Are there biases in borehole databases of weathered basement aquifers affecting their reliability to estimate aquifer productivity?

Abstract Groundwater from weathered basement aquifers (WBAs) is a strategic water resource, but its potential for resource development remains poorly characterized. Meanwhile, groundwater abstraction from WBAs is growing to meet water and food security needs in most sub-Saharan African countries. In this study, the impact of the instructions given to water borehole drillers on the characterization of WBA properties was assessed by modelling WBA short-term productivity using a novel numerically based, stochastic modeling approach. The numerical modeling of 10,000 synthetic WBAs reveals systematic biases in borehole databases due to instructions given to water borehole drillers, such as the discharge target ( Q target ) together with the maximum allowed borehole depth ( Z max ) and the minimum allowed borehole depth ( Z min ). Insufficient drilling depth below the base of the saprolite ( Z ≤ 35 m) leads to undersampling of deeper water-bearing fracture, causing an underestimation of aquifer productivity and fractured-layer thickness (> 10%). These biases persist across discharge targets (0.5–10 m 3 /h) and are exacerbated by shallow drilling. Moreover, traditional borehole-database-processing methodologies, fail to account for instructions given to water borehole drillers, misrepresenting useful aquifer thickness (Lu) by up to 100%. To enhance accurate use of drilling databases, several approaches could be considered: (1) classify boreholes by exploration depth and instantaneous discharge, excluding shallow borehole data (≤ 35 m below the base of saprolite); (2) archive drilling instructions ( Q target , Z max and Z min ) to contextualize data limitations; and (3) adopt robust indicators like specific capacity and depth-dependent yield instead of instantaneous discharge.