Estimating Soil Organic Carbon Using VIS/NIR Spectroscopy with SVMR and SPA Methods

Xiaoting Peng,Wenxiu Gao,Tiezhu Shi,Aihong Song,Yiyun Chen

doi:10.3390/rs6042699

Abstract

With 298 heterogeneous soil samples from Yixing (Jiangsu Province), Zhongxiang and Honghu (Hubei Province), this study aimed to combine a successive projections algorithm (SPA) with a support vector machine regression (SVMR) model (SPA-SVMR model) to improve the estimation accuracy of soil organic carbon (SOC) contents using the laboratory-based visible and near-infrared (VIS/NIR, 350−2500 nm) spectroscopy of soils. The effects of eight spectra pre-processing methods, i.e., Log (1/R), Log (1/R) coupled with Savitzky-Golay (SG) smoothing (Log (1/R) + SG), first derivative with SG smoothing (FD), second derivative with SG smoothing (SD), SG, standard normal variate (SNV), mean center (MC) and multiplicative scatter correction (MSC), on SPA-based informative wavelength selection were explored. The SVMR model (i.e., SVMR without SPA) and SPA-PLSR model (i.e., SPA combined with partial least squares regression (PLSR)) were developed and compared with the SPA-SVMR model in order to evaluate the performance of SPA-SVMR. The results indicated that the variables selected by SPA and their distributions were strongly affected by different pre-processing methods, and SG was the optimal pre-processing method for SPA-SVMR model development; the SPA-SVMR model using SG pre-processing and 28 SPA-selected wavelengths obtained a better result (R2V = 0.73, RMSEV = 2.78 g∙kg−1 and RPDV = 1.89) and outperformed the SVMR model (R2V = 0.72, RMSEV = 2.83 g∙kg−1 and RPDV = 1.86) and the SPA-PLSR model (R2V = 0.62, RMSEV = 3.23 g∙kg−1 and RPDV = 1.63). Most of the spectral bands used by the SPA-SVMR model over the near-infrared region were important wavelengths for SOC content estimation. This study demonstrated that the combination of SPA and SVMR is feasible and reliable for estimating SOC content from the VIS/NIR spectra of soils in regions with multiple soil and land-use types.

Highlights

Soil organic carbon (SOC) has considerable influence on soil quality and plant growth, and it governs various physical, chemical and biological processes in the soil environment [1,2]
This study demonstrated that the combination of successive projections algorithm (SPA) and support vector machine regression (SVMR) is feasible and reliable for estimating soil organic carbon (SOC) content from the Visible and near-infrared (VIS/NIR) spectra of soils in regions with multiple soil and land-use types
Various quantitative methods have been applied to VIS/NIR hyperspectral data to estimate SOC content, such as multiple linear regression (MLR) [9], principal component regression (PCR) [10] and partial least squares regression (PLSR) [9,11]

Summary

Introduction

Soil organic carbon (SOC) has considerable influence on soil quality and plant growth, and it governs various physical, chemical and biological processes in the soil environment [1,2]. A rapid and economical method for estimating SOC content can improve environmental monitoring, modeling and precision agriculture [3,4,5]. Various quantitative methods have been applied to VIS/NIR hyperspectral data to estimate SOC content, such as multiple linear regression (MLR) [9], principal component regression (PCR) [10] and partial least squares regression (PLSR) [9,11]. PLSR, developed by Wold et al [12], has become the most commonly-used calibration method for SOC estimation, because it can successfully model the linear relationship between spectral data and chemical components, especially when multi-dimension and multi-collinearity exist in raw spectra data. Nonlinearity between the spectra data and chemical components often exists due to instrument variations (lamp aging and sensor sensitivity) [13]

Objectives

Methods

Results

Discussion

Conclusion