Testing groundwater for contaminants such as arsenic or fluoride is time-consuming and costly. Prediction maps indicating the probability of encountering high contaminant concentrations can help in undertaking targeted drinking water surveys.
We have been developing innovative methods for producing such maps using machine learning with groundwater quality point data and geospatial predictor variables (e.g. geology, soil, climate) related to the processes of contaminant accumulation. Such maps have been created for various countries, for example, Burkina Faso, China, India, Pakistan and Vietnam, as well as at the global scale.
The Groundwater Assessment Platform (GAP) includes our country-level and global maps along with maps shared by our partners as well as groundwater quality measurement data and datasets of geology, climate soil and population. In addition to viewing and exploring these maps and datasets, users can upload and share their own data of any contaminant or area of interest. GAP also allows users to create their own maps by providing an easy-to-use statistical modelling framework. GAP includes help and support information (user manual) describing how to navigate and utilize GAP, including uploading data to producing hazard maps.