coefficients of evaporation for snow/graupel depend upon the vapor pressure of ice (erf_esati). When T>273.16, the ice melts and thus its vapor pressure is 0. To alleviate this issue, we limit the temperature in the SAM initialization so that esati is computed at T=273.16 if T>273.16. This keeps the coefficients bounded and should not affect evaporation rates since that process requires finite q_{s/g} and that condition requires T<273.16.
This PR prevents a divide by 0 case with the SAM microphysics. Namely, the coefficients of evaporation for snow/graupel depend upon the vapor pressure of ice (
erf_esati). WhenT>273.16, the ice melts and thus its vapor pressure is 0. To alleviate this issue, we limit the temperature in the SAM initialization so thatesatiis computed atT=273.16ifT>273.16. This keeps the coefficients bounded and should not affect evaporation rates since that process requires finiteq_{s/g}and that condition requiresT<273.16.