Groundwater Level Estimation in Northwest Region of Bangladesh Using Hybrid Locally Weighted Linear Regression and Gaussian Process Regression Modeling

The knowledge about the role of the underlying variables on groundwater level (GWL) fluctuation at local scale in the drought-prone urban areas of Bangladesh is still not explored. To better insight into the relative contribution of underlying factors on GWL fluctuation, this study proposed a novel hybrid ensemble modeling framework based on locally weighted linear regression (LWLR) and four Gaussian Process Regressions (GPRs) e.g., poly kernel, Pearson universal kernel (PUK), radian basis function (RBF) and normalized poly kernel. The proposed framework has employed to predict GWL at six wells in the drought-prone local areas of North-western urban region of Bangladesh, where GWL is declining rapidly. The rainfall, temperature (Tave), soil moisture (SM), normalized difference vegetation index (NDVI), Indian Ocean Dipole (IOD), Southern Oscillation Index (SOI), Nina3.4, and population growth rate for the period 1993-2017 were utilized as inputs to developed GWL models. The best input combination was explored using the best subset regression model and sensitivity analysis, and the optimal input combination was applied in LWLR and GPRs to estimate monthly GWL fluctuation. The hybrid LWLR-GPR-PUK model, on average, improves the prediction accuracy from 10 to 50% during the training stage and 20-70% during the testing stage compared to other models. The proposed modeling approach can act as a promising substitute tool to estimate GWL fluctuation, especially in drought-prone local areas in urban regions where groundwater data scarcity hinders the physical law-based model development.