Graph-inverted Ghanaian aquifers under aridification
Description
This dataset supports the study “Graph-inverted Ghanaian aquifers under aridification”. It contains processed borehole, hydrochemical, climate, graph-network and model-ready data used to reconstruct groundwater evolution pathways across northern Ghanaian aquifers under aridification. The study domain covers the Northern, North East, Upper East and Upper West regions of Ghana and includes Voltaian sedimentary aquifers, Birimian fractured basement aquifers, granitoid fractured basement systems and shallow regolith/alluvial groundwater. The dataset includes metadata for 160 boreholes represented as graph nodes, wet- and dry-season hydrochemical observations, regional climate and aridification indicators for 1991–2025, graph edge weights describing hydrochemical similarity and hydrogeological connectivity, and a model feature matrix for classifying aquifer evolution states. The data support analysis of recharge-weathering pathways, evaporative salinity evolution, geogenic fluoride mobilisation, anthropogenic nitrate contamination and ion-exchange transitions. The dataset is intended for hydrochemical facies analysis, groundwater-quality assessment, graph-constrained inverse hydrogeochemical modelling, process classification, uncertainty analysis and reproducibility of the aquifer-evolution framework. Coordinates are anonymized or spatially masked where necessary to protect community water-supply infrastructure while preserving regional-scale analytical reproducibility.
Files
Steps to reproduce
Open Aquifers_Dataset_v1.xlsx or the corresponding CSV files. Use Well_ID to link Wells_Nodes, Hydrochemistry_Seasonal, Graph_Edges and Model_Feature_Matrix. Use Sample_ID to identify individual wet- and dry-season hydrochemical observations. Screen hydrochemical records using Charge_Balance_Error_pct; records within ±5% are suitable for quantitative hydrochemical and inverse-modelling analysis. Use Hydrochemistry_Seasonal for facies classification, ion-ratio analysis, fluoride/nitrate/salinity assessment and seasonal comparison. Use Climate_Aridification to reconstruct rainfall, temperature, PET, NDVI and aridity-index trajectories from 1991 to 2025. Use Graph_Edges to reconstruct the borehole graph, where edge weights combine hydrochemical similarity, spatial proximity, climate-geology similarity and flow-path plausibility. Use Model_Feature_Matrix to reproduce aquifer evolution labels and pathway classifications for recharge-weathering, salinity evolution, fluoride mobilisation, nitrate contamination and ion-exchange transitions. Where PHREEQC inverse modelling is applied, use hydrochemical pairs linked by graph edges and constrain candidate reactions using aquifer type, saturation indices, major-ion chemistry and plausible mineral phases. Compare wet- and dry-season records to evaluate aridification-related hydrochemical displacement.
Institutions
- Ghana Atomic Energy CommissionGreater Accra, Accra