HydroGeoML

Coupled hydrogeophysical inversion and machine learning for improved estimation of hydrological parameters

Groundwater is a vital source of fresh water, heavily stressed by overuse and climate change. Accurately estimating groundwater recharge is crucial but difficult. This project presents a novel hydrogeophysical approach that combines geoelectrical monitoring, hydrological modeling, and machine learning:

Key points of the project:

• Problem: Increasing stress on groundwater resources and high uncertainty in groundwater recharge estimation.
• Solution: Development of a hydrogeophysical approach based on geoelectrical monitoring, hydrological modeling, and machine learning.
• Methods:
  • Development of a hydrogeophysical inversion approach to estimate hydraulic conductivity and spatially resolved recharge rates.
  • Investigation of machine learning approaches for efficient estimation of hydraulic conductivities.
  • Application of optimized survey techniques to improve spatial resolution.
• Validation: The approaches will be tested and evaluated using synthetic models, laboratory experiments, and real field data.
• Goal: Provision of spatially resolved, accurate estimates of groundwater recharge rates to improve the management of groundwater recharge facilities.

This project aims to develop innovative methods to support the management and optimization of groundwater resources through more precise information on groundwater recharge.