基于ENDVI-SI3特征空间的盐渍化反演模型及风险评估

Salinization inversion model based on ENDVI-SI3 characteristic space and risk assessment

土壤盐渍化是干旱和半干旱地区面临的最严重环境风险,利用特征参量建立特征空间的遥感方法为土壤盐渍化的及时监测与反演提供了更有效、更经济的工具和技术。目前反演盐渍化的特征参量多选用归一化植被指数(normalized difference vegetation index,NDVI)和盐分指数(salinity index,SI),缺乏精细化分析与地区适用性。以内蒙古乌拉特前旗为研究区,基于Landsat8 OLI数据,选用引入短波红外波段的增强型归一化植被指数(enhanced normalized difference vegetation index,ENDVI)和半干旱区反演效果最优的盐分指数3(salinity index 3,SI3)构建ENDVI-SI3特征空间,建立改进型盐渍化监测指数(improved salinization monitoring index,ISMI)模型。结果表明,ISMI与土壤含盐量相关系数达0.82,反演精度优于NDVI,EDNVI和SI3(-0.66,-0.70和0.75),在ISMI基础上实现了内蒙古乌拉特前旗土壤盐渍化的定量反演分析与风险评估,为半干旱区盐渍化反演特征空间中特征参量的选取提供了优化思路。

Soil salinization is the most severe environmental risk in arid and semi-arid areas.The remote sensing method that constructs a characteristic space based on characteristic parameters provides an effective and economical tool and technique for the timely monitoring and inversion of soil salinization.Presently,the normalized difference vegetation index(NDVI)and the salinity index(SI)are mainly selected as the characteristic parameters for salinization inversion,while refined analysis and regional applicability are lacking.This study investigated Urad Front Banner in Inner Mongolia based on the Landsat8 OLI data.The ENDVI-SI3 characteristic space was constructed using the enhanced normalized difference vegetation index(ENDVI)that introduced the shortwave infrared band and the salinity index 3(SI3)with the best inversion effect for semi-arid areas.Accordingly,the improved salinization monitoring index(ISMI)model was built.The results show that the correlation coefficient between ISMI and soil salt content was up to 0.82,and the inversion precision of the ISMI model was higher than that of NDVI,EDNVI,and SI3(-0.66,-0.70,and 0.75,respectively).Based on the ISMI,this study achieved the quantitative inversion analysis and risk assessment of soil salinization in Urad Front Banner.This study provides an approach for selecting the optimal characteristic parameters of the characteristic space in the salinization inversion of semi-arid areas.