Comparison Of Cloud Motion Wind Datasets From Different Deriving Systems And Their Applications In 3DVAR Typhoon Forecast

The western Pacific Ocean is the most frequent sea area where tropical cyclones take place. The East Asia, especially China and Korea are all under the influence of typhoon. In the past, due to a lack of information from observation above the sea, the research of situation before a typhoon landing was rather limited. In recent years, the application of a series of unconventional high resolution spatiotemporal data, for example, satellite cloud motion wind (CMW), complements the defect of observational data above the sea. The assimilation of the unconventional data in models can effectively improve the typhoon track forecast.This study includes comparing of cloud motion wind datasets from different deriving systems and also comparing CMW data with radiosonde data. These CMW data are from CWIS (cloud motion wind inferring system developed by NUIST), NSMC (China Satellite Meteorological Center) and JMSC (Japan Meteorological Satellite Center). In addition, numerical experiments and diagnostic analyses are conducted for the effect of CWIS CMW data on typhoon forecasting in WRF 3DVAR and analyzed the difference of numerical experiments of three kind of CMW data. This study could be giving a clue of improving the typhoon forecasting. The results are as follows:Firstly, it was evaluated by comparing three different kinds of CMW data each other and comparing CMW data with radiosonde data for evaluating the CWIS data. The wind speeds measured by CWIS data, are similar to NSMC data and JMSC data, but the wind directions of CWIS data relatively greater than other data. Although there are some errors in CWIS data, it insures good reliability of the weather situation.Secondly, Both assimilation of one channel and two channels of CWIS data have remarkable improvement in forecasting the track of typhoon Masha (200509). The typhoon track forecasted by assimilation of two channels (the time interval is 3 hours) is much better.Thirdly, CWIS, NSMC and JMSC data have different grade of improvement in forecasting the track of typhoon Khanun (200515), Three systems yield 49%, 19% and 39% respective reduction in forecast mean error. GTS (radiosonde, surface) data simulation has some negative effect on forecasting typhoon track, but when it is combined with CMW data, the track forecasting is greatly improved. Especially when combined with JMSC data, the result is better than the assimilation of JMSC data and GTS data respectively. The assimilations of three kinds of CMW data and GTS data have 26 %, 15% and 50% in forecast mean error, respectively.Fourthly, it was analyzed typhoon Nabi (200514) caused heavy rainfalls, which mainly happen in the coastal area of Korea. These heavy rainfalls take placed due to inverted trough, CMW data has good simulation of the typhoon main precipitation area and core.Fifthly, Convective schemes have great influence on the track of typhoon Nabi. Compared with Betts-Miller-Janjic (BMJ) scheme, Kain-Fritcz (KF) scheme can better simulate typhoon track. When using KF scheme, choosing micro-physical scheme is better than no choosing scheme. The results got by Ferrier, WSM6 and Lin scheme are really close to the observation. JMSC CMW data using different parameterization schemes has different effect on typhoon forecasts. In KF scheme, typhoon track only by Kessler scheme is a better improvement than controlled experiment. In BMJ scheme, every micro-physical method can provide better results than controlled experiment.