Spatial urban water demand forecasting using Multi-criteria analysis and GIS: A case study of Edinburgh, Scotland

Abstract

The planning of future drinking water supply is one of the most vital aspects of water resource management. Accurate forecasts of future water demand allows water utility companies to manage their assets and ensure the capability of the network to meet consumer expectations. Common forecasting methods often involve quantitative temporal demand analysis and future trends based on population and household projections; however, the spatial variability pattern of these estimations is often disregarded. In this dissertation, a Spatial Multi-Criteria analysis method is employed to model the local future water demand in the city of Edinburgh, based on the forecasting of future urban growth. Different criteria were selected and processed in a Geographic Information System environment. The factors and constraints identified were used within the Analytical Hierarchy Process framework to identify which areas have more probability of experiencing urban development, and subsequently an increase in water demand can be expected. According to their potential for that purpose, different priority classes were allocated. A Spatial Future Water Demand Priority map was produced, identifying the areas where the likelihood of demand increase is higher, which could be used for asset planning purposes. A validation method using previous spatial demand variation rates was used to provide an indicative accuracy measure of the model.