Abstract

Predictive statistical models, or Machine Learning, are increasingly used in Economics. Here, we explore the value added Of such models in the case of water quality. Clean water is key for sustainable development. However, large data gaps limit our understanding of global hotspots of water quality and their evolution over time. We use a data-driven approach to provide monthly estimates of surface water quality globally between 1992 and 2010 at a 0.5° spatial resolution. We assess water quality hotspots for six indicators relevant for the Sustainable Development Goals. Poor water quality is a global problem that impacts low- and high-income countries with different pollutants: past economic developments have not solved the problem of water
pollution but have transformed it. Low-income countries are particularly exposed to organic pollution while high-income countries are particularly exposed to high nitrate levels (pollutants of prosperity). Water quality is also directly impacted
by climate variables. A drier climate with higher temperature has a negative impact on water quality. Climate change will therefore further exacerbate water quality problems.

Co-authors : Michelle T. H. van Vliet, , Esha Zevari, , Richard Damania, Jason Russ, Giovanna Ribeiros, Aude-Sophie Rodella, Frederic Mortier

Practical information