An Analysis of the Potential for Extreme Temperature Change Based on Observations and Model Simulations

The study analyzes observational climate data for June–August 1977–2004 and simulations of current and future climate scenarios from a nested GCM/regional climate model system to assess the potential for extreme temperature change over the eastern United States. Observational evidence indicates that anomalously warm summers in the eastern United States coincide with anomalously cool eastern Pacific sea surface temperatures, conditions that are conducive to geopotential ridging over the east, less frequent precipitation, and lower accumulated rainfall. The study also found that days following nighttime rain are warmer on average than daytime rain events, emphasizing the importance of the timing of precipitation on the radiation balance. Precipitation frequency and eastern Pacific sea surface temperature anomalies together account for 57% of the 28-yr variance in maximum surface temperature anomalies. Simulation results show the sensitivity of maximum surface air temperature to the moist convection parameterization that is employed, since different schemes produce different diurnal cycles and frequencies of precipitation. The study suggests that, in order to accurately project scenarios of extreme temperature change, models need to realistically simulate changes in the surface energy balance caused by the interannual variation of these precipitation characteristics. The mesoscale model that was realistic in this respect predicted much warmer mean and maximum surface air temperatures for five future summers than the parallel GCM driving simulation.