Smart Aquifer Monitoring in Fire-Prone Regions Using Ground-Level Wireless Sensor Networks: A Bolivian Case Study

Abstract

Water management is becoming increasingly critical as the availability of this vital resource declines—a trend driven by human activity and forest fires. In the face of growing water scarcity, understanding aquifer dynamics is essential. Aquifer recharge is typically estimated using soil water balance models; however, these rely on detailed analyses of water distribution processes and are often constrained by uncertainties related to land surface changes and limited data availability. To address these challenges, this study presents the design, development, and preliminary results of a low-cost Internet of Things (IoT) system aimed at supporting aquifer recharge monitoring. The system comprises four end devices that continuously collect hydrological data, including albedo, soil moisture, temperature, and sap flow. Additionally, we implemented anomaly detection models to analyse the collected data in near real time for irregularities and sensor bursts. Initial findings from the system’s in situ deployment in the Chiquitania region of Santa Cruz, Bolivia, demonstrate its potential for continuous monitoring and prompt anomaly detection. This work represents a Phase I prototype with preliminary validation; absolute flux accuracy and species-specific calibration will be addressed in future campaigns.