Predicting groundwater phosphate levels in coastal multi-aquifers: A geostatistical and data-driven approach

Md Abdullah-Al Mamun; Abu Reza Md Towfiqul Islam; Mst Nazneen Aktar; Md Nashir Uddin; Md Saiful Islam; Subodh Chandra Pal; Aznarul Islam; A B M Mainul Bari; Abubakr M Idris; Venkatramanan Senapathi

doi:10.1016/j.scitotenv.2024.176024

Predicting groundwater phosphate levels in coastal multi-aquifers: A geostatistical and data-driven approach

Sci Total Environ. 2024 Sep 4:953:176024. doi: 10.1016/j.scitotenv.2024.176024. Online ahead of print.

Affiliations

¹ Department of Data Science, Tampere University, Finland.
² Department of Disaster Management, Begum Rokeya University, Rangpur 5400, Bangladesh; Department of Development Studies, Daffodil International University, Dhaka 1216, Bangladesh. Electronic address: towfiq_dm@brur.ac.bd.
³ Department of Disaster Management, Begum Rokeya University, Rangpur 5400, Bangladesh.
⁴ Department of Civil Engineering, Dhaka University of Engineering and Technology, Gazipur, Bangladesh.
⁵ Department of Soil Science, Patuakhali Science and Technology University, Dumki, Patuakhali 8602, Bangladesh.
⁶ Department of Geography, The University of Burdwan, Purba Bardhaman, West Bengal 713104, India.
⁷ Department of Geography, Aliah University, 17 Gorachand Road, Kolkata 700014, India.
⁸ Department of Industrial and Production Engineering, Bangladesh University of Engineering and Technology, Dhaka 1000, Bangladesh.
⁹ Department of Chemistry, College of Science, King Khalid University, Abha 62529, Saudi Arabia.
¹⁰ PG and Research Department of Geology, National College (Autonomous), Tiruchirappalli 620001, Tamil Nadu, India. Electronic address: venkatramanan@nct.ac.in.

PMID: 39241889
DOI: 10.1016/j.scitotenv.2024.176024

Abstract

The groundwater (GW) resource plays a central role in securing water supply in the coastal region of Bangladesh and therefore the future sustainability of this valuable resource is crucial for the area. However, there is limited research on the driving factors and prediction of phosphate concentration in groundwater. In this work, geostatistical modeling, self-organizing maps (SOM) and data-driven algorithms were combined to determine the driving factors and predict GW phosphate content in coastal multi-aquifers in southern Bangladesh. The SOM analysis identified three distinct spatial patterns: K⁺Na⁺pH, Ca²⁺Mg²⁺NO₃^-, and HCO₃^-SO₄^2-PO₄^3-F^-. Four data-driven algorithms, including CatBoost, Gradient Boosting Machine (GBM), Long Short-Term Memory (LSTM), and Support Vector Regression (SVR) were used to predict phosphate concentration in GW using 380 samples and 15 prediction parameters. Forecasting accuracy was evaluated using RMSE, R², RAE, CC, and MAE. Phosphate dissolution and saltwater intrusion, along with phosphorus fertilizers, increase PO₄^3- content in GW. Using input parameters selected by multicollinearity and SOM, the CatBoost model showed exceptional performance in both training (RMSE = 0.002, MAE = 0.001, R² = 0.999, RAE = 0.057, CC = 1.00) and testing (RMSE = 0.001, MAE = 0.002, R² = 0.989, RAE = 0.057, CC = 0.998). Na⁺, K⁺, and Mg²⁺ significantly influenced prediction accuracy. The uncertainty study revealed a low standard error for the CatBoost model, indicating robustness and consistency. Semi-variogram models confirmed that the most influential attributes showed weak dependence, suggesting that agricultural runoff increases the heterogeneity of PO₄^3- distribution in GW. These findings are crucial for developing conservation and strategic plans for sustainable utilization of coastal GW resources.

Keywords: Bangladesh; Coastal region; Data mining; Groundwater resource; Phosphate; Spherical semivariogram.