Description: <DIV STYLE="text-align:Left;"><P><SPAN><SPAN>This dataset consists of a globally normalized map (from 0 to 1) of pixels ranked for their relative importance in delivering clean water to downstream beneficiaries. Water pollution is estimated based on the Human Footprint on Water Quality (HFWQ) index. The HFWQ is a measure of the extent to which water runoff is drawn from contaminating human land uses, both point (urban, roads, mining, oil and gas) and non-point (unprotected cropland, unprotected pasture) sources. The HFWQ is calculated relative to all runoff by cumulating the downstream runoff from polluting and non-polluting land uses and expressing the former runoff as a proportion of the total runoff.Realised provisioning is high where the prevailing climate and land use generate high volumes of clean water which can be used (and reused) by large numbers of downstream users. The greater the downstream population, number of dams and actual water available, the greater the service provided. If there is plenty of water, but no people or dams, then there is no realised service. In this way, not all water provides a direct service, only that water that is accessed and used.All analyses were carried out using the WaterWorld (Mulligan 2013) and Co$ting Nature (Mulligan et al. 2010) ecosystem services assessment tools (www.policysupport.orgMulligan, M. (2013). WaterWorld: a self-parameterising, physically based model for application in data-poor but problem-rich environments globally. Hydrology research 44, 5; 748-769.Mulligan, M.A. Guerry, K. Arkema, K. Bagstad and F. Villa (2010) Capturing and quantifying the flow of ecosystem services in Silvestri S., Kershaw F., (eds.). Framing the flow: Innovative Approaches to Understand, Protect and Value Ecosystem Services Across Linked Habitats. UN Environment Programme World Conservation Monitoring Centre, Cambridge, UK. ISBN 978-92-807-3065-4.</SPAN></SPAN></P></DIV>