East China Summer Rainfall Variability of 1958–2000: Dynamical Downscaling with a Variable-Resolution AGCM

Abstract

A variable-grid atmospheric general circulation model, namely, Laboratoire de Météorologie Dynamique-zoom, version 4 (LMDz4), with a local zoom over eastern China, is driven by 40-yr European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40) data and is used as a downscaling tool of summer rainfall variability for the period 1958?2000. During the integration, the model temperature and wind were nudged to the ERA-40 data through a relaxation procedure. The performance of the LMDz4 in simulating the regional rainfall features is thoroughly assessed through a comparison to both rain gauge data and the reanalysis product. The dynamical downscaling improves not only the climatology of the monsoon major rainband but also the interannual variability modes of rainfall over eastern China in comparison with that of the ERA-40 data. The added values of LMDz4 are evident in both the spatial patterns of dominant rainfall variability modes and the associated temporal variations. A comparison of rainfall averaged over several typical regions shows improvement as a better-matched variability and a reduced root-mean-square error, except for the region over the lower reaches of the Yellow River valley, where the model shows bias because of the northward shift of the monsoon rainband. This rainband shift is caused by the stronger low-level southerlies and the lower specific humidity over southern China. The stronger southwestern wind transports excessive water vapor northward, and the underestimation of specific humidity implies that air masses need to go farther north to reach condensation. Both favor a northward shift of the major rainband. The analysis demonstrates that a variable-resolution AGCM can be a useful tool for the dynamical downscaling of rainfall variability over eastern China, although the rainband bias remains evident as with many other regional climate models.