Scale-Aware and Definition-Aware Evaluation of Modeled Near-Surface Precipitation Frequency Using CloudSat Observations

Abstract

CloudSat's 94-GHz Cloud Profiling Radar provides unique near-global observations of precipitation frequency and intensity. Here CloudSat-based diagnostics for near-surface precipitation frequency are implemented in publicly available software that is widely used for climate model evaluation. The new diagnostics are “definition aware” and “scale aware.” As a result, the diagnostics enable robust assessment of modeled near-surface precipitation frequency at a range of intensity classes. The new diagnostics are used to evaluate precipitation frequency in a state-of-the-art climate model, the Community Earth System Model version 1 (CESM1). CESM1 rains and snows too frequently, a bias that is especially pronounced for light rain. Conversely, while rare in both observations and CESM1, the heaviest rainfall events occur too infrequently in CESM1. Though the spatial distribution of snowfall events matches observations well, CESM1 also exhibits excessive snow frequency biases. Despite these biases, projected CESM1 changes in reflectivity-based diagnostics provide interesting insights into what a future 94-GHz radar could detect in a warmer world. With 3 °C of global warming, a future CloudSat-class mission would detect substantial conversion of snow to rain at midlatitudes, a narrowing of the Tropical Pacific rain belt, increased light rain in subtropics, and increased snow frequency in polar regions. The future CESM1 simulations also provide evidence that present-day spatial and magnitude biases imprint themselves on precipitation frequency changes. In summary, new precipitation frequency diagnostics for a range of precipitation intensities robustly expose climate model biases and inform expectations for observable future precipitation changes in a warming world.