中国农业气象

• 论文 • 上一篇    下一篇

基于NDVI-LST模型对辽宁月尺度土壤水分的反演

张启霖,殷红,纪瑞鹏,武晋雯,张海旭   

  1. 1.沈阳农业大学,沈阳110866;2.中国气象局沈阳大气环境研究所,沈阳110016;3.沈阳市苏家屯区气象局,沈阳110101
  • 收稿日期:2017-02-13 出版日期:2017-11-20 发布日期:2017-11-17
  • 作者简介:张启霖(1992-),满族,硕士生,研究方向:气象灾害遥感研究。E-mail:18804050011@163.com
  • 基金资助:
    国家自然科学基金青年科学基金项目(31600350)

Retrieving on Monthly Soil Moisture in Liaoning Province Based on NDVI-LST Module

ZHANG Qi-lin,YIN Hong,JI Rui-peng,WU Jin-wen,ZHANG Hai-xu   

  1. 1.Shenyang Agricultural University, Shenyang 110866, China;2.Institute of Atmospheric Environment(IAE),CMA, Shenyang 110016;3.Metelogical Bureau of Shenyang Sujiatun District, Shenyang 110101
  • Received:2017-02-13 Online:2017-11-20 Published:2017-11-17

摘要: 利用MODIS卫星数据集中的归一化植被指数NDVI(Normalized Difference Vegetation Index,NDVI)与地表温度LST(Land Surface Temperature,LST)数据建立NDVI-LST模型,对辽宁省2015年5-8月的土壤水分进行反演,建立土壤水分与干边斜率的相关关系,得到该模型反演下的土壤水分。结果表明:(1)该时间段的NDVI-LST实际模型能够形成类似“三角”的特征空间,与理论模型吻合,利用该模型反演的5-8月土壤水分含量与地面实际监测结果相关性较高,除8月外,相关系数均在0.8以上,反演结果空间布局与地面实际土壤水分基本一致;(2)8月土壤水分反演结果不理想,相关系数为0.48,反演和地面实际空间特征差异也较大,其原因是8月NDVI对7月降水极少的响应时间的延后。整体而言,NDVI-LST模型反演土壤水分的试验结果较理想,可为利用卫星遥感手段快速反演辽宁月尺度的土壤水分、干旱灾害防御评估等决策工作提供一种新思路。

关键词: MODIS, NDIV-LST模型, 土壤水分, 特征空间

Abstract: To verify the application of NDVI(Normalized Difference Vegetation Index) and LST(Land surface Temperature)module in Liaoning Province, the NDVI and LST data from MODIS May to August in 2015 were used to establish the module and used as the slope from the module fitting curves to compute the soil moisture. The results showed that: (1) the module of NDIV/LST had the same feature with the theory module which was a triangle module, the soil moisture computed from this module had high correlation coefficient with the measured soil moisture values. The values of correlation coefficient were all above 0.8 except for August, and the spatial distribution of computed soil moisture was same with the measured values except for August. (2) The result of computed soil moisture in August was not ideal, the correlation coefficient was only 0.48, which was possibly due to NDVI had a delayed reflection of July precipitation and the influence of August drought in Liaoning province. Overall, the result form retried NDVI-LST module was ideal. It could provide a new idea for the quick retrieval of soil moisture in Liaoning and provide the decision making for disaster prevention and mitigation.

Key words: MODIS, NDVI-LST module, Soil moisture, Feature space