Modelling the impact of climate and landscape changes on snow distribution and melt in regions with limited data

The quantity and distribution of snow across landscapes and timing of the spring snowmelt is key to a diverse range of processes, from the hydrological cycle and glaciation through to ecological and human-environment interactions. Many snow-covered landscapes are remote, inaccessible and lack observation data, especially at high resolutions and spanning multi-decadal time periods. Models are therefore valuable tools for understanding and simulating temporal and spatial variations in snow cover. The aim is to determine the most robust method of modelling snow distribution and melt across regional landscapes with limited data availability, and to apply models to understand and project variations in snow cover as a result of landscape and climate change. Physically based, high resolution snow distribution and melt models are tested through fieldwork in Sweden and Norway at research sites with detailed landscape and climate data. The impact of pseudo-limiting input data spatially and temporally on model performance and uncertainty is assessed. Methods of snow model transferral (including parameter estimation and transfer) between areas of different spatial scales and over varying time periods are explored alongside the effects on model uncertainty, with the use of additional field data from research sites in North America and Finland. The impact of variations in topography, vegetation and climate on snow distribution and melt is assessed through both fieldwork and model application, with exploration of the implications for landscape processes and populations.