论文部分内容阅读
选用Landsat 8资料构建了地表温度(Ts)与归一化植被指数(NDVI)的Ts-NDVI特征空间,计算了温度植被干旱指数(TVDI)。利用MODIS温度产品数据和实地野外采样数据进行精度验证确定煤田火区TVDI阈值。通过对遥感影像的地表热异常信息进行定性与定量分析继而对煤田温度异常区边界信息进行挖掘。结果表明:(1)利用野外实测土壤相对含水量进行验证,反演值与实测值的相关系数R~2=0.66,表明干旱指数的反演精度较高,相关性较好。(2)TVDI模型对温度呈现出较高的敏感性,二维散点图集中在1∶1线上,对NDVI的敏感性较低,有利于识别温度异常区。(3)利用MVC最大合成法,建立TVDI-MVC作为精度验证数据,火区面积为5.03 km~2,其中TVDI-SC提取火区精度最大为98.50%,TVDI-SW_2提取火区精度最小为88.98%。可见煤田温度异常区范围较广,潜在的灾情恶化较严重。
The Ts-NDVI feature space of surface temperature (Ts) and normalized vegetation index (NDVI) was constructed using Landsat 8 data, and the temperature vegetation drought index (TVDI) was calculated. Using MODIS Temperature Product Data and Field Field Sampling Data for Accuracy Verification to Determine TVDI Threshold in Coalfield Fire Area. Through the qualitative and quantitative analysis of the surface thermal anomaly information of remote sensing images, the boundary information of the anomalous area of coalfields is excavated. The results showed that: (1) The correlation coefficient R ~ 2 = 0.66 between the inversion and the measured data was validated by field measured soil relative water content, indicating that the inversion accuracy of the drought index is high and the correlation is good. (2) The TVDI model shows a high sensitivity to temperature. The two-dimensional scattergram is concentrated on the 1: 1 line, which is less sensitive to NDVI, which is good for identifying the temperature anomalies. (3) TVDI-MVC was established as accuracy verification data using MVC maximum synthesis method. The area of fire area was 5.03 km ~ 2, TVDI-SC extracted the highest fire area accuracy of 98.50%, TVDI-SW_2 extracted fire area minimum of 88.98 %. Shows a wide range of coal anomalous temperature anomalies, the potential disaster worse.