Evaluation of Precipitation from Reanalyses Products Using Gauge-Based Analysis of Daily Precipitation over East Asia

During last decades, several reanalysis datasets have been produced for climate research. Such reanalyses include the National Centers for Environmental Prediction-National Center for Atmospheric Research (NCEP-NCAR) reanalyses covering the period from 1948 until present, the 40-yr European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analyses (ERA-40) from 1957 to 2002, and the Japanese Reanalysis (JRA-25) from 1979 to 2004. However, the present generation of reanalysis has many problems that make them sub-optimal or even unusable for regional climate studies. Therefore, the applicability evaluation of the reanalyzed products is crucial for their intelligent utilization, the improvement of the data assimilation technique and the reliability of climate change research. In this study, a new gauge-based analysis of daily precipitation (regard as observations) developed by the ‘Asian Precipitation Highly-Resolved Observational Data Integration Towards Evaluation of the Water Resources (APHRODITE)’ project will be used to validate the precipitation products from NCEP-NCAR, NCEP-DOE, ERA-40 and JRA-25 reanalysis over East Asia monsoon region (5°-60°N, 65°-155°E) for a 25-yr period from 1979 to 2003. The reliability and applicability represented by reanalyzed precipitation in climate research will be analyzed through multi-statistical diagnostic analysis methods on different spatio-temporal scales, especially in seasonal and interannual variation over East Asia. The main results about this study reveal that the ERA-40 and JRA-25 are better than NCRP-NCAR and NCEP-DOE to describe the spatial distribution and temporal variation represented by observations in most regions of East Asia, but the NCEP-DOE has a better ability to represent the variation of Meiyu belt in the Yangtze-Huai River Valley of China. With regard to the magnitude of the precipitation difference among the reanalyes and the observations, the JRA-25 is closer to the observed precipitation than others over most domains. However, some uncertainties existed in the study results are still needed to discus due to the difference of data resolution and other issues.