Spatial Vulnerability Assessments for Water Resources Management - Cases from major Asian river basins with a focus on spatial unit of analysis and the use of big and open data

This dissertation looks at the use of spatial analysis with big and open data for water-related vulnerability assessments in major river basins of Monsoon Asia. Special focus is on the spatial unit of analysis by exploring various ways to define it and by examining systematically the related Modifiable Areal Unit Problem (MAUP). The extent and availability of spatial data have grown rapidly. This big and open spatial data, when combined, mapped and analysed, increases our understanding of interlinked issues and provides support for decision-making. However, the seemingly transparent way of map overlay and zonal analysis require closer examination. This is particularly important, when GIS and spatial analysis are applied for water resources management, which involves actors, values, and demands from various sectors and drivers of change on multiple scales. In Monsoon Asia (covering the area from China to eastern Afghanistan, with a population of 3.52 billion) the drivers of change include: climate change, population growth, urbanisation and various development pressures. The region has major and transboundary river basins making management of water resources particularly challenging. This dissertation includes four case studies that draw findings from three scales: regional, basin and subbasin. Both data-driven and a priori methods were utilised in defining the spatial unit of analysis and new approaches to finding appropriate spatial units of analysis were developed. Based on the case studies, this dissertation demonstrates that the big and open spatial data is extremely useful for water resources management. Yet, the findings indicate that the scale influences profoundly the applicability and performance of the spatial datasets. Moreover, the spatial unit of analysis through the MAUP has significant influence in the analysis results. A multizonal and multiscale approach was found to minimise the negative effects of MAUP. Through such approach it is possible to find appropriate spatial unit of analysis. The findings reinforce the importance of reporting explicitly the choices and assumptions behind the spatial units of analysis. Classifying spatial data to avoid accumulation of uncertainty and identification of data gaps is strongly recommended. Finally, simplicity should be emphasised when conducting vulnerability assessments to ensure comparability. However, also more complex methods were found to have potential to support the process of analysis. The findings help to develop spatial approaches to vulnerability assessments, and thus, enhance the applicability of big and open data for water resources management.