GEOSTATISTICAL/GIS MODELLING OF GROUNDWATER ARSENIC HAZARD AND ATTRIBUTABLE HEALTH RISKS IN CAMBODIA
The risk maps were generated in three steps. Firstly, a hazard map, based on 500 m grid squares, of arsenic in shallow groundwater was developed using regression kriging of 977 data points with environmental covariates, including those derived from soil type, geology, slope, and elevation, as independent variables. Secondly, exposure to groundwater arsenic was estimated through analysis of Cambodian government census data. Finally, estimates of cases of arsenic induced arsenicosis, skin cancer and internal cancers were calculated using the dose-response relationships of Yu [4].
Our model indicates several thousand cases of groundwater-arsenic attributable sequela per annum, supporting the conclusions of previous workers that exposure to groundwater arsenic is a major concern for public health in Cambodia.
There are considerable uncertainties in many of the parameters used in the modeling. In particular there is a relative paucity of exposure data, the applicability of dose-response relationships established for other populations to the Cambodian context remains to be fully tested and there is no comprehensive provision for secular changes in hazard, population or exposure. Nevertheless, this study can be considered as a good framework for rapid health risk assessment on a country and regional scale, even though further work is required to improve the models.
Acknowledgements: This work is an output of the FP6 European Commission funded Asia-Link project ‘CALIBRE: Cambodia and Laos Initiative for Building human Resources for the Environment’ (Contract No. KH/Asia-Link/04 (142-966)). The contents of this document do not necessarily reflect the views of any of the funders.
References: [1] Polya et al. (2005). Min. Mag., 69(5), 807; [2] Lado et al. (2008). Applied Geochemistry, 23, 3010; [3] Winkel et al. (2008) Nature Geoscience, 1, 536 ; [4] Yu, et al.. (2003). Water Resources Research, 39, 17.