Tests of Single-Column Model Physics Using RACORO Data and Bayesian Calibration

The RACORO dataset provides high-quality observations of shallow convection and forcing conditions over a 3-day period within the mid-continent US. Previous single column model experiments were challenged to produce many of the characteristics of the RACORO observations, generally pointing to limitations in generating turbulent kinetic energy within the boundary layer and an over-eager triggering of deep convection. Similar failures also exist in global GCMs. Here we revisit this problem with an updated version of the Community Single Column Model that makes use of the CLUBB parameterization of shallow convection and Bayesian calibration to ask the broader question of whether such data sets and testing framework is useful for diagnosing and potentially resolving issues that arise in global climate models.