Shaanxi Provincial Land Engineering Construction Group Co., Ltd, Xi'an, Shaanxi 710075, China; Institute of Water Resources and Hydro-electric Engineering, Xi'an University of Technology, Xi'an, Shaanxi 710048, China; Key Laboratory of Degraded and Unused Land Consolidation Engineering, the Ministry of Natural Resources, Xi'an, Shaanxi 710075, China.
Institute of Water Resources and Hydro-electric Engineering, Xi'an University of Technology, Xi'an, Shaanxi 710048, China.
Spectrochim Acta A Mol Biomol Spectrosc. 2021 Sep 5;258:119823. doi: 10.1016/j.saa.2021.119823. Epub 2021 Apr 20.
Soil organic matter (SOM) is an important index used to evaluate soil fertility and nutrient availability, and it is also an important component of precision agriculture. In this study, in order to quickly and efficiently estimate the SOM content of farmland soil, we took 190 farmland soil samples in Jingbian County and measureed the SOM content of the samples in the lab and collected the corresponding Vis-NIR spectroscopy data. Based on the six pretreatment methods, a competitive adaptive weighting algorithm (CARS) is used for characteristic wavelength selection. Random forest (RF) regression is used to establish the predictive SOM model. The results indicate that after the CARS algorithm screens the different spectral variables, the optimal variable sets of the seven spectral variables are 15, 40, 30, 23, 20, 26, and 23, respectively. The accuracy of the model is improved after the CARS algorithm screens the different spectral variables. A total of 15 characteristic variables from the 2151 spectral wavelengths were used as the optimal spectral variable subset; RF shortened the training time required during the SOM modeling process and dramatically improved the model's accuracy and predictive ability, and the R of the validation set increased from 0.21 to 0.96, and the RPD increased from 0.46 to 3.02. The RPIQ increased from 1.25 to 4.41. Among the tested models, the CR-RF model produced the best results. The R and RMSE values of the calibration set are 0.91 and 0.49, and the R, RMSE, RPD, and RPIQ values of the validation set are 0.96, 0.51, 3.02, and 4.41, respectively. Accurate prediction of the SOM of the cultivated layer in the study area was realized.
土壤有机质（SOM）是评估土壤肥力和养分供应的重要指标，也是精准农业的重要组成部分。本研究以靖边县 190 个农田土壤样本为研究对象，在实验室测量了土壤样本的 SOM 含量，并采集了相应的可见-近红外光谱数据。基于 6 种预处理方法，采用竞争自适应加权算法（CARS）进行特征波长选择。采用随机森林（RF）回归建立预测 SOM 模型。结果表明：CARS 算法筛选不同光谱变量后，7 个光谱变量的最优变量集分别为 15、40、30、23、20、26、23；CARS 算法筛选不同光谱变量后，模型精度提高。2151 个光谱波长中共选择了 15 个特征变量作为最优光谱变量子集；RF 缩短了 SOM 建模过程中的训练时间，显著提高了模型的准确性和预测能力，验证集的 R2 从 0.21 增加到 0.96，RPD 从 0.46 增加到 3.02，RPIQ 从 1.25 增加到 4.41。在测试的模型中，CR-RF 模型效果最佳，校准集的 R 和 RMSE 值分别为 0.91 和 0.49，验证集的 R、RMSE、RPD 和 RPIQ 值分别为 0.96、0.51、3.02 和 4.41。实现了研究区耕作层 SOM 的准确预测。
