Collaborative Research: EaSM-3: Understanding the Development of Precipitation Biases in CESM and the Superparameterized CESM on Seasonal to Decadal Timescales

University Of Washington, Seattle WA

Investigators

Abstract

The issue of bias in state-of-the-art global models that are used to study decadal climate variability and prediction has been a long-standing problem in the field of climate science. Biases in tropical precipitation, circulation, and sea-surface temperature (SST) affect simulations of seasonal to interannual, decadal, and longer-time climate variations. Additionally, under this effort three doctoral students will be trained for careers in climate modeling, contributing to national scientific workforce capabilities. The goal of this project is to reduce tropical precipitation biases in Community Earth System Model (CESM) and a new version of the CESM that incorporated the superparameterized cloud scheme, SP-CESM. Tropical precipitation biases have been attributed to poor cumulus parameterizations, ocean coupling feedbacks, ocean eddy mixing deficiencies, land-atmosphere interaction, and remote effects of insufficient low clouds. Sources of error will be separated by looking at the initial development of the biases over the first decade of simulations. Cumulus parameterization errors develop within days to weeks, local ocean coupling errors develop in months, and the remote effect of midlatitude low cloud biases may take years to affect tropical precipitation. This initial rainfall and SST bias development will be used to test improvements to the cloud microphysical parameterizations in CESM and SP-CESM, as well as changes to the cumulus parameterization in CESM. Changes affecting Southern Hemisphere midlatitude low cloud cover will be a focus. The goal will be to significantly reduce the tropical SST and precipitation bias in both models for both initialized multiyear forecasts and long-term projections while maintaining good ENSO variability.