Scale matters: Effects of temporal and spatial data resolution on water scarcity assessments

摘要

Water scarcity can have severe socio-economic and ecological consequences but can be alleviated by adapting water management plans and by increasing preparedness. Such adaptation measures require reliable estimates of potential water scarcity, which are derived at a certain temporal and spatial scale. The choice of scale might affect the outcome of water shortage assessments and conclusions drawn from them. It has been stressed that water scarcity assessments should be performed at a monthly instead of an annual scale, but it has not yet been investigated whether water scarcity assessments could profit from further increasing the temporal resolution. Furthermore, it has been shown that national scale assessments might hide regions potentially affected by water scarcity, and it remains unclear what scale nationwide assessments should focus on. In this study, we therefore investigate the effect of choosing a specific catchment size as the unit of nationwide water scarcity assessments. We compute water supply and demand on a dataset from Switzerland at two spatial scales, a regional scale consisting of large hydrological catchments and a local scale consisting of medium-sized hydrological catchments, and use them to derive water scarcity estimates. Our results show that water scarcity assessments performed at a monthly compared to an annual scale significantly affect the outcomes of water scarcity assessments, while a sub-monthly resolution does not lead to very different results than monthly assessments. In addition, our results also demonstrate that water scarcity can be over- or underestimated at a regional scale compared to at a local scale, at both a monthly and an annual scale and for both normal and extreme discharge conditions. These differences between the local and regional scale estimates can be explained by the within-region variability of local catchment characteristics, i.e. differences between local and regional estimates are larger in regions with diverse local catchment characteristics. We conclude that regional-scale assessments might not be sufficient in regions with highly variable physiographical and meteorological catchment characteristics and/or differences in water demand, infrastructure, and management. Therefore, water scarcity assessments should be conducted at a scale where the variability in these factors is small.