Visible-near infrared reflectance spectroscopy for estimating soil total nitrogen contents in the Sanjiang Yuan Regions, China: A case study of Yushu county and Maduo county,Qinghai province

摘要

The objective of this paper was to evaluate the prediction ability of visible and near-infrared reflectance spectroscopy(VNIRP) to estimate soil total nitrogen concentration (TN) in the Sanjiang Yuan regions, Qinghai province, China by using partial least squares regression (PLSR) method. A total of 149 soil samples (0-30cm) were collected respectively from Yushu county and Maduo county in August 2012. The soil total nitrogen concentrations were measured using Vario EL Ⅲ elemental analyzer (Germany Elementar Inc), and the soil spectral reflectance of with spectral range 350-2500nm was measured using ASD FieldSpec 4 in the laboratory. PLSR model was used to relate the reflectance spectra and its five different pre-processing transformations including the first order derivative reflectance (FDR), the second order derivative reflectance(SDR), the logarithm of the inverse of the reflectance (Log (1/R)), the depth of the band (Band depth) and the first order derivative of the logarithm of reflectance (( Log R)') to the soil total nitrogen contents measured. The results showed that the combination of visible-near infrared reflectance spectroscopy and PLSR to estimate TN produced a good prediction ability with R~2 ＞ 0.76 and RPD ＞ 2.01 in the Sanjiang Yuan regions. The best prediction ability was obtained by (Log R)' with R~2_(CV) = 0.87, R~2_V =0.88 and RPD ＞2.67, followed by Log (1/R), Band Depth, SDR, FDR and Reflectance. Visible and near infrared spectroscopy provided a rapid and effective method for estimating soil TN from five different soil types in the Sanjiang Yuan Regions, Qinghai province.