Impacts, Metrics, and Synoptic-Scale Drivers
A new multi-institutional effort to improve understanding and prediction of monsoon rainfall will launch in Fall 2018. This effort is a partnership between William Boos at UC Berkeley, Travis O’Brien at the Lawrence Berkeley National Lab, and Paul Ullrich at UC Davis.
Much of the water supply for low-latitude land regions, including the southwestern U.S., is delivered by propagating vortices and waves embedded within seasonal-mean, continental-scale monsoon circulations. These propagating disturbances—which include monsoon depressions, easterly waves, and orographically trapped moisture surges—frequently produce extreme precipitation that challenges water management networks, hydropower generation, and natural ecosystems. Furthermore, the influence of greenhouse gas emissions, atmospheric aerosols, and other products of energy generation on seasonal-mean and extreme monsoon precipitation remains poorly understood, in part because the global models and data used to study planetary-scale monsoon flow are only now beginning to achieve resolutions needed to represent the transient disturbances that produce much of the mean and extreme monsoon rainfall. This project will improve understanding of a variety of transient atmospheric disturbances in monsoon regions, the precipitation extremes they produce, and their interactions with larger-scale monsoon winds. The work will focus on synoptic (2-12 day) time scales, which include important events such as Gulf of California moisture surges, East Pacific easterly waves, and waves in the subtropical jets that lie on the poleward edge of monsoon regions. Metrics and statistical models will be developed for the North American monsoon; the applicability of these metrics and models to altered atmospheric states will be improved through comparison with the South Asian monsoon, which shares important dynamical characteristics with the North American monsoon.
The main research tasks involve (1) developing feature-tracking algorithms to identify synoptic-scale disturbances in large ensembles of observational data and numerical model output, including output from variable resolution integrations of the Energy Exascale Earth System Model (E3SM), (2) estimating the multidimensional sensitivities of disturbance genesis, growth, and rain rate to properties of the larger-scale atmospheric state in these ensembles, and (3) constructing physically based statistical models for disturbance genesis, amplification rate, propagation speed, and precipitation rate. Rather than improving prediction of individual synoptic events, the aim is to understand, measure, and project variations in the ensemble of synoptic-scale precipitation events as a function of natural and forced modes of variability. The statistical models developed here will be derived from mechanistic understanding of individual classes of synoptic-scale phenomena. This approach will provide process-based statistical projections of changes in precipitation extremes together with process-based assessment of model bias.
Project benefits and outcomes include improved understanding of the physical processes responsible for variations in extreme rainfall in the southwestern U.S. and other monsoon regions, rigorous characterization of E3SM bias in simulating the full rainfall distribution, and creation of dynamically based metrics for extreme monsoonal rainfall. These metrics will be incorporated into existing metrics packages at Department of Energy facilities, including, but not limited to, the Toolkit for Extreme Climate Analysis (TECA). The project will train two graduate students and a postdoctoral scientist.