Department of Natural Resources Engineering, University of Hormozgan, Bandar-Abbas, Hormozgan, Iran.
Department of Electrical and Computer Engineering, University of Hormozgan, Bandar-Abbas, Hormozgan, Iran.
Environ Sci Pollut Res Int. 2021 Aug;28(29):39432-39450. doi: 10.1007/s11356-021-13503-7. Epub 2021 Mar 23.
Understanding the spatial distribution of soil salinity is required to conserve land against degradation and desertification. Against this background, this study is the first attempt to predict soil salinity in the Jaghin basin, in southern Iran, by applying and comparing the performance of four deep learning (DL) models (deep convolutional neural networks-DCNNs, dense connected deep neural networks-DenseDNNs, recurrent neural networks-long short-term memory-RNN-LSTM and recurrent neural networks-gated recurrent unit-RNN-GRU) and six shallow machine learning (ML) models (bagged classification and regression tree-BCART, cforest, cubist, quantile regression with LASSO penalty-QR-LASSO, ridge regression-RR and support vectore machine-SVM). To do this, 49 environmental landsat8-derived variables including digital elevation model (DEM)-extracted covariates, soil-salinity indices, and other variables (e.g., soil order, lithology, land use) were mapped spatially. For assessing the relationships between soil salinity (EC) and factors controlling EC, we collected 319 surficial (0-5 cm depth) soil samples for measuring soil salinity on the basis of electrical conductivity (EC). We then selected the most important features (covariates) controlling soil salinity by applying a MARS model. The performance of the DL and shallow ML models for generating soil salinity spatial maps (SSSMs) was assessed using a Taylor diagram and the Nash Sutcliff coefficient (NSE). Among all 10 predictive models, DL models with NSE ≥ 0.9 (DCNNs was the most accurate model with NSE = 0.96) were selected as the four best models, and performed better than the six shallow ML models with NSE ≤ 0.83 (QR-LASSO was the weakest predictive model with NSE = 0.50). Based on DCNNs-, the values of the EC ranged between 0.67 and 14.73 dS/m, whereas for QR-LASSO the corresponding EC values were 0.37 to 19.6 dS/m. Overall, DL models performed better than shallow ML models for production of the SSSMs and therefore we recommend applying DL models for prediction purposes in environmental sciences.
了解土壤盐分的空间分布对于防止土地退化和沙漠化至关重要。在此背景下，本研究首次尝试应用和比较四种深度学习（DL）模型（深度卷积神经网络-DCNNs、密集连接深度神经网络-DenseDNNs、递归神经网络-长短期记忆-RNN-LSTM 和递归神经网络-门控循环单元-RNN-GRU）和六种浅层机器学习（ML）模型（袋装分类和回归树-BCART、cforest、cubist、带 LASSO 惩罚的分位数回归-QR-LASSO、岭回归-RR 和支持向量机-SVM）来预测伊朗南部 Jaghin 盆地的土壤盐分。为此，我们对 49 个环境 Landsat8 衍生变量（包括数字高程模型（DEM）提取的协变量、土壤盐分指数和其他变量（如土壤类型、岩性、土地利用）进行了空间映射。为了评估土壤盐分（EC）与控制 EC 的因素之间的关系，我们采集了 319 个表层（0-5 厘米深度）土壤样本，基于电导率（EC）测量土壤盐分。然后，我们应用 MARS 模型选择了控制土壤盐分的最重要特征（协变量）。我们通过泰勒图和纳什-苏特克利夫系数（NSE）评估了 DL 和浅层 ML 模型生成土壤盐分空间图（SSSMs）的性能。在所有 10 个预测模型中，DL 模型的 NSE≥0.9（DCNNs 是最准确的模型，NSE=0.96）被选为四个最佳模型，它们的性能优于 NSE≤0.83 的六个浅层 ML 模型（QR-LASSO 是最薄弱的预测模型，NSE=0.50）。基于 DCNNs，EC 值范围在 0.67 到 14.73 dS/m 之间，而对于 QR-LASSO，相应的 EC 值在 0.37 到 19.6 dS/m 之间。总体而言，DL 模型在生成 SSSMs 方面的性能优于浅层 ML 模型，因此我们建议在环境科学中应用 DL 模型进行预测。
