BUET-IWFM

COMPLETED RESEARCH

GWMap: Applying Machine Learning to map groundwater levels in Bangladesh

Research Funding: Earth Microsoft Azure Compute Grant

Faculty Researcher: Dr Sara Nowreen

Collaboration: Lead: Department of Computer Science and Engineering, BUET; Institute of Water and Flood Management (IWFM), BUET

Duration: July 2019 - December 2021


The overarching goal was to develop a groundwater map of Bangladesh. This was achieved along two dimensions. Firstly, a model was trained to identify the abstraction technology (suction mode vs. force mode) at a given location. Secondly, the Groundwater Level (GWL) at any point in Bangladesh was predicted with reasonable accuracy by training models to capture the intricate relationships between hydrogeological factors and GWL. This work is intended to assist policymakers in planning integrated water resources management for sustainable use. In summary, groundwater abstraction or withdrawal can be restricted where necessary to conserve environmental flow, and appropriate abstraction technologies can be recommended for different regions of Bangladesh.