项目名称: 基于先验信息约束的红外光谱反演污染气体云团方法研究
项目编号: No.41505020
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: 吴军
作者单位: 中国科学院合肥物质科学研究院
项目金额: 20万元
中文摘要: 红外光谱在污染云团遥感探测方面独具优势,然而传统的遥测红外光谱分析方法专注于光谱本身,忽略了对光谱形成和传输过程中已知信息的利用,导致其在复杂背景、厚云团、非等温条件、多种干扰组分并存时难以实现准确的云团信息反演。项目提出基于先验信息约束的红外光谱分析方法,利用辐射传输工具,探索先验信息在红外光谱反演中的作用机理和约束方法,以抑制非目标因素的干扰、增强被动遥测污染云团的能力。拟解决如下问题:(1)基于辐射传输原理,开展背景、目标、测量环境和仪器辐射特性等先验信息变化规律研究;(2)在此基础上,利用最大后验概率模型,开展基于先验信息约束的污染气体云团红外光谱物理反演方法研究;(3)实现多元信息反演方法的验证和优化,拓展被动红外遥测技术的应用范围。基于先验信息约束的红外光谱反演算法,增加了参与反演的信息量,有望进一步提高被动红外遥测技术在污染气体监测中的应用能力。
中文关键词: 先验信息;污染云团;红外光谱;反演;辐射传输
英文摘要: Infrared spectrum have unique advantages in pollutant vapor detection. But traditional analytical methods mainly focused on the observed spectrum, ignored the use of available informations such as spectral structure of normal atmosphere, temperature and pressure conditions etc. Which make accurate pollutant vapor retrieval unavailable under more complicated conditions such as thick cloud, non-isothermal and multi mixture conditions. This project proposed a method using a priori information to constrain the analysis of remotely sensed infrared spectrum. Radiative transfer theories are used to learn about the contributions and constrain method of different a priori informations in retrieval. Three aspects would be discussed:(1)Study the probability distributions of a priori informations concerning background, target, measurement equipment and environment.(2)Develop a maximum posterior based retrieval method which could make use of multiple a priori informations as well as corresponding constrains.(3)Validating and optimizing the retrieval methods to ensure the efficent use of multi a priori informations, and try to strike a blance between reliability of algorithm and a priori information content. By introducing constrained a priori information into the retrieval procedure, this method may break the limitations on pollutant vapor remote sensing.
英文关键词: a priori information;pollutant vapor plume;infrared spectrum;retrieval;radiative transfer