Urban water resources allocation and shortage risk mapping with support vector machine method

Abstract

The assessment of urban water shortage risk is crucial to the development of management strategies that enable a more systematic allocation of water resources and effectively reduce societal losses. The present work proposes a comprehensive approach to the thorough evaluation of water shortage risk in an urban basin. We focus on a case study in Jilin City, demonstrating the suitability of the proposed method. The development and utilization of hydric resources in Jilin City is both evaluated and simulated using the MIKE BASIN model, involving hydrological data from 1956 to 2010. Furthermore, this investigation provides a comparison between measured and analog quantity results during the period 2001–2010, yielding an acceptable average relative error of 4.478 %. Five indicators, namely, hazard rate, restorability, vulnerability, recurrence period, and risk level are investigated with the purpose of organizing the assessment system prior to mapping the level of water shortage risk with the SVM model. The data calculated from both the MIKE BASIN model and the emergency conditions are used to formulate five indicators of water shortage assessment. Finally, the level of water shortage risk is determined for six sub-basins in Jilin City. Several schemes are capable of improving the water supply and alleviating the shortage conditions.