FY-2E资料空间响应订正及对强对流监测改进

建立空间响应匹配滤波 (spatial-response matched filter, SRMF) 方法，针对强对流低温研究目标，开展我国风云二号E星 (FY-2E) 红外亮温订正计算，并选取2013—2014年典型对流天气进行统计分析，从对流的空间分布、发展过程、云团结构等多角度进行方法性能评估。结果表明:对于对流云团结构，SRMF方法可改进FY-2E卫星红外波段对对流云团识别的准确度，减小高温背景对低温对流云团的邻近像元效应，增加了FY-2E卫星对中尺度对流内部小尺度精细化结构的揭示能力；对于对流空间分布，SRMF方法降低了对流判识空间分布统计误差，减少极短时间、极小范围强对流天气的漏判；对于对流识别时间响应，SRMF方法能够正确且提前显现出云团由弱对流向强对流的发展潜势，提高FY-2E卫星探测仪器对强对流天气的临近预警能力。

Performance Improvement for FY 2E Convection Monitoring Using a Spatial response Matched Filter Method

In China, severe convective weather system often causes sudden disasters, and its occurrence time and falling area is difficult to be forecasted. The improvement of convection monitoring and forecasting closely depends on the advance of the monitoring capability. Geostationary satellites can provide a large range, full day cloud information, so they may be the most practical tools for monitoring the convection. In these years, the convective cloud identification method is mainly based on infrared channels of satellites using brightness temperature threshold. The accuracy of the brightness temperature is crucial for convective cloud identification, which depends on not only the satellite calibration, but also the satellite instruments sensitivity, especially for the mesoscale and small scale targets. Based on the observation performance and principle of FY 2E meteorological satellite, a spatial response matched filter (SRMF) method for FY 2E is set up and applied for convection observations in the thermal infrared band, the MTSAT/JAMI is used as the reference standard of the revised method, and the spatial response of VISSR and JAMI is matched with the same infrared band. Accordingly, the infrared channel brightness temperature is corrected. Furthermore, some typical convection examples are selected during summer of 2013 and 2014 for statistics, the SRMF performance is evaluated focusing on convection spatial distribution, development process and convective cloud inner structure. Results indicate recovered images after SRMF processing show more sensitivity of convective cloud identification. For small scale convective core in the mesoscale cloud and very short time convection, it has significant improvement for reducing effect from the high temperature background smoothing, and also it enhances the ability to reveal the mesoscale and finer scales cloud. Besides, after SRMF processing, the convection distribution statistical error is reduced, improving the missing problem for the short time and small scope convective cloud. The SRMF method also enhances the characterization ability for the convection development potential. These results all indicate that the SRMF is more suitable for convection nowcasting, such a progress is believed to be beneficial to convection monitoring and forecasting. In the future, monitoring operational work on the convective weather, for small scale and mesoscale clouds, this SRMF technique could be applied to reduce the overall observation error, and then improve the spatial resolution for the deep convective cloud top identification. The method can also be extended to the detail inner structure of tropical cyclones.