王远谋,郁凡. 2017. 主要降水与非降水云类的全天时反演试验研究[J]. 气象学报, ():-, doi:10.11676/qxxb2017.041
主要降水与非降水云类的全天时反演试验研究
An Experiment of the Whole-Day Cloud Classification About Principal Rain or No-rain Cloud
投稿时间:2016-08-25  最后修改时间:2016-12-31
DOI:10.11676/qxxb2017.041
中文关键词:  SBDART模式  单位特征空间  云分类  卫星反演
英文关键词:SBDART  UFSM  Cloud Classification  Satellite Inversion
基金项目:41275030; 国家自然科学基金
作者单位E-mail
王远谋 南京大学大气科学学院 njuwym9327@163.com 
郁凡 南京大学大气科学学院 yufan@nju.edu.cn 
摘要点击次数: 243
全文下载次数: 221
中文摘要:
      利用SBDART模式正演了主要云类在多光谱通道上的光谱特性,分析了在由不同通道组成的二维光谱空间中的云类分布;又以MTSAT-2多光谱静止卫星资料与CloudSat云廓线雷达分类产品匹配,建立了包含诸主要云类光谱特性的数据对。在此基础上,采用单位特征空间归类分析法,反演并分析了各典型云类在相应二维光谱空间中的分布范围和高频集群分布特点;结合正、反演结果,应用最小距离分类与最大似然估计方法进行了云分类试验研究。经过单位特征空间方法的调整,最终确定了较为准确的适用于全天时云分类的判据。检验结果表明:根据多通道的云光谱特性区分高、中、低云和对流性及非对流性降水云是可行的,由此能够实现全天时云系生消、移动的连续可靠监测。
英文摘要:
      This paper have done forward-modelings about the spectrum characteristics in multispectral wavebands of principal kinds of cloud, analyzed the distributions in different two-dimensional(2-D) spectral space. We also composed satellite data in multispectral wavebands from Multi-function Transport Satellites (MTSAT-2) and cloud classification products from CloudSat Cloud-Profiling Radar(CPR), to establish data pairs containing the spectrum characteristics of principal kinds of cloud. On the basis, a method named Unit-Feature Spatial Classification Method (UFSM) was introduced to retrieve the distribution range and the location of high-frequency feature points of cloud in the 2-D spectral space. Then two different kinds of statistics methods came out: Minimum Distance Classification and Maximum Likelihood Estimation, which were combined with results from forward-modelings and inversion to be applied in cloud classification experiment. Next, we adjusted the UFSM to bring out comparatively accurate cloud-classification criterions which can be used in whole-day finally. From the results of tests we have got some conclusions: Firstly, it was feasible that high, middle, low cloud, cloud with/without vertical development could be distinguished according to the spectral characteristics, that whole-day cloud’s dissipation and mobile monitoring could also been achieved reliably and continuously.
查看全文   查看/发表评论  下载PDF阅读器
分享按钮