A recent study investigates a novel use of machine learning analyses of satellite-derived Essential Climate Variables (ECVs) to forecast environmental cholera risk in coastal India, home to more than 200 million inhabitants and a region affected by recurrent cholera outbreaks. A machine learning algorithm, trained on clinical cholera data reported in the study region between 2010 and 2018, was used to recognise patterns with the seven satellite-derived ECV datasets and make testable predictions of cholera risk.
With a reliability success rate of 89%, the study demonstrates that chlorophyll-a concentration, sea surface salinity, sea level anomalies and land surface temperature are key predictors of the cholera outbreaks in the region. Furthermore, while previous studies have utilised remotely-sensed ECVs, this study for the first time investigates the use of satellite sea surface salinity data in machine learning analyses of a combination of ECVs to detect the risk of cholera outbreaks.
With over half of the global cholera cases reported by the World Health Organisation between 2010 and 2016 coming from the countries bordering the northern Indian Ocean, the study findings can contribute to improved early-warning systems and mitigation strategies of cholera-risk in affected areas, and ultimately save lives.
Published in the International Journal of Environmental Research and Public Health, the study was carried out by a team of scientists from the European Space Agency Climate Office and the Plymouth Marine Laboratory (PML). The study is part of the international PODCAST project on environmental reservoirs of Vibrio cholerae and climate in the northern Indian Ocean bordering countries led by UK-PML Marie-Fanny Racault co-author of the study. It contributes to the GEO Blue Planet Working Group on water-associated diseases. The Working Group aims to identify benefits, best practices and feasibility of incorporating Earth observation measurements into early-warning systems for water-associated diseases. To achieve this, the working group provides a forum to exchange information, share data and coordinate activities. Find out more about this working group here.
To access the study paper, click here. (open access)
Citation: Campbell, A.M., Racault, M.-F., Goult, S., Laurenson, A., 2020. Cholera Risk: A Machine Learning Approach Applied to Essential Climate Variables. International Journal of Environmental Research and Public Health 17, 9378.