| With the development of numerical weather forecast model, it is possible to have high resolution data of radar applied to numerical forecast model to improve the numerical forecast. But in detecting process of Doppler weather radar, there will be non-precipitation echoes that lead to quality problems, affecting the application of radar data in business.Therefore, removing clutter in order to improve the quality of radar data for the practical application of radar data has important significance. However, our current status and the use of radar data lag far behind foreign countries, especially the quality control research of Doppler radar data. Therefore, how to remove clutters appeared in the detecting process and have high quality radar data applied to numerical forecast system to improve the accuracy of numerical forecast, has become an urgent problem.This paper mainly researches and develops a quality control scheme of Doppler radar echoes. In the quality control process, firstly a statistical of different echoes with the horizontal and vertical structure characteristic was done; Based on the review of researching radar data quality control at home and abroad, a new quality control scheme based on fuzzy mathematics theory of fuzzy c-means (FCM) algorithm was developed. The experiments result show that the proposed scheme can effectively identify precipitation echoes and non-precipitation echoes, and is a kind of a new try in radar reflectivity data quality control.Subsequently, the rainwater mixing ratio retrieved by reflectivity factor, the vertical velocity and radial velocity were put into GRAPES 3DVAR directly in order to improve a good initial field for GRAPES numerical forecast model. Using GRAPES 3DVAR assimilation system, a simulate test of a strong convective weather in northern Anhui province on June 5,2009 was done. Four groups testing results indicate that radar data can effectively adjust the initial fields and can improve the forecast of short-term precipitation.
|
PDF Full Text Request