基于TM/ETM+光谱特征空间的土壤湿度遥感监测

Remote sensing monitoring of soil Moisture on the basis of TM/ETM+ spectral space

基于神东矿区TM/ETM+影像波段3、波段4、波段5和波段7反射率数据,建立二维光谱特征空间,根据土壤湿度在此空间中的分布规律,利用特征空间中任意一点到原点的距离大小表征土壤水分变化状况,提出了不依赖土壤背景线而变化的TM/ETM+土壤湿度监测指数(SMMI),并与实测土壤湿度数据进行相关性分析.结果表明:SMMI能更好地反映地表0～5cm深度土壤湿度,且基于SWIR(波段7)-Nir(波段4)特征空间的SMMI(7,4)(R2=0.541 0)是最优的,而在监测10cm深度土壤湿度方面,基于SWIR(波段7)-SWIR(波段5)特征空间的SMMI(7,5)更有优势.以SMMI(7,4)作为土壤湿度监测指标,评估了神东矿区20a来土壤湿度时空变化特点,结果表明土壤湿度20a来呈上升趋势,这与矿区植被环境改善紧密相关.

The reflectance data of band3, band4, band5 and band7 derived from TM/ETM＋ image in Shendong Mining Area were applied to establish two-dimensional spectral space. According to the distribution characteristics of soil moisture in the spectral space, using the distance from any point of space to the origin was to present the change of soil moisture, and propose the soil moisture monitoring index （SMMI） of TM/ETM＋which did not rely on the change of soil background line. Besides, correlation analysis was carried out for the index and the data of actually measuring soil moisture. The results show that SMMI can better reflect the soil moisture at depth 0-5 cm on the basis of the optimal SMMI（7,4） （R2 =0. 541 0） of SWlR （bandT）-Nir （band4） spectral space. With respect to monitoring the 10 cm soil moisture, SMMI （7,5） obtained from SWIR （bandT）-SWIR（bandS） spectral space has more advan-tages. Based on SMMI （7,4）, the spatio-temporal change features of soil moisture situation in Shendong Mining Area was evaluated. The results indicate that soil moisture gradually rises in twenty years, because of improvement of vegetation environment.