The Research Of AMSU Data Using In The Heavy Rain Simulation With MM5-adjoint Model

Nowadays, observed data from meteorological satellites are used in many kinds of numerical weather prediction models. For the density of these data more than routine observations both in time and space, the use of them increased the improved the model output fields and lead to the advance of the numerical weather prediction technology. In this paper, the AMSU(Advanced Microwave Sounding Units) data from the NOAA-16 polar orbit satellite are used, after directly assimilated by the GRAPES-3Dvar system, the output data are the variables of the MM5 model, then can be used in the MM5-adjoint model. A heavy rainfall is taken for an example, compared with the variable fields before assimilating AMSU data to discuss the effect of AMSU in the prediction of summer heavy rainfall in this country.In this paper, the AMSU data from NOAA-16 satellite are used in the simulation of the heavy rainfall during July 4~5 2003, which occurred in Yangze-Huaihe River. The results suggested that: (1) When AMSU are added into the model, the initial fields could have been changed, especially in temperature and humidity. (2) Compared with the simulation results from MM5 model using only conventional observations, it is found that retrievals have ability to reveal clearly the structure of temperature and humidity fields, obviously enhance the intensity and the region of model simulated precipitation, so that the accuracy of MM5-Adjoint model rainfall prediction and it's Ts grades are greatly improved. (3)AMSU data increased the quantity of temperature and wind, could clearly revealed the occurrence and change of the low vortex which lead to the rainfall. The results suggest that AMSU data could improve the simulation of the summer heavy rainfall in China, and has latent capacity to ameliorate the operational weather prediction.