Precipitation Extremes

Abstract

We present comparisons of precipitation extremes for the high resolution (T341) and low-resolution (T85) global AMIP simulation ensembles for the period 1979-2005 with the spectral version of CAM4 from the ultra high-resolution climate modeling project. We design a correlation-based regionalization framework to quantify precipitation extremes, where samples of extreme events for a grid box may also be drawn from neighboring grid boxes with statistically equal means and statistically significant temporal correlations. We model precipitation extremes with the Generalized Extreme Value (GEV) distribution fits to time series of annual maximum precipitation. We capture the non-stationary behavior of extremes by building suitable linear models of GEV distribution and estimate trends and dependence of extremes on natural modes of variability, namely the El Nino Southern Oscillation (ENSO) and North Atlantic Oscillation (NAO). Our preliminary analysis reveals that the high-resolution model substantially improves the simulation of stationary precipitation extreme statistics when compared to NOAA Climate Prediction Center (CPC) Gauge-based Unified Daily Precipitation Data, similar to previous work. Observational data exhibits significant non-stationary behavior of extremes only over some parts of the world. While the high-resolution simulations improve upon the low-resolution model in simulating the non-stationary behavior, the trends are statistically significant only over some regions.