Water resources management and cost efficiency analysis of different scenarios

Water resources management is a complex issue since it bears many challenges to be resolved and addressed explicitly. Today, we are more concerned than ever to manage earth’s resources in a sustainable way. Naturally uneven distributed water resources have potential to create stress over water-poor areas in the world, and water scarcity seems to be the biggest fear of human kind nowadays. Water quality, in fact, is as important as water quantity. It is actually the first condition to be realized when dealing with the management of water resources since poor quality conditions of water resources affect ecosystem functioning in the habitats of many aquatic species and their surrounding environments as a chain. Once we cannot reach the good water quality in our water bodies we cannot proceed in any way to utilize continuously this vital resource. There have always been water laws in order to comply with the necessary water quality criteria. These laws sometimes stressed discharge limits for the receiving water bodies and, sometimes required treatment of pollutants at the source. With the advent of Water Framework Directive in 2000 the necessity of an integrated and co-ordinated approach while managing water resources became inevitable. The co-ordinated approach requires the formation of river basin districts and following management plans for every district. These plans must involve watershed specific measures which must be tailored according to needs of the land and the water bodies. The indispensable step in this procedure is the identification of prevailing stressors in the undertaken area. In other words, right diagnosis is essential for the right treatment. Water Framework Directive adopts a holistic approach towards the sustainable management of all types of inland waters, and coastal waters. The very crucial and important phase of this management cycle is the identification of pressures and stressors that are acting directly or indirectly on the status of water resources and, development of appropriate measures for the purposes of full or partial restoration goals. Development of tools and methods is crucial in order to find out the most effective restoration measure. Especially quantitative computer modelling is one of the most useful tools in terms of projection of future scenarios where the water resources can be maintained in their natural integrity. Scenario analyses provide great help in finding the most cost-effective alternative in proposed restoration proposals. This doctoral dissertation focuses on water resources management in the perspective of developing tools and methods and, measuring the effectiveness and/or cost-effectiveness of various scenarios in an attempt to support decision making process. The research is organized on a case study basis; each case study represents a different water body in terms of geography and physico-chemical properties. The aim was to analyse and understand underlying factors that are affecting the system’s integrity and propose meaningful set of restoration goals. For this purpose we utilized different quantitative models. First case study is a very particular river system in the northern part of Lombardy region. Lambro-Seveso-Olona river system, well known being the most polluted tributary of Italy’s longest river Po; it was modelled through the USEPA QUAL2Kmodel, allowing to consider various scenarios for improving the quality status of the system and to evaluate them in terms of cost and effectiveness.. Second case concerns the Venice Lagoon and its watershed which were analysed through SWAT (Soil and Water Assessment Tool, USDA Model) and Aquatox (USAEPA model) models. While SWAT enabled to apportion the different sources of nutrient load in the watershed, Aquatox enabled to evaluate the trophic status of the Lagoon taking into account the same loads coming from the watershed. In addition, a meta-model was also developed using a neural network approach on the basis of SWAT outputs. The aim was to develop a neural network capable to provide scenario predictions as accurate as the SWAT model but obtained, in shorter time thanks to both the higher computational efficiency and the quickest scenario implementation. Finally, in the last case study multivariate statistical methods were used in order to disentangle the multiple pressures and stresses, affecting the water resources of whole Lombardy region.