Hydroclimatic sustainability assessment of changing climate on cholera in the Ganges-Brahmaputra basin. (October 2017)

• Record Type: Journal Article
• Authors: Nasr-Azadani, Fariborz; Khan, Rakibul; Rahimikollu, Javad; Unnikrishnan, Avinash; Akanda, Ali; Alam, Munirul; Huq, Anwar; Jutla, Antarpreet; Colwell, Rita
• Is Part Of: Advances in water resources. Volume 108(2017)
• Page Start: 332
• Page End: 344
• Publication Date: 2017-10
• Subjects: Cholera -- Climate change -- Downscaling streamflow -- GBM basin -- Infectious diseases; Hydrology -- Periodicals; Hydrodynamics -- Periodicals; Hydraulic engineering -- Periodicals; 551.48
• DOI: 10.1016/j.advwatres.2016.11.018
• Languages: English
• ISSNs: 0309-1708

Abstract

Highlights: Machine learning algorithms were found to be useful to link GCM outputs with river discharge. Combined low flow of the two rivers may increase in the future. High flows increases for first half of this century, decreasing thereafter. Significant shifts noted in the streamflow suggesting that cholera dynamics may become unpredictable. Abstract: The association of cholera and climate has been extensively documented. However, determining the effects of changing climate on the occurrence of disease remains a challenge. Bimodal peaks of cholera in Bengal Delta are hypothesized to be linked to asymmetric flow of the Ganges and Brahmaputra rivers. Spring cholera is related to intrusion of bacteria-laden coastal seawater during low flow seasons, while autumn cholera results from cross-contamination of water resources when high flows in the rivers cause massive inundation. Coarse resolution of General Circulation Model (GCM) output (usually at 100 – 300 km)cannot be used to evaluate variability at the local scale(10–20 km), hence the goal of this study was to develop a framework that could be used to understand impacts of climate change on occurrence of cholera. Instead of a traditional approach of downscaling precipitation, streamflow of the two rivers was directly linked to GCM outputs, achieving reasonable accuracy (R 2 = 0.89 for the Ganges and R 2 = 0.91 for the Brahmaputra)using machine learning algorithms (Support Vector Regression-Particle Swarm Optimization). Copula methods were used to determine probabilistic risks of cholera under several discharge conditions. Key results, using model outputs from ECHAM5, GFDL, andHadCM3for A1B and A2 scenarios, suggest that the combined low flow of the two rivers may increase in the future, with high flows increasing for first half of this century, decreasing thereafter. Spring and autumn cholera, assuming societal conditions remain constant e.g., at the current rate, may decrease. However significant shifts were noted in the magnitude of river discharge suggesting that cholera dynamics of the delta may well demonstrate an uncertain predictable pattern of occurrence over the next century.