Abstract
Four recent reanalyses—the 55-yr Japanese Reanalysis Project (JRA-55), Interim ECWMF Re-Analysis (ERA-I), NCEP Climate Forecast System Reanalysis (CFSR), and NASA Modern-Era Retrospective Analysis for Research and Applications (MERRA)—are assessed to clarify their quality in representing the diurnal cycle over East Asia. They are found to present similar patterns/structure and summer progress of the mean wind diurnal cycle, whereas they exhibit some differences in diurnal amplitude, particularly for the low-level meridional wind. An evaluation with intense soundings suggests that the amplitude difference mainly results from the diurnal variation of mean bias that differs among reanalyses. The root-mean-square (RMS) error is found to have a diurnal variation more evident in CFSR and MERRA than that in JRA-55 and ERA-I, which strongly affects the representation of the varying diurnal amplitude at the peak hours of RMS error.
Compared with satellite-derived rainfall, the four reanalyses are shown to reproduce well the rainfall diurnal cycle over East Asia in terms of large-scale terrain contrast, summer progress, and interannual variability. JRA-55 even presents a long-term increase of morning rainfall percentage over the east China plain over the past four decades, consistent with rain gauge observations. The four reanalyses exhibit some considerable discrepancies at regional scale; JRA-55 gives the best capture of the rainfall diurnal cycle over the Tibetan Plateau and the eastward propagation to the eastern lees. These results suggest that new reanalyses are potentially applicable for studying the large-scale diurnal variability over East Asia, whereas their different preferences, especially at regional scale, should be of concern in data application.
