项目名称: 基于双频双极化微波链路的降水类型识别与反演
项目编号: No.41505135
项目类型: 青年科学基金项目
立项/批准年度: 2016
项目学科: 天文学、地球科学
项目作者: 刘西川
作者单位: 中国人民解放军国防科技大学
项目金额: 21万元
中文摘要: 利用微波链路在近地面大气中传播的衰减、极化等效应进行降水测量是近年来出现的降水测量新方法,是现有降水测量手段的一种有效补充手段,对于提高灾害性降水的准确监测和预警能力具有重要意义。但是现有单频单极化微波链路对降水类型的识别能力不足,直接限制了降水强度反演精度的进一步提高。为了解决这一问题,本项目在前期研究的基础上,提出了基于双频双极化微波链路的降水类型识别与反演方法。从雨、雪、雨夹雪等不同类型降水对微波传播的影响出发,研究建立时间序列、功率谱、梯度谱等不同函数空间上的微波传播特征提取算法,建立相应特征库;研究特定频段微波传播特征的相关性,建立基于Fisher线性判别分析的双频与双极化微波链路联合识别和反演降水的模型。深入挖掘微波链路测量降水技术的潜力,在利用微波链路进行降水识别与反演方面取得创新和突破。
中文关键词: 自动识别;降水类型;微波链路;反演
英文摘要: Microwave links is a new method for precipitation measurement in recent years. it can measure the precipitation by the attenuation and polarization of microwaves travelling in the atmosphere near the surface, which is a promising complement to existing precipitation measurements. It has important implications for the monitoring and warning of severe precipitation. However, the existing single-frequency and single-polarization microwave links can not identify the precipitation types, which limit the improvement of the inversion accuracy of precipitation intensity. To address this problem, we propose a precipitation identification and inversion method by dual-frequencies and dual-polarizations microwave links based on our previous work. Starting from the effect of different precipitation (rain, snow, mixed, and etc.) on the microwave propagation, investigate the feature extraction algorithms of time series, power spectrum, and gradient spectrum, and establish the corresponding feature library. Studies the correlation relationship of features at specific frequencies, at last this project will establish the model of precipitation types identification and inversion by the dual-frequencies and dual-polarizations microwave links based on the Fisher linear discriminant analysis. This method will develop the potential of microwave links, and realize a breakthrough and innovation for precipitation measurement and inversion by microwave links.
英文关键词: Automatic indentification;Precipitation types;Microwave links;Inversion