Land surface models (LSMs) have been found to poorly simulate Arctic snowpack properties leading to large errors in process representation (e.g. metamorphism), simulation of the ground surface regime and surface energy budgets (Domine et al., 2019). Hydrological and weather forecasts rely heavily on LSMs to accurately portray these key earth system processes to generate reliable forecasts in order to provide water security to large populations, inform water resource management and aid decision making (Liston and Elder, 2006). Evaluating the capacity of LSMs to simulate snowpack properties will improve the reliability of outputs and produce forecasts that will make meaningful impacts to human life.
LSM deficiencies can be identified and evaluated using comparison techniques to field measurements of snowpack properties. Snowpack properties are known to differ over spatial scales and refer to the Specific Surface Area (SSA), density, depth, thermal conductivity and snow water equivalent (SWE). Spatial heterogeneity is an aspect that LSMs do not currently consider, making application of forecasts difficult especially over differing scales (Bokhorst et al., 2016). This project aims to constrain the uncertainty in LSM simulations of Arctic snowpack properties across the differing spatial scales (e.g. pan-arctic, regional, local, point) of Arctic Canada tundra and taiga biomes (Trail Valley Creek and Cambridge Bay). Using improved spatially distributed field measurements of snowpack properties combined with statistical comparison techniques (Leonardini et al., 2020), the accuracy of current LSM simulations (e.g. Canadian operational Soil, Vegetation and Snow (SVS) LSM) can be evaluated. This evaluation will be used to identify model deficiencies which can be used to address uncertainties in the reliability of current hydrological and weather forecasts. Project outputs will aid development of key ONE Planet research themes through an improved understanding of our capacity to model earth system processes that are highly vulnerable to rapid climate change. The overall objective of this research is to constrain the uncertainty in current LSM simulations of snowpack microstructure properties across Arctic tundra and taiga biomes. More specific research objectives are as follows:-
1 Produce a comprehensive dataset of snowpack microstructure properties gathered through well-constrained field measurements that reflect spatial variability across Arctic tundra and taiga biomes (Northern Canada – Trail Valley Creek and Cambridge Bay). 2 Identify and detect deficiencies in the capability of LSMs to simulate snowpack properties through comparison of measured results using statistical indicators. 3 Determine the impact of model deficiencies on process representation, which have important implications for global weather and hydrological forecasting.