基于多光谱遥感的耕地土壤有机质定量反演

Quantitative Inversion of Soil Organic Matter in Cultivated Land Based on Multi Spectral Remote Sensing

以重庆市江津区LandSat 8(L8)多光谱遥感影像和地面样点测土配方数据,开展了研究区L8地表反射率与土壤有机质的相关性分析,然后采用逐步回归分析的方法建立了研究区土壤有机质含量的遥感反演模型,最后利用模型计算了研究区土壤有机质含量分布。结果表明:研究区L8各波段地表反射率与土壤有机质含量均呈负相关,双侧显著性检验均<0. 01,其中,第6波段相关系数达到最大值为-0. 840;以第4、5、6和7波段地表反射率建立的土壤有机质含量多元回归反演模型结果较好,模型判定系数R^2=0. 741;研究区土壤有机质含量遥感反演结果表明,研究区南部和北部边缘的土壤有机质含量较高,沿江浅丘阶地和中部深丘谷地的土壤有机质主要为10～30 g/kg,呈现不规律分布。

Relativity between Landsat 8 OLI surface reflectance and ground-measured data of soil organic matter( SOM) content in Jiangjin district of Chongqing was analyzed,then a retrieval model for SOM content was built based on multi-regression analysis method,finally SOM content of study area was estimated by using retrieval model. Results showed that Landsat 8 reflectance of site was negatively correlated with organic content of soil,and bilateral significant test were less than 0. 01,and a maximal correlation coefficient appeared in 6 th waveband,which was-0. 840. In addition,retrieval result of SOM content based on reflectance of band 4,5,6 and 7 showed better performance,with a R^2 of 0. 741. Map of SOM content showed an irregular spatial patch distribution in study area,where southern and northern edges of study area had bigger value,and organic content of soil in shallow hilly terrace along river and central deep hilly valley was mainly between 10 ~ 30 g/kg.