The effect of snow: How to better model ground surface temperatures
Introduction
Through its role as an insulator, snow cover plays a key role in the Arctic climate system by controlling heat exchanges between the atmosphere and the ground surface. This strongly influences the dynamics of the active layer and underlying permafrost (e.g., Goodrich, 1982, Ling and Zhang, 2004, Sazonova and Romanovsky, 2003, Shiklomanov and Nelson, 1999, Zhang, 2005). Warming and thawing of near surface permafrost have already been observed in parts of the Arctic, and permafrost degradation is expected to continue through the 21st century, with significant impacts on infrastructure and ecosystems (Callaghan et al., 2011, Instanes and Anisimov, 2008, Oberman, 2008). A number of studies point to the potential for a significant positive climate feedback related to carbon release from thawing permafrost (Grosse et al., 2011, Schaefer et al., 2011). Improving the veracity of projected changes in the active layer and permafrost conditions requires better parameterization of snow thermal properties, in particular, snow thermal conductivity. While often assumed to be constant throughout the entire snow season, in reality snow thermal conductivity depends on many factors, such as air and snow temperatures, snow density, and grain structures, and hence varies with time and position within the snow layer (Sturm et al., 1997).
Past studies of snow thermal conductivity have made use of field observations, laboratory experiments, and theoretical frameworks (e.g., Abel, 1893, Fukusako, 1990, Mellor, 1977, Pitman and Zuckerman, 1967, Sturm et al., 1997, Sturm et al., 2002, Yen, 1962). Brun et al. (2013) and Domine et al. (2013) recently employed physically-based approaches making use of outputs from atmospheric reanalyses to simulate snowpack properties and soil temperatures.
Here, we present an approach to simulate snow “effective” thermal conductivities on a daily basis using the Geophysical Institute Permafrost Laboratory (GIPL) numerical transient model (Jafarov et al., 2012, Nicolsky et al., 2007, Sergueev et al., 2003). Effective conductivity (hereafter simply referred to as conductivity) includes the combined effect of conduction through the ice grains, conduction through the air in the void spaces, and radiative exchange across the void spaces (Anderson, 1976, Marks and Dozier, 1992, Morin et al., 2010). We use an inverse modeling approach originally introduced by Tipenko and Romanovsky (2002), and similar to that used by Sergienko et al. (2008).
The GIPL model simulates ground temperatures and the seasonal freeze/thaw layer dynamics, and has been successfully validated against ground temperature measurements in shallow boreholes across Alaska (Jafarov et al., 2013, Nicolsky et al., 2009, Romanovsky and Osterkamp, 2000). The model incorporates the effects of air temperature, snow, soil moisture and multi-layered soil thermal properties (Nicolsky et al., 2007, Sergueev et al., 2003). We use as model input measurements collected from four permafrost monitoring stations in Alaska, including snow depth, air temperatures and ground temperatures and obtain time series of snow thermal conductivity over the entire snow season. We show that the obtained thermal conductivity values improve the simulation of the ground surface temperature dynamics for the entire snow season.
Throughout the manuscript, we use the terms “estimated” and “reconstructed” interchangeably. The term “estimated snow conductivity” refers to the obtained snow conductivity as a result of using the inverse method, whereas the term “reconstructed” refers to the estimated snow conductivity after a moving-average filter is applied. The estimated and reconstructed values are marked by “ˆ” and “ˇ” signs, correspondingly.
Section snippets
Physical model
The GIPL model solves the 1-D heat equation with phase changes (Carslaw and Jaeger, 1959):where T (x,t) is the temperature and L [Jm− 3] is the volumetric latent heat fusion of water. Here, t stands for time and x ∈ (xu,xl) is the spatial variable with the ground surface at x = 0. The upper boundary xu = xu (t) depends on time in order to track the evolution of snow cover. The quantity xu (t) is equal to the snow cover depth when snow is present, or is zero
Method evaluation
To evaluate the performance of the proposed method, we use the data measured at the DH site during the winter of 2009–2010. The data include high-precision measurements of snow depth, air and ground temperature. After estimating the soil thermal properties in the upper 1 m of the soil column at this station, we consider several numerical experiments by using the measured air temperatures and snow depths. In particular, we calculate the ground surface temperature for three different cases.
In the
Validation
The thermal conductivities obtained at the DH station are large in the middle of the snow season > 0.4 Wm− 1 K− 1 (see Fig. 6). We hypothesize that there are two primary reasons for these large values. First, there maybe a problem with the observations. Ground observations reveal that the soil at the DH site is highly saturated and water commonly pools on the ground surface in summer. When the water pool freezes, a thin ice layer forms and embeds a temperature sensor at the ground surface.
