Prediction of Soil Organic Carbon in a New Target Area by Near-Infrared Spectroscopy: Comparison of the Effects of Spiking in Different Scale Soil Spectral Libraries

Abstract

Near-infrared (NIR) spectroscopy is widely used to predict soil organic carbon (SOC) because it is rapid and accurate under proper calibration. However, the prediction accuracy of the calibration model may be greatly reduced if the soil characteristics of some new target areas are different from the existing soil spectral library (SSL), which greatly limits the application potential of the technology. We attempted to solve the problem by building a large-scale SSL or using the spiking method. A total of 983 soil samples were collected from Zhejiang Province, and three SSLs were built according to geographic scope, representing the provincial, municipal, and district scales. The partial least squares (PLS) algorithm was applied to establish the calibration models based on the three SSLs, and the models were used to predict the SOC of two target areas in Zhejiang Province. The results show that the prediction accuracy of each model was relatively poor regardless of the scale of the SSL (residual predictive deviation (RPD) < 2.5). Then, the Kennard-Stone (KS) algorithm was applied to select 5 or 10 spiking samples from each target area. According to different SSLs and numbers of spiking samples, different spiked models were established by the PLS. The results show that the predictive ability of each model was improved by the spiking method, and the improvement effect was inversely proportional to the scale of the SSL. The spiked models built by combining the district scale SSL and a few spiking samples achieved good prediction of the SOC of two target areas (RPD = 2.72 and 3.13). Therefore, it is possible to accurately measure the SOC of new target areas by building a small-scale SSL with a few spiking samples.

Dataset variability and carbonate concentration influence the performance of local visible-near infrared spectral models

“…Due to the high complexity within spectral libraries, the application of a general model to the local context leads to high prediction errors. Recent research has shown that the localization of these infrared models substantially improves the predictive performance in a local context, for example by spiking (Brown, 2007;Li et al, 2020;Ng et al, 2022;Wetterlind and Stenberg, 2010;Zhao et al, 2021), memory-based learning (Ramirez-Lopez et al, 2013), resampling algorithms (Lobsey et al, 2017) or deep learning (Shen et al, 2022). However, for analyzing small-scale variability (field or farm level), a local model is often still the best choice because it achieves the lowest prediction errors.…”