Intercomparison and diagnosis of Mei-Yu rainfall derived from satellite data

The Chinese term Mei-Yu refers to a regional rainy season over the Yangtze River Valley in East China, which is initiated by the planetary-scale rainband advancing across the East Asian continent during the Asian summer monsoon period. As an important regional feature of the Asian summer monsoon, characteristics of Mei-Yu rainfall have been investigated for decades by many researchers. However, a unified definition of onset and withdrawal dates of Mei-Yu rainfall has been difficult to agree upon, partly because of the limitations of the raingauge datasets available. In a recent study, Janowiak and Xie used estimated precipitation from the pentad version of the GPCP (Global Precipitation Climatology) global precipitation analysis to determine onset and withdrawal dates for monsoon precipitation in many parts of the globe. In this study we use several different estimates of large-scale precipitation, mainly based on satellite data but in some cases including gauge observations, to examine the annual and interannual variation of precipitation over the Mei-Yu region (110°~125°E, 25°~35° N) and surrounding areas. We have two main goals: to determine those aspects of Mei-Yu rainfall that are robust across the various data sets, and to examine the sensitivity of features such as onset and withdrawal date to variations in input datasets. We will utilize the CMAP (CPC Merged Analysis of Precipitation) and GPCP (Global Precipitation Climatology) composite datasets, along with precipitation estimated from various satellite observations such as TOVS (TIROS Operational Vertical Sounder), TRMM (Tropical Rainfall Measuring Mission) precipitation radar and TRMM Microwave Imager, GPI (GOES Precipitation Index) from geostationary satellite infrared data, OPI (OLR Precipitation Index) from NOAA AVHRR data, and passive microwave data from SSM/I (Special Sensor Microwave/Imager). The data sets to be used have different characteristics, cover different time periods, and have different spatial and temporal resolutions. We will intercompare the results derived from different data sets and attempt to arrive at a consensus description of Mei-Yu rainfall. Our results based on these satellite-derived products will be compared to those based on an Optimum Interpolation analysis of land gauge data and to the published results of previous work.