Given two or more parameterizations that represent the same physical process in a climate model, scientists are sometimes faced with difficult decisions about which scheme to choose for their simulations and analysis. These decisions are often based on subjective criteria, such as “which scheme is easier to use, is computationally less expensive, or produces results that look better?” Uncertainty quantification (UQ) and model selection methods can be used to objectively rank the performance of different physical parameterizations by increasing the preference for schemes that fit observational data better, while at the same time penalizing schemes that are overly complex or have excessive degrees-of-freedom. Following these principles, we are developing a perturbed-parameter UQ framework to assist in the selection of parameterizations for a climate model. Preliminary results will be presented on the application of the framework to assess the performance of two alternate schemes for simulating tropical deep convection (CLUBB-SILHS and ZM-trigmem) in the U.S. Dept. of Energy's ACME climate model. 

This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344, is supported by the DOE Office of Science through the Scientific Discovery Through Advanced Computing (SciDAC), and is released as LLNL-ABS-675799.