首页 | 本学科首页   官方微博 | 高级检索  
     检索      

HY-2散射计风速降雨影响的神经网络校正
引用本文:王婧,解学通,夏丽华.HY-2散射计风速降雨影响的神经网络校正[J].地理信息世界,2017,24(1).
作者姓名:王婧  解学通  夏丽华
作者单位:广州大学地理科学学院,广东广州,510006
基金项目:国家自然科学基金项目,广州市产学研协同创新重大专项项目,国家863计划项目,广东省科技计划项目
摘    要:海洋二号搭载的笔形圆锥扫描微波散射计(HY2-scat)是国内第一个业务化运行的,可提供大量实时海面风场数据的微波传感器。由于Ku波段散射计测风原理和微波传输特性,受到降雨影响的散射计反演风场数据准确度降低。降雨导致的微波传播路径衰减,雨滴对微波直接后向散射导致的回波能量增加和雨滴对海表面毛细波的干扰等综合效应,使得降雨条件下散射计测风风速计算值偏高,风向计算值偏差较大。针对散射计反演风速受降雨影响的特点引入神经网络模型,使用准确度较高的NWP数值预报模式风场数据作为参考,对受降雨影响的HY-2散射计反演L2B级标准风场数据产品进行校正,改进HY-2散射计反演风矢量在降雨条件下的准确度。与受降雨影响的散射计反演风场风速偏差相比较,经过神经网络校正后的风速偏差减小,说明该方法适用于改善受降雨影响的HY-2散射计测风风速精度。

关 键 词:微波散射计  降雨影响  神经网络模型

The Correction Effect of Rainfall on HY-2 Scatterometer Wind Field Retrieval by Neural Network Model
WANG Jing,XIE Xuetong,XIA Lihua.The Correction Effect of Rainfall on HY-2 Scatterometer Wind Field Retrieval by Neural Network Model[J].Geomatics World,2017,24(1).
Authors:WANG Jing  XIE Xuetong  XIA Lihua
Abstract:The HY2-scat is the first pen-shaped cone-scan scatterometer in China to provide a large amount of realtime sea surface wind data.Due to the principle of Ku-band scatterometer and the microwave transmission characteristics,the accuracy of wind field data retrieval by scatterometer was reduced.The calculated value of the wind speed of the scatterometer is overestimated and the deviation of the calculated wind direction is increased due to the attenuation of rainfall and the interference of raindrops to the capillary waves in the sea surface make.Using the NWP surface wind field simulation data with high accuracy as a reference,the neural network model can be used to correct the data of the L2B-level wind field data from the HY2 scatterometer,and improve the HY2 scatterometer inversion wind vector accuracy in rainfall conditions.Comparing with the scatterometer retrieval wind speed deviation,the wind speed deviation of the wind field corrected by the neural network is reduced,which shows that the method is suitable for improving the wind speed accuracy of the HY2 scatterometer wind field retrieval affected by rainfall.
Keywords:microwave scatterometer  rainfall effect  neural network model
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号