摘要
为了支持短时临近预报,利用新一代数值预报技术,结合高时空分辨的新一代探测所提供的观测信息,开发了基于快速更新循环同化的数值预报系统。基于GRAPES预报模式及其三维变分,开发了逐时循环同化,包括雷达、卫星、飞机、常规地面和探空等观测资料的同化模块,并采用nudging技术,引入雷达回波,订正模式的初始云水、雨水等信息,实现快速更新预报。多种测试和汛期连续试验表明,系统运行稳定可靠。逐时同化场合理,与实况基本一致。通过1个月的滚动预报综合分析和个例预报分析均表明,预报稳定有效,与观测分布基本一致,初步具备开展短时临近预报的能力。
In order to support short-term nowcasting, a new generation of numerical prediction techniques （such as the latest developments of GRAPES model and its three-dimensional variational assimilation） is used, together with observations provided by new-generation means that are highly resolved both temporally and spatially, to develop a rapid updating cycle assimilation and forecasting system. The system provides hourly assimilations of radars, satellites, aircraft, conventional surface observations and soundings; the rapid updating forecasting can be achieved through nudging, revision of the initial cloud water mixing ratio, correction of rain water mixing ratio and other information and introducing radar echoes. A variety of tests and consecutive seasonal experiments show that the system is stable and reliable. Analysis showed that the assimilation result, situation forecasting, and precipitation forecasting from the system are basically consistent with corresponding observations, indicating that the system has short-term nowcasting capabilities.
出处
《热带气象学报》
CSCD
北大核心
2010年第1期49-54,共6页
Journal of Tropical Meteorology
基金
公益性行业(气象)科研专项(GYHY200706014)
"十一五"国家科技支撑计划重点项目(2006BAC02B)
973计划项目(2004CB418307
新技术推广CMATG2005M17
社会公益研究专项(2005DIB3J110)共同资助
关键词
临近预报
数值预报
快速循环同化
nowcasting
numerical prediction
rapid update cycle
