塔里木河流域绿洲土壤盐分遥感反演模型研究

Remote Sensing Retrieval Model of Soil Salinity in Oasis of Tarim River Basin

以塔里木河流域阿拉尔—沙雅段为研究区,通过多光谱指数组转换和指标筛选,基于最小二乘原理和多元二次模型比较不同光谱指标与土壤电导率之间的相关性,构建基于光谱指数组的土壤盐渍化监测模型。结果表明:生长期NDVI与土壤电导率相关性最好,相关系数为-0.783,休耕期各光谱指标与土壤电导率相关性明显低于生长期。光谱指标的不同数学变换可以一定程度改变与土壤电导率的相关性,NDVI、MSAVI(1/R)、多元二次模型与土壤电导率成极显著相关关系,决定系数R^(2)分别为0.8328、0.8096、0.6790,且模型验证精度较高,可用于研究区土壤盐渍化监测研究。

Taking the Alar-Shaya section of Tarim River Basin as the study area,this paper constructed a soil salinization monitoring model based on spectral index group by transforming and selecting multi-spectral index group,comparing the correlation between different spectral indexes and soil electrical conductivity based on the least square principle and multiple quadratic models.The results showed that the correlation between NDVI and soil conductivity was the best in the growing period,and the correlation coefficient was-0.783.The correlation between the spectral indexes and soil conductivity in the fallow period was significantly lower than that in the growing period.Different mathematical transformations of spectral indexes can change the correlation with soil conductivity to a certain extent.NDVI,MSAVI(1/R)and multiple quadratic models are significantly correlated with soil conductivity,and the determination coefficient R^( 2) is 0.8328,0.8096 and 0.6790,respectively.Moreover,the model verification accuracy is high.It can be used for soil salinization monitoring in the study area.