Soil saline-alkalization evaluation basing on spectral reflectance characteristics

The hyperspectral reflectance and soil parameters (organic matter, pH, electrical conductivity( EC)) of saline soils in the west of Jilin province were measured and studied. The first derivate and continuum removal methods were introduced to build soil spectral indices, and the relationship between soil parameters and spectral indives was analyzed to make certain the best spectral bands and spectral indices which could uncover the degree of soil salinity. Then the hyperspectral models were built to evaluate soil saline-alkalization level. The results are as follows: saline soil has five apparent absorption vales in the spectral domain between 400 and 2500nm, and the five vales of vegetation-uncovered saline soil are much larger than the others' because of higher salt content; the correlation between soil parameters of saline soils in different places has the commonness of positive relationship between pH and EC, both of which negatively correlates with OM; the correlation between soil parameters and spectral indices derived from the first derivate and continuum removal methods is much more significant than that between soil parameters and reflectance, and the indices derived from continuum removal can be used more widely. The hyperspectral models based on soil spectral indices can be used to predict pH values of saline soil and evaluate the degree of soil salinity.