基于FY-3卫星多源遥感资料融合的青藏高原积雪分布研究

It is very important for Tibetan Plateau - atmosphere coupled system effects of global climate change if we get seasonal snow-covered plateau temporal precision characteristic information. In this project, we integrated a variety of satellite remote sensing advantage and study the existing mature snow cover remote sensing technology. The FY-3 satellite multi-source observation data (from the optical to the microwave) was applied to achieve the high spatial and temporal resolution of Tibetan Plateau snow parameters using the joint inversion algorithms and automation comprehensive quantitative inversion technique. It should conduct ground verification to meet batch processing of historical data and the subsequent demand for real-time product. The main studies include: (1) the study of FY-3 data quality control, and the Tibetan Plateau complete remote sensing image generation through seamless image mosaic method; (2) In order to solve the defects of the passive microwave remote sensing low spatial resolution, we study passive microwave remote sensing image un-mixing and its super-resolution reconstruction algorithm; (3) the study of multi-temporal and multi-spectral image fusion for removal of cloud cover on the snow; (4) for the problems in the existed optical and microwave like MODIS and AMSR-E snow equivalent inversion algorithm, the high spatial and temporal resolution parameters of snow cover inversion algorithm for the Tibetan Plateau applying FY-3 multi-source data was studied and established at last. For the ultimate goal, it should provide scientific and technical database which can meet the research and business needs of the plateau snow information for a longer period of time.

获取青藏高原积雪覆盖的高精度时空特性信息，对研究青藏高原地气耦合系统变化及全球气候效应具有重要意义。本项目综合多种卫星遥感观测优势，在现有高原积雪遥感监测技术基础上，以FY-3卫星多源观测资料为研究对象，实现多源观测资料联合反演青藏高原区域高时空分辨率积雪参数算法并进行地面验证，以满足对历史资料批量处理和后续实时产品延续的需求。研究内容包括：研究FY-3资料一级数据产品质量控制，实现青藏高原地区完整遥感图像生成；为解决被动微波遥感低空间分辨率的缺陷，研究被动微波遥感图像的混合像元分解及其超分辨率重建算法；研究多时相多光谱融合的积雪图像云覆盖去除方法；针对已有光学和被动微波遥感的雪水当量反演算法以及积雪雪盖判识算法中存在的问题，研究并建立基于FY-3多源卫星资料的青藏高原区域高时空分辨率的积雪参数定量反演算法及数据产品。为建立更长时段、满足科研与业务需要的高原积雪资料库提供科学技术基础。

获取青藏高原积雪覆盖的高精度时空特性信息，对研究青藏高原地气耦合系统及其全球气候效应具有重要意义。目前国内外主流遥感积雪产品在青藏高原地区的判识精度显著偏低，针对这一问题，本项目综合多源卫星观测优势，在现有高原积雪遥感监测技术基础上，以FY-3卫星资料为主要数据来源，发展了适用于青藏高原复杂地形的积雪遥感判识算法，经实测数据验证结果表明，本项目积雪判识算法可有效提高青藏高原地区的积雪判识精度，并优于国际主流积雪产品。在本项目积雪判识算法的基础上，本研究设计并实现了青藏高原积雪长时间序列自动化批量提取软件系统，测试结果表明该系统具有操作简单、过程可视化、大批量快速处理等优点，可节省大量人力资源，提高工作效率，为积雪遥感动态监测提供良好的技术支持。此外，在使用本项目开发系统获取长时间序列青藏高原积雪覆盖数据集的基础上，深入研究分析了长时间序列青藏高原积雪覆盖时空变化特征，并取得了大量重要科学结论。