Evaluation of Land Surface Scheme SABAE-HW in Simulating Snow Depth, Soil Temperature and Soil Moisture within the BOREAS Site, Saskatchewan

The Soil Atmosphere Boundary, Accurate Evaluation of Heat and Water (SABAE-HW) model is a multilayered, one-dimensional, physically based version of the Canadian Land Surface Scheme (CLASS) and uses the same methodologies as CLASS, version 2.6. SABAE provides an improved interface for groundwater modelling to simulate soil moisture, soil temperature, energy fluxes and snow depth for a wide range of soil and vegetation. This paper reports the results of the first field comparison of SABAE-HW using an extensive ten-year dataset from the Boreal Ecosystem Atmosphere Study (BOREAS) and the Boreal Ecosystem Research and Monitoring Sites (BERMS) project, an area in central Saskatchewan, Canada, rich in terms of hydrological and meteorological data. The model is also independently tested and verified with the Simultaneous Heat and Water (SHAW) model, which is an unsaturated-zone transport model. Two boundary conditions are considered at the bottom of the soil profile: a water table boundary condition and a unit gradient boundary condition. There was substantial agreement between the results of the simulations and observations in terms of snow depth and soil temperature. Snow depth and soil temperature were simulated reasonably well by SABAE, with correlation values of 0.96 and 0.98, respectively. However, there were some discrepancies for simulated soil temperature in winter. General agreement was obtained in terms of unfrozen soil moisture results, especially at greater depths, but there were general similarities in observed and simulated soil moisture trends in winter. An average correlation of 0.55 was found for SABAE while the correlation for SHAW was much smaller (less than 0.30), which indicates a better fit between simulated and field data by SABAE. Although a unit gradient boundary condition does not influence soil moisture, it was found that unit gradient boundary runs resulted in increased bias towards overestimation of the soil temperature. Thus, a safer and more accurate approach, we believe, is to adopt a first type boundary (i.e., water table) condition at the bottom of the domain. This has implications for climate and weather modelling in general. The result of this field testing demonstrated the potential and high accuracy of SABAE-HW as a Canadian model capable of simulating snow depth, snow temperature, soil moisture, energy fluxes, and we believe it is now appropriate to include this land surface scheme with its counterparts.