Abstract
We evaluate the statistics of global precipitation extremes of an ensemble (4 realizations) of low resolution (ne30) and high-resolution (ne120) ACME v0 FAMIP simulations (1979-2005). A regionalization framework is applied to improve the sample size of extreme events by using data from neighboring regions with a homogeneous climate. The generalized extreme value theory is used to quantify the statistics and non-stationary (tele-connections here) behavior of precipitation extremes and validated against the global land NOAA Climate Prediction Center (CPC) gauge-based analysis data. While the high-resolution model generally produces stronger extremes improving upon the low-resolution model, it also generates excessive extremes as compared to CPC data. The high-resolution model also improves the simulation of tele-connections of extremes with low-frequency phenomenon like the North Atlantic Oscillation (NAO) as compared to its low-resolution counterpart. These diagnostics for climate extremes are part of ACME Tier1b atmospheric diagnostics package and follow a software design similar to the ACME coupled diagnostics package.