Does Data Availability Constrain Temperature-Index Snow Models? A Case Study in a Humid Boreal Forest

Temperature-index (TI) models are commonly used to simulate the volume and occurrence of meltwater in snow-fed catchments. TI models have varying levels of complexity but are all based on air temperature observations. The quality and availability of data that drive these models affect their predictive ability, particularly given that they are frequently applied in remote environments. This study investigates the performance of non-calibrated TI models in simulating the subcanopy snow water equivalent (SWE) of a small watershed located in Eastern Canada, for which some distinctive observations were collected. Among three relatively simple TI algorithms, the model that performed the best was selected based on the average percent bias (Pbias of 24%) and root mean square error (RMSE of 100 mm w.e.), and was designated as the base TI model. Then, a series of supplemental tests were conducted in order to quantify the performance gain that resulted from including the following inputs/processes to the base TI model: subcanopy incoming radiation, canopy interception, snow surface temperature, sublimation, and cold content. As a final test, all the above modifications were performed simultaneously. Our results reveal that, with the exception of snow sublimation (Pbias of 5.4%) and snow surface temperature, the variables mentioned above were unable to improve TI models within our sites. It is therefore worth exploring other feasible alternatives to existing TI models in complex forested environments.

Aerodynamic and Radiative Controls on the Snow Surface Temperature

The snow surface temperature (SST) is essential for estimating longwave radiation fluxes from snow. SST can be diagnosed using finescale multilayer snow physics models that track changes in snow properties and internal energy; however, these models are heavily parameterized, have high predictive uncertainty, and require continuous simulation to estimate prognostic state variables. Here, a relatively simple model to estimate SST that is not reliant on prognostic state variables is proposed. The model assumes that the snow surface is poorly connected thermally to the underlying snowpack and largely transparent for most of the shortwave radiation spectrum, such that a snow surface energy balance among only sensible heat, latent heat, longwave radiation, and near-infrared radiation is possible and is called the radiative psychrometric model (RPM). The RPM SST is sensitive to air temperature, humidity, ventilation, and longwave irradiance and is secondarily affected by absorption of near-infrared radiation at the snow surface that was higher where atmospheric deposition of particulates was more likely to be higher. The model was implemented with neutral stability, an implicit windless exchange coefficient, and constant shortwave absorption factors and aerodynamic roughness lengths. It was evaluated against radiative SST measurements from the Canadian Prairies and Rocky Mountains, French Alps, and Bolivian Andes. With optimized and global shortwave absorption and aerodynamic roughness length parameters, the model is shown to accurately predict SST under a wide range of conditions, providing superior predictions when compared to air temperature, dewpoint, or ice bulb calculation approaches.

Modeling the snow surface temperature with a one-layer energy balance snowmelt model

Snow surface temperature is a key control on and result of dynamically coupled energy exchanges at the snow surface. The snow surface temperature is the result of the balance between external forcing (incoming radiation) and energy exchanges above the surface that depend on surface temperature (outgoing longwave radiation and turbulent fluxes) and the transport of energy into the snow by conduction and meltwater influx. Because of the strong insulating properties of snow, thermal gradients in snow packs are large and nonlinear, a fact that has led many to advocate multiple layer snowmelt models over single layer models. In an effort to keep snowmelt modeling simple and parsimonious, the Utah Energy Balance (UEB) snowmelt model used only one layer but allowed the snow surface temperature to be different from the snow average temperature by using an equilibrium gradient parameterization based on the surface energy balance. Although this procedure was considered an improvement over the ordinary single layer snowmelt models, it still resulted in discrepancies between modeled and measured snowpack energy contents. In this paper we evaluate the equilibrium gradient approach, the force-restore approach, and a modified force-restore approach when they are integrated as part of a complete energy and mass balance snowmelt model. The force-restore and modified force-restore approaches have not been incorporated into the UEB in early versions, even though Luce and Tartoton have done work in calculating the energy components using these approaches. In addition, we evaluate a scheme for representing the penetration of a refreezing front in cold periods following melt. We introduce a method to adjust effective conductivity to account for the presence of ground near to a shallow snow surface. These parameterizations were tested against data from the Central Sierra Snow Laboratory, CA, Utah State University experimental farm, UT, and subnivean snow laboratory at Niwot Ridge, CO. These tests compare modeled and measured snow surface temperature, snow energy content, snow water equivalent, and snowmelt outflow. We found that with these refinements the model is able to better represent the snowpack energy balance and internal energy content while still retaining a parsimonious one layer format.

Evaluating the Utah Energy Balance (UEB) snow model in the Noah land-surface model

Noah (version 2.7.1), the community land-surface model (LSM) of National Centers for Environmental Predictions-National Center for Atmospheric Research (NCEP-NCAR), which is widely used to describe the land-surface processes either in stand-alone or in coupled land-atmospheric model systems, is recognized to underestimate snow–water equivalent (SWE). Noah's SWE bias can be attributed to its simple snow sub-model, which does not effectively describe the physical processes during snow accumulation and melt period. To improve SWE simulation in the Noah LSM, the Utah Energy Balance (UEB) snow model is implemented in Noah to test alternate snow surface temperature and snowmelt outflow schemes. Snow surface temperature was estimated using the force-restore method and snowmelt event is regulated by accounting for the internal energy of the snowpack. The modified Noah's SWE simulations are compared with the SWE observed at California's NRCS SNOTEL stations for 7 water years: 2002–2008, while the model's snow surface temperature is verified with observed surface-temperature data at an observation site in Utah. The experiments show that modification in Noah's snow process substantially reduced SWE estimation bias while keeping the simplicity of the Noah LSM. The results suggest that the model did not benefit from the alternate temperature representation but primary improvement can be attributed to the substituted snowmelt process.

Modeling the snow surface temperature with a one-layer energy balance snowmelt model

\\label{sec:abstract} Snow surface temperature is a key control on energy exchanges at the snow surface, particularly net longwave radiation and turbulent energy fluxes. The snow surface temperature is in turn controlled by the balance between various external fluxes and the conductive heat flux, internal to the snowpack. Because of the strong insulating properties of snow, thermal gradients in snow packs are large and nonlinear, a fact that has led many to advocate multiple layer snowmelt models over single layer models. In an effort to keep snowmelt modeling simple and parsimonious, the Utah Energy Balance (UEB) snowmelt model used only one layer but allowed the snow surface temperature to be different from the snow average temperature by using an equilibrium gradient parameterization based on the surface energy balance. Although this procedure was considered an improvement over the ordinary single layer snowmelt models, it still resulted in discrepancies between modeled and measured snowpack energy contents. In this paper we examine the parameterization of snow surface temperature in single layer snowmelt models from the perspective of heat conduction into a semi-infinite medium. We evaluate the equilibrium gradient approach, the force-restore approach, and a modified force-restore approach. In addition, we evaluate a scheme for representing the penetration of a refreezing front in cold periods following melt. We also introduce a method to adjust effective conductivity to account for the presence of ground near to a shallow snow surface. These parameterizations were tested against data from the Central Sierra Snow Laboratory, CA, Utah State University experimental farm, UT, and Subnivean snow laboratory at Niwot Ridge, CO. These tests compare modeled and measured snow surface temperature, snow energy content, snow water equivalent, and snowmelt outflow. We found that with these refinements the model is able to better represent the snowpack energy balance and internal energy content while still retaining a parsimonious one layer format.