Study on Characteristics of Visible and Near Infrared Reflectance Spectra of Soil Organic Matter
To optimize the calculation speed of the soil organic matter(SOM)content prediction model,50 soil samples were used to extract and screen the characteristic spectral bands of SOM by using Pearson correlation coefficient analysis method.The results showed that the reflectance spectra of the soil was strongly negatively correlated with SOM content in the range of 555~662 nm,with a characteristic band at 601 nm.After different preprocessing methods,the reflectance spectra of soil and SOM content showed a strong correlation,and the characteristic bands increased,mainly around 601,1 221,1 410,1 665,1 880,2 110 and 2 200 nm.By analyzing the Pearson correlation curves between the reflectance spectra at different soil moisture levels and SOM content,the characteristic bands were observed around 601,1 450,1 930 and 2 200 nm.With an increase in soil moisture content,the correlation of the characteristic band at 601 nm gradually weakened,while the correlation of the moisture-related characteristic bands at 1 450,1 930 and 2 200 nm increased.When the soil moisture content exceeded 10%,the reflectance spectra of soil and SOM content showed a positive correlation.The identified characteristic bands of SOM and soil moisture provided theoretical support for the development of a rapid detection model for SOM.