Radarsat-2四极化影像在河套灌区土壤含盐量反演中的应用

Using Radarsat-2 Four Polarization Images in Salinity Inversion of Hetao Irrigated Soil

以位于河套平原的巴彦淖尔市杭锦后旗陕坝镇为试验区域,利用Radarsat-2四极化数据,配合DEM数据计算得到研究区域的后向散射系数(Sigma Nought)值。通过分析得出Sigma Nought值与土壤含盐量之间的反演关系,并以此为依据对整幅影像进行决策树分类,以确定其盐渍化程度及分布。通过58个点建立的反演模型的决定系数R2=0.753。研究发现此模型的遥感反演效果较好,具有一定的参考和应用价值。

Using Shanba Hangjinhouqi area as the experiment object, Radarsat-2 four polarization data and DEM data were used to calculate the backscattering coefficient （Sigma Nought）. By analyzing Sigma Nought and soil salinity,Decision Trees were used to classify the entire image. Coefficient of determination of the inversion model established by 58 points is 0.753. Remote sensing based on this model was found to be better, with certain values of reference and application.