基于多模型神经网络的湿度廓线反演研究  

Research on humidity profile inversion based on multi model neural network

作　　者：王金虎[1,2,3,4] 肖安虹 陈后财 王昊亮[1,2] 刘萱[5] 蔡海强 WANG Jinhu;XIAO Anhong;CHEN Houcai;WANG Haoliang;LIU Xuan;CAI Haiqiang(Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters,Nanjing University of Information Science and Technology,Nanjing 210044,China;Key Open Laboratory of Aerosol and Cloud Precipitation,China Meteorological Administration,Nanjing 210044,China;Key Laboratory of Middle Atmosphere and Global Environment Observation,Chinese Academy of Sciences,Beijing 100029,China;Nanjing Xinda Institute of Safety and Emergency Management,Nanjing 210044,China;China Research Institute of Radiowave Propagation,Qingdao 266107,China)

机构地区：[1]南京信息工程大学气象灾害预报预警与评估协同创新中心,南京210044 [2]中国气象局气溶胶与云降水重点开放实验室,南京210044 [3]中国科学院中层大气和全球环境探测重点实验室,北京100029 [4]南京信大安全应急管理研究院,南京210044 [5]中国电波传播研究所,青岛266107

出　　处：《电波科学学报》2024年第1期181-190,共10页Chinese Journal of Radio Science

基　　金：国家自然科学基金(41905026);江苏省自然科学基金(BK20170945);南京信息工程大学人才启动基金(2016r028);江苏省333工程高层次人才培养资助(第三层次)。

摘　　要：为提升微波辐射计对大气廓线探测的精度,利用ARM大气观测站提供的地基微波辐射计、毫米波测云雷达以及探空数据,构建了两种添加不同云信息的反向传播神经网络(back propagation neural network,BPNN)模型(添加入云和出云高度的C-BPNN模型与添加雷达反射率因子的Z-BPNN模型)与一种未添加云信息的BPNN模型(记为BPNN0),并对反演结果进行了对比,结果表明:C-BPNN模型和Z-BPNN模型在任何天气下(有云或无云),得到的反演误差都小于BPNN0模型;C-BPNN相较于另外两种模型反演结果具有更高的稳定性。对3种模型各自反演结果最好的个例分析发现,C-BPNN与Z-BPNN模型主要的误差存在于高空无云但是相对湿度却出现跃变的情况,说明神经网络模型对初始权值与阈值较为敏感,因此通过遗传算法(genetic algorithms,GA)对BPNN模型进行优化。经GA优化后的反演结果表明:BPNN0模型与C-BPNN模型具有明显优化效果,而Z-BPNN模型优化效果则不明显。In order to improve the accuracy of the microwave radiometer in detecting the atmospheric profile,two back propagation neural network(BPNN)models with different cloud information(C-BPNN model with cloud inlet and cloud outlet height added and Z-BPNN model with radar reflectivity factor added)were constructed using the ground-based microwave radiometer,millimeter wave cloud radar and radiosonde data provided by US ARM atmospheric observatory.The inversion results were compared with the BP neural network model without cloud information(BPNN0).The results show that the retrieval errors of C-BPNN model and Z-BPNN model in any weather(with or without clouds)are smaller than those of BPNN0 model;C-BPNN inversion results are more stable than the other two models.Through the analysis of the case with the best inversion results of the three models,it is found that the main error of C-BPNN and Z-BPNN models exists in the situation that there is no cloud at high altitude but the relative humidity has a jump,which indicates that the neural network model is sensitive to the initial weight and threshold value.Therefore,the BPNN model is optimized by genetic algorithm(GA).The inversion results after GA optimization show that BPNN0 model and C-BPNN model have obvious optimization effects,however,the optimization effect of Z-BPNN model is not obvious.

关 键 词：地基微波辐射计 毫米波雷达 湿度廓线 反向传播神经网络(BPNN) 遗传算法(GA) 

分 类 号：P407.1[天文地球—大气科学及气象学]