地理学报 ›› 2003, Vol. 58 ›› Issue (4): 494-502.doi: 10.11821/xb200304002

• 环境遥感 • 上一篇    下一篇

基于陆面模式和遥感技术的地表温度比较

高志强,刘纪远   

  1. 中国科学院地理科学与资源研究所,北京 100101
  • 收稿日期:2002-10-23 修回日期:2003-01-06 出版日期:2003-07-25 发布日期:2003-07-25
  • 作者简介:高志强 (1966-), 男, 副研, 主要从事遥感应用研究。E-mail:gaozq@igsnrr.ac.cn
  • 基金资助:

    国家自然科学基金项目 (90202002); 973项目资助(2002CB4125)

The Comparison of Land Surface Temperature with CLM and Split Window Retrieving Method

GAO Zhiqiang, LIU Jiyuan   

  1. Institute of Geographic Sciences and Natural Resources Research, CAS, Beijing 100101, China
  • Received:2002-10-23 Revised:2003-01-06 Online:2003-07-25 Published:2003-07-25
  • Supported by:

    National Natural Science Foundation of China, No.90202002; National 973 Program, No.2002CB4125

摘要:

基于我国中部的样带基础上,采用陆面模式 (Common Land Surface Model CLM) 和遥感结合技术,对研究的样带区域的地表温度利用CLM模式进行模拟和利用遥感分裂窗技术进行反演;以观测地表温度为真值,分别比较模拟地表温度和反演地表温度空间分布特征和误差大小,分析模拟方法和反演方法对计算地表温度的可适性和应用范围,为进行大面的地表温度计算提供选择和参考依据。通过研究发现,模式模拟地表温度同实测地表温度分布的大的格局吻合非常的好, 但是因地貌类型及地表覆盖的影响,模拟地表温度对水域的模拟温度误差较大,大于3%以上,对耕地模拟误差偏小,在 (-3)% 以内,对裸地、草地和林地模拟温度吻合非常的好。反演地表温度同观测地表温度相差较大,分裂窗反演方法适合地表覆盖为草地和林地状态的地表温度的计算,裸露和农耕区域反演的地表温度误差特别的大。

关键词: 地表温度, 模式模拟, 分裂窗反演, 遥感

Abstract:

In this paper, the Common Land Surface Model (CLM) and split window retrieving method were employed to compute and retrieve land surface temperature along transects, and then the spatial distribution and errors of two kinds of land surface temperatures from different ways were analyzed and compared based on the observed temperature, the applicability of two ways was decided by the comparison results. The conclusions drawn by this study are as follows: the land surface temperature simulated with CLM matches best with that from observed in distribution, but because of the impacts of different topographic types and land covers the errors of water surface temperature simulated with CLM to land surface temperature are greater (errors > 3%), the errors of cultivated land surface temperature are smaller (errors < -3%), the surface temperatures of barren land, grassland and forest match best with observed values. The land surface temperatures with split window method differ obviously from that observed. The split window method is applicable to retrieve the temperature of grassland and forest, but it has significant errors to retrieve the surface temperatures of barren land and cultivated land.

Key words: temperature, modeling simulatation, split window’s retrieving, remote sensing