skip to main content

SciTech ConnectSciTech Connect

Title: Evaluation of CMIP5 continental precipitation simulations relative to satellite-based gauge-adjusted observations

Numerous studies have emphasized that climate simulations are subject to various biases and uncertainties. The objective of this study is to cross-validate 34 Coupled Model Intercomparison Project Phase 5 (CMIP5) historical simulations of precipitation against the Global Precipitation Climatology Project (GPCP) data, quantifying model pattern discrepancies and biases for both entire data distributions and their upper tails. The results of the Volumetric Hit Index (VHI) analysis of the total monthly precipitation amounts show that most CMIP5 simulations are in good agreement with GPCP patterns in many areas, but that their replication of observed precipitation over arid regions and certain sub-continental regions (e.g., northern Eurasia, eastern Russia, central Australia) is problematical. Overall, the VHI of the multi-model ensemble mean and median also are superior to that of the individual CMIP5 models. However, at high quantiles of reference data (e.g., the 75th and 90th percentiles), all climate models display low skill in simulating precipitation, except over North America, the Amazon, and central Africa. Analyses of total bias (B) in CMIP5 simulations reveal that most models overestimate precipitation over regions of complex topography (e.g. western North and South America and southern Africa and Asia), while underestimating it over arid regions. Also, while mostmore » climate model simulations show low biases over Europe, inter-model variations in bias over Australia and Amazonia are considerable. The Quantile Bias (QB) analyses indicate that CMIP5 simulations are even more biased at high quantiles of precipitation. Lastly, we found that a simple mean-field bias removal improves the overall B and VHI values, but does not make a significant improvement in these model performance metrics at high quantiles of precipitation.« less
 [1] ;  [1] ;  [2]
  1. Univ. of California, Irvine, CA (United States). Dept. of Civil and Environmental Engineering
  2. Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
Publication Date:
OSTI Identifier:
Report Number(s):
Journal ID: ISSN 2169-897X
Grant/Contract Number:
AC52-07NA27344; R11AP81451; OISE-1243543
Accepted Manuscript
Journal Name:
Journal of Geophysical Research: Atmospheres
Additional Journal Information:
Journal Volume: 119; Journal Issue: 4; Journal ID: ISSN 2169-897X
American Geophysical Union
Research Org:
Lawrence Livermore National Laboratory (LLNL), Livermore, CA
Sponsoring Org:
USDOE; United States Bureau of Reclamation (USBR); National Science Foundation (NSF)
Country of Publication:
United States
54 ENVIRONMENTAL SCIENCES; 58 GEOSCIENCES precipitation; CMIP5; satellite data; GPCP