The Impact of NOAA Satellite Soundings on the Numerical Analysis and Forecast System of the People's Republic of China. Part 1. FGGE Data. Part 2. December 1987 Data

The impact of NOAA satellite sounding data on the numerical analysis and forecast system of the People's Republic of China has been investigated. Four experiments were conducted. They are: (1) NOSAT; which uses conventional data only and provides benchmark statistics for the performance of the forecast model, (2) SAT; which includes satellite and conventional data (3) ALLSAT; which uses satellite data only, except for conventional surface data required to define the 1000 mb height reference level, and (4) DATGUESS; which uses the ALLSAT analysis as the first guess to which conventional dat are then added. Seven areas deemed to have adequate conventional data for definition of truth. The impact of satellite data was then assessed by computing systematic errors, changes in S1 skill score and root mean square errors for 24 and 48 h forecasts of Sea Level Pressure (SLP) and 500 mb height. Overall, the Chinese model is found to underforecast the strength of features in the mean SLP fields, i.e., pressures are too low near high pressure systems and too high near low pressure systems. The same type of pattern is also seen in the mean 500 mb height forecast. In general, satellite data have been shown to reduce systematic error near where the data have been introduced in the vicinity of low pressure systems. A positive impact on root-mean-square errors of up 10 11.5% was observed over land for SLP and 8% over ocean for 500 mb height for forecast period of 48 h. However, only a meager increase of 5.5% in S1 skill score was observed at SLP for 48 h forecast over oceans.