Seasonal recharge forcing and hydrogeological controls on groundwater contaminant vulnerability: a process-informed spatial machine-learning framework in the Middle Benue Trough, Nigeria.

<p dir="ltr">This dataset supports the study <b>“Seasonal recharge forcing and hydrogeological controls on groundwater contaminant vulnerability: a process-informed spatial machine-learning framework in the Middle Benue Trough, Nigeria.”</b> It comprises <b>600 groundwater observations from 100 monitoring locations</b> across five Local Government Areas—Awe, Doma, Keana, Lafia and Obi—sampled during six seasonal campaigns in 2024.</p><p dir="ltr">The dataset integrates groundwater hydrochemical measurements with temporally matched environmental predictors, including antecedent precipitation, soil moisture, runoff, hydrogeomorphological variables and land-cover indicators.

The principal contaminants modelled are <b>Mn, Pb, Cd, Cr and As</b>. Associated modelling outputs include grouped and spatial cross-validation results, hierarchical H1–H4 predictor evaluations, explainability metrics, spatial prediction surfaces, and uncertainty-supported benchmark-exceedance products.</p><p dir="ltr">The dataset was developed to support reproducible analysis of seasonal recharge forcing, contaminant-specific groundwater vulnerability, spatial transferability and predictive uncertainty in the Middle Benue Trough.</p>