At the decadal-to-century climate time scale, warming causes extreme precipitation intensity to increase; at the weather-to-interannual time scales, extreme precipitation intensity decreases after the local temperature exceeds a certain threshold. The latter phenomenon, known as “negative scaling”, hinders extrapolation of precipitation extremes in a warming climate. Recently, it was found that when controlled for moisture limitation (as reflected by atmospheric saturation deficit), extreme precipitation intensity increases monotonically with surface air temperature and shows no negative scaling at the weather timescale; moreover, in saturated atmosphere, the observed extreme precipitation-temperature scaling ratio does not vary much across different regions and is close to the theoretical Clausius-Clapeyron rate. However, climate models produce a wide range of scaling ratio that is significantly higher in the tropics than the extratropics. This team will investigate potential causes for model biases in reproducing the emergent extreme precipitation-temperature relationship, and how these biases may propagate to influence the projection of future extremes. Among others, a major focus will be on two interacting factors, the roles of soil moisture feedback and mesoscale convective systems.

The proposed project will use sub-daily observational data from meteorological stations, gridded precipitation data derived from satellite remote sensing, global reanalysis products, an observation based mesoscale convective system tracking dataset, existing simulations of Energy Exascale Earth System Model (E3SM) and other Earth System Models (ESMs) at different spatial resolutions, and new E3SM experiments. The overarching goal is to better understand and predict future changes of extreme precipitation over land. Specifically, the team proposes to (1) Assess the performance of E3SM and other ESMs at different resolutions in reproducing the extreme precipitation scaling relationships with surface air temperature and saturation deficit and identify the sources of model biases; (2) Quantify how model-projected future changes of extreme precipitation are related to the model’s scaling ratio under saturation in the present-day climate, and to develop emergent constraints for future projections; (3) Investigate how local and regional land surface conditions and feedback may influence extreme precipitation intensity and its scaling relationship with environmental variables, and how the land surface impact may change in a warmer climate; (4) Evaluate the role of mesoscale convective systems in the extreme precipitation scaling with temperature and saturation deficit, and how model biases in simulating mesoscale convective systems precipitation may contribute to model biases in the scaling relationships.   

The proposed research will advance understanding of the earth system’s water cycle and provide guidance for future model development effort to reduce biases and uncertainties related to extreme precipitation modeling and projection. Two emergent metrics for assessing model performance in simulating extreme precipitation will be developed for inclusion in the Program for Climate Model Diagnosis and Intercomparison Metrics Package.