Zha Yannan, Yang Yao
Guangzhou Institute of Technology, Guangzhou, Computer Simulation Research and Development Center, 465 Huanshi East Road, Guangzhou, 510075, China.
Guangdong Provincial Key Laboratory of Agricultural & Rural Pollution Abatement and Environmental Safety, College of Natural Resources and Environment, Joint Institute for Environment & Education, South China Agricultural University, 483 Wushan St., Guangzhou, 510642, China.
Sci Rep. 2024 Jul 17;14(1):16505. doi: 10.1038/s41598-024-67175-7.
Predicting soil heavy metal (HM) content is crucial for monitoring soil quality and ensuring ecological health. However, existing methods often neglect the spatial dependency of data. To address this gap, our study introduces a novel graph neural network (GNN) model, Multi-Scale Attention-based Graph Neural Network for Heavy Metal Prediction (MSA-GNN-HMP). The model integrates multi-scale graph convolutional network (MS-GCN) and attention-based GNN (AGNN) to capture spatial relationships. Using surface soil samples from the Pearl River Basin, we evaluate the MSA-GNN-HMP model against four other models. The experimental results show that the MSA-GNN-HMP model has the best predictive performance for Cd and Pb, with a coefficient of determination (R) of 0.841 for Cd and 0.886 for Pb, and the lowest mean absolute error (MAE) of 0.403 mg kg for Cd and 0.670 mg kg for Pb, as well as the lowest root mean square error (RMSE) of 0.563 mg kgfor Cd and 0.898 mg kg for Pb. In feature importance analysis, latitude and longitude emerged as key factors influencing the heavy metal content. The spatial distribution prediction trend of heavy metal elements by different prediction methods is basically consistent, with the high-value areas of Cd and Pb respectively distributed in the northwest and northeast of the basin center. However, the MSA-GNN-HMP model demonstrates superior detail representation in spatial prediction. MSA-GNN-HMP model has excellent spatial information representation capabilities and can more accurately predict heavy metal content and spatial distribution, providing a new theoretical basis for monitoring, assessing, and managing soil pollution.
预测土壤重金属（HM）含量对于监测土壤质量和确保生态健康至关重要。然而，现有方法往往忽视了数据的空间依赖性。为了弥补这一差距，我们的研究引入了一种新颖的图神经网络（GNN）模型，即基于多尺度注意力的图神经网络重金属预测模型（MSA-GNN-HMP）。该模型集成了多尺度图卷积网络（MS-GCN）和基于注意力的GNN（AGNN）来捕捉空间关系。利用珠江流域的表层土壤样本，我们将MSA-GNN-HMP模型与其他四个模型进行了评估比较。实验结果表明，MSA-GNN-HMP模型对镉（Cd）和铅（Pb）具有最佳的预测性能，镉的决定系数（R）为0.841，铅为0.886，镉的最低平均绝对误差（MAE）为0.403mg/kg，铅为0.670mg/kg，镉的最低均方根误差（RMSE）为0.563mg/kg，铅为0.898mg/kg。在特征重要性分析中，经纬度是影响重金属含量的关键因素。不同预测方法对重金属元素的空间分布预测趋势基本一致，镉和铅的高值区域分别分布在流域中心的西北部和东北部。然而，MSA-GNN-HMP模型在空间预测中表现出卓越的细节表征能力。MSA-GNN-HMP模型具有出色的空间信息表征能力，能够更准确地预测重金属含量和空间分布，为土壤污染的监测、评估和管理提供了新的理论依据。
