Studies on soil organic matter content mapping using EO-1 hyperion data

摘要

Soil organic matter content is an important indicator for soil fertility. The use of hyper-spectrum can conduct quantitative inversion on soil organic matter content for precision agriculture. Taking soil types in Jilin province of China as the research object, the author conducted the spectral measurements on treated soil samples under laboratory conditions by using ASD FieldSpec FR, and established the multiple regression model through the correlation analysis between the different variations of the spectral reflectance of the soil samples and soil organic matter content (SOM). SOM distribution map of study area was then obtained by using Hyperion images applied from the regression model through the band calculation module of ENVI software. Results showed that SOM content inversion regression model, which was established by employing the first derivate differential spectral reflectance at the wavelength of 492 nm, 663 nm, 1221 nm, 1317 nm and 2130 nm, possessed the best prediction accuracy. The coefficient of determination R2 was 0.909. The precision of SOM distribution map by using Hyperion images was 0.737. The hyper spectral inversion model of SOM could provide a new approach for the rapid determination of SOM for the precision agriculture.