基于定量遥感的中国草地综合顺序分类

Grassland classification is a fundamental need of grassland science. Meanwhile developing a comprehensive grassland classification system is also a challenge for grassland science because of the multivariable and multi-functional feature of grassland system. Comprehensive and Sequential Classification System of Grassland(CSCS), one of well known grassland classification systems in China, is advanced in quantification indicators. However, the result of basic level classification often does not match well with the actual grassland type. This project points out that there are at least two points needing to improve at the basic classification level of CSCS. The first one, grasslands are grouped into classes according to an index of precipitation and temperature, which characterize the near-surface atmosphere hydrothermal conditions instead of the actual habitat of low grass plant. The second one, data of precipitation and temperature coming from ground observation of weather station can only characterize conditions within a small area, but always be extrapolated to be on behalf of feature of the region. Then, suggestions are proposed that areal data of land surface temperature and soil moisture inverted by quantitative remote sensing are introduced to the basic classification level of CSCS to replace parameters of precipitation and atmosphere temperature from ground observation. Firstly, the regular relationship between grassland types and soil hydrothermal conditions is studied, the theoretical basis and classifying principles are set up; and then, a set of classifying indicators is defined and a classifying program is given at the basic classification level；finally, spatial database of grassland classification in study area is established based on the above work. The project aims at increasing the fidelity of parameters; improving the quantitative study level; carrying forward the practical progress of CSCS and deepening the application of remote sensing in grassland classification from appearances to the genetic degree.

草地分类是草地科学研究的基础理论问题。目前草地分类系统多属经验性分类,难以客观、准确地反映草地类型分异的实质。草地综合顺序分类系统(CSCS)在世界草地分类系统中处于领先地位,其重要贡献是在基本单位“类”一级实现了定量分类。但在实际应用中,由于数据限制,出现以点代面,以相邻生境代实际生境的问题,对CSCS科学严谨性和实际推广应用提出了挑战。本研究提出在CSCS“类”的划分中,用面状的、由定量遥感反演的土壤水热数据替换点状的、地面站点的大气水热数据作为分类指标的解决思路。在研究草地类型与土壤水热条件规律性联系的基础上,提出利用土壤水热条件划分草地类型的理论依据,确定分类原则；建立“类”的划分指标标准,拟定分类方案；建立研究区草地分类空间数据库。旨在为增强草地分类指标的确限性、提高草地分类定量化研究水平提供新思路,推进CSCS的实用化进程,将遥感在草地分类中的应用由表相深入到发生学层面。

草地综合顺序分类系统(CSCS)在世界草地分类系统中处于领先地位，但在实际应用中，由于数据限制，出现以点代面、以相邻生境代实际生境的问题，对CSCS科学严谨性和实际推广应用提出了挑战。本研究提出在CSCS“类”的划分中，用面状的、由定量遥感反演的土壤水热数据替换点状的、地面站点的大气水热数据作为分类指标的解决思路。本项目主要进行了以下研究：①土壤水热条件与大气水热条件关系的定量研究；②地表参数遥感反演的质量控制研究；③土壤水分遥感反演模型对比与筛选研究；④基于土壤水热条件的草地类型划分标准研究。取得以下成果：①总结了一套利用土壤水分和地表积温为指标、以综合顺序分类为体系的草地分类方法；②开发了多模型协同土壤水分反演软件，可实现数据质量控制、数据补偿、区域选择、镶嵌、裁剪，根据阈值选择模型进行土壤水分反演并与实测数据拟合、结果可视化等功能；③研究成果还包括2003-2020年1km每日中国草地土壤水分系列图像。相比目前已经发布的土壤水分产品，在空间分辨率和时间分辨率以及时间跨度方面具有优势，可以较为详细地刻画近20年来全球气候变化背景下中国草地土壤水分的变化状况；④研究生成了2003-2020年1km每年中国草地地表积温系列图像。可以提供近20年来全国地表积温时空变化分析的数据，为提高草地研究领域应对全球气候变化的能力提供数据支持；⑤研究最终成果为利用多年平均土壤水分和地表积温划分的中国草地综合顺序分类图，可为草地资源的分类和合理利用提供基础数据。项目研究成果可以提高草地分类的定量化水平，推进草地综合顺序分类法的实际应用进程，并为草地分类研究提供新方法和新思路，也可为合理开发利用草地资源、提高生态敏感区应对气候变化能力等提供支持。

内容获取失败，请点击重试