Preliminary Study on the Application of GPS Observations to a Mesoscale Numerical Model

Based on observations from 11 stations inside the GPS (global positioning system) observation network,study is performed both on adjustment of the MM5 initial humidity field by means of, and nudging assimilations of, G-PW (short for GPS-sensed atmospheric precipitable water) for a rainfall event happening in the Yangtze delta during June 23-24, 2002. Results show that adjusting the initial moisture field through G-PW will enhance pronouncedly the ability of the initial field to depict vapor distribution, thereby harnessing errors of atmospheric PW prediction at an early stage of model integration to improve more markedly the prediction of 6-h rainfall and, in contrast, nudging assimilations of G-PW show insigni cant amelioration of model prediction, with less effect on the result by using a bigger nudging coefficient. On the whole, compared to successive nudging assimilations of G-PW into the MM5, greater amelioration occurs in 6-h rainfall prediction from the G-PW adjusted initial moisture field. Also, evidence suggests that the improvement of 6-h rainfall prediction with G-PW in correcting the initial humidity field is realized mainly through the amelioration of the ability of grid-scale rainfall prediction while the nudging scheme achieves the improvement largely by bettering sub-grid scale rainfall prediction.