Seasonal and Regional Biases in Cmip5 Precipitation Simulations

Zhu Liu,Ali Mehran,Thomas J. Phillips,Amir AghaKouchak

DOI: https://doi.org/10.3354/cr01221

IF: 1.459

2014-01-01

Climate Research

Abstract:This study provides insight into how CMIP5 climate models perform in simulating summer and winter precipitation at different geographical locations and climate conditions. Precipitation biases in the CMIP5 historical (1901 to 2005) simulations relative to the Climatic Research Unit (CRU) observations are evaluated over 8 regions exhibiting distinct seasonal hydroclimates: moist tropical (Amazonia and central Africa), monsoonal (southern China), moist continental (central Europe), semi-arid (western United States and eastern Australia), and polar (Siberia and Canada). While the bias and monthly quantile bias (MQB) reflect no substantial differences in CMIP5 summer and winter precipitation simulations at the global scale, strong seasonality and high inter-model variability are found over the selected moist tropical regions (i.e. Amazonia and central Africa). In the semi-arid regions, high inter-model precipitation variability is also displayed, especially in summer, while the median of simulations is an overestimate of both winter and summer precipitation. In Siberia and central Europe, most CMIP5 models underestimate summer precipitation, and overestimate it in winter. Also, the MQB values decrease as the choice of quantile thresholds increase, implying that the underestimation of summer precipitation is primarily associated with biases in lower quantiles of the precipitation distribution. While the CMIP5 models exhibit similar behaviors in simulating high-latitude winter precipitation, they differ substantially in summer simulations for the selected Canadian and Siberian regions. Finally, in the monsoonal southern China region, CMIP5 models exhibit large overall precipitation biases in both summer and winter, as well as at higher quantiles.