A spatiotemporal dissolved oxygen prediction model based on graph attention networks suitable for missing data.

The accurate prediction of dissolved oxygen concentration is crucial for the effective prevention and control of water pollution. A spatiotemporal prediction model for dissolved oxygen content that is suitable for missing data is proposed in this study. The model utilizes a module based on neural controlled differential equations (NCDEs) to handle missing data and graph attention networks (GATs) to capture the spatiotemporal relationship of dissolved oxygen content. To enhance the performance of model, it is optimized from three aspects: an iterative optimization method based on the k-nearest neighbor graph is proposed to enhance the quality of graph; Shapley additive explanations model (SHAP) is used to select the main features into model, enabling it to handle multiple features; and a fusion graph attention mechanism is introduced to improve the robustness of model to noise. The model is evaluated using data from water quality monitoring sites in Hunan Province, China, from January 14, 2021, to June 16, 2022. The proposed model outperforms other models in long-term prediction (step = 18), with MAE of 0.194, NSE of 0.914, RAE of 0.219, and IA of 0.977. The results demonstrate that constructing appropriate spatial dependencies can enhance the accuracy of dissolved oxygen prediction models, and the NCDE module confers robustness to missing data in the model.

Infant Exposure to PCBs and PBDEs Revealed by Hair and Human Milk Analysis: Evaluation of Hair as an Alternative Biomatrix.

Human hair, as an emerging biological monitoring matrix, has begun to be used in various human exposure studies, but little research has been done on persistent organic pollutants (POPs), especially for the body burden of POPs in infants. In this study, 36 breast-fed infants in Shanghai were recruited for a study to determine their exposure to POPs, including 12 dioxin-like polychlorinated biphenyls (dl-PCBs), 6 indicator PCBs, and 8 polybrominated diphenyl ethers (PBDEs) in the inner layer (internal) and outer layer (external) of infant hair and human milk. The similarity or difference of the POP distribution pattern or concentration among these matrices was investigated, and only weak correlations (r < 0.4) were observed between the POP concentration in human milk and infant hair (internal or external). POPs in human milk have a different profile than those in infant hair, while they have stable concentration ratios (0.58-2.72), similar distribution patterns, fine Spearman's rank correlations, and tangled principal component analysis (PCA) plots in each POP family between external and internal hair samples. The result suggested that POPs in internal hair can be easily affected by those in external hair, but POPs in human milk seem to have little contribution to the POP profile in internal hair. Although infant hair cannot reflect the POPs from diet or from body burden, it can be an ideal biomatrix that estimates infant exposure to POPs from exogenous sources like house dust when considering the similar pattern of POPs and their proper accumulation period in hair.