顾及混合像元的遥感地表温度时空变分融合方法研究

项目名称： 顾及混合像元的遥感地表温度时空变分融合方法研究

项目编号： No.41501376

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： 吴鹏海

作者单位： 安徽大学

项目金额： 20万元

中文摘要： 基于遥感的地表温度是研究地-气间物质与能量交换的重要参数，在天气和气候系统、城市热岛等领域应用广泛。但由于技术设计上的相互制约，卫星反演地表温度产品无法同时满足高时间、高空间分辨率需求。本研究针对传统方法的不足，提出顾及混合像元的遥感地表温度时空变分融合方法：基于多时相的低分辨率温度数据与预先定义端元构建正则化解混函数以获取温度变化端元；在变分框架下将该温度变化端元与高空间分辨率温度数据构建时空融合模型；设计合适的参数求解函数和引入温度相关地理要素以提高模型参数的自适应性和精确性。本项目研究的融合方法，可以提高遥感地表温度数据的时空分辨率，在诸多领域具有重要的应用价值；也可以有效解决现有方法难以克服的技术难题，具有重要的研究价值。

中文关键词： 混合像元；降尺度；时空融合；变分模型；地表温度

英文摘要： Land surface temperature is an important parameter of the matter and energy exchange within the atmosphere and land, and is widely used to weather and climate system, urban heart island, and so on. However, limited by the technical conditions, the LSTs inversed by remote sensing can not meet the need of high spatial resolution and high temporal resolution. In order to improve the defect of traditional methods, this study proposes a spatio-temporal LST variational fusion model considering mixed pixels. Based on multi-temporal coarse resolution LSTs and predefined endmembers, the method extracts endmembers of temperature change by building a regularization unmixing function. Then a spatio-temporal variational fusion model is built based on the endmembers and the high resolution LST. The adaptivily and the accuracy of the model can be improved by designing suitable functions and introducing geographic elements related to LST. The method researched and developed in this program, can improve spatial resolution and temporal resolution for LST, shows important application value in many areas. The proposed method can solves effectively technical problems of the existing methods and has important research value.

英文关键词： Mixed pixel;Downscaling;Spatio-temporal fusion;Variational model ;LST