Conclusions
The objective of this study was to recover thermal properties of snow based on field observations of temperature and snow depth, and hence to improve modeling of ground temperature dynamics when no formulation of snow properties is available. We show that by using the time varying thermal conductivity we can improve the quality of ground temperature simulations when compared with typical approaches that use constant values for snow conductivity. The estimated snow thermal conductivities have
Acknowledgments
We thank M. Sturm for sharing his data with us and for his advice, critiques and reassurances along the way. We thank R. Daanen for his help with finding validation datasets. We also thank J. Stroh and S. Higgins for their valuable comments and corrections. This research was funded by the State of Alaska and by the National Science Foundation under grant ARC-0856864. Support was also provided by the U.S. Geological Survey, Alaska Climate Science Center and by the Arctic, Western Alaska, and
References (60)
- et al.
A physical snowpack model for the Swiss avalanche warning: Part I: numerical model
Cold Reg. Sci. Technol.
(2002) - et al.
A numerical model for surface energy balance and thermal regime of the active layer and permafrost containing unfrozen water
Cold Reg. Sci. Technol.
(2004) - et al.
In-situ monitoring of the time evolution of the effective thermal conductivity of snow
Cold Reg. Sci. Technol.
(2010) - et al.
Estimation of soil thermal properties using in-situ temperature measurements in the active layer and permafrost
Cold Reg. Sci. Technol.
(2009) - et al.
Frost growth parameters in a forced air stream
Int. J. Heat Mass Transfer
(1991) - et al.
Analytic representation of the active layer thickness field, Kuparuk River Basin, Alaska
Ecol. Model.
(1999) Daily variations of temperature in snow and the relation between the thermal conductivity of snow and its density
Meteorol. Vestn.
(1893)- et al.
Parameter Estimation in Engineering and Science
(1977) - et al.
Simulation of northern Eurasian local snow depth, mass and density using a detailed snowpack model and meteorological reanalyses
J. Hydrometeorol.
(2013)
Changing permafrost and its impacts
Numerical and experimental investigations of the effective thermal conductivity of snow
Geophys. Res. Lett.
Conduction of Heat in Solids
Seasonal evolution of snow permeability under equi-temperature and temperature-gradient conditions
Cryosphere Discuss.
A new snow parameterization for the Meteo-France climate model
Clim. Dyn.
Thermophysical properties of ice, snow, and sea ice
Int. J. Thermophys.
The influence of snow cover on the ground thermal regime
Can. Geotech. J.
Vulnerability of high latitude soil carbon in North America to disturbance
J. Geophys. Res. Biogeosci.
Climate change and arctic infrastructure
Numerical modeling of permafrost dynamics in Alaska using a high spatial resolution dataset
Cryosphere Discuss.
The effects of fire on the thermal stability of permafrost in lowland and upland black spruce forests of interior Alaska in a changing climate
Environ. Res. Lett.
Characteristics of the low-level temperature inversion along the Alaskan arctic coast
Int. J. Climatol.
Non-conductive heat transfer associated with frozen soils
Glob. Planet. Chang.
Convergence properties of the Nelder–Mead simplex method in low dimensions
SIAM J. Optim.
The surface energy balance of a polygonal tundra site in northern Siberia – Part 2: winter
Cryosphere
Climate and energy exchange at the snow surface in the Alpine Region of the Sierra Nevada: 2 snow cover energy balance
Water Resour. Res.
Engineering properties of snow
J. Glaciol.
Using in-situ temperature measurements to estimate saturated soil thermal properties by solving a sequence of optimization problems
Cryosphere
Contemporary permafrost degradation of Northern European Russia
Establishing long-term permafrost observatories for active-layer and permafrost investigations in Alaska: 1977–2002
Permafr. Periglac. Process.
Cited by (26)
Response of peat-rich permafrost to a warming climate on the northeast Tibetan Plateau
2021, Agricultural and Forest MeteorologyCitation Excerpt :Numerical models, however, are better and preferable because they can simulate the dynamic changes of the thermal state of permafrost in the past, present, and future. Such models include the Noah land surface model (Pitman, 2003), the CoupModel (Zhou et al., 2013), and also the GIPL 2.0 model (Alexiades, 1992; Jafarov et al., 2012,2014; Marchenko et al., 2008) used in this paper. The above equations are solved by the finite difference method obtained by Godunov splitting (Marchuk, 1975).
Sentinel-1 based soil freeze/thaw estimation in boreal forest environments
2021, Remote Sensing of EnvironmentSnow cover controls seasonally frozen ground regime on the southern edge of Altai Mountains
2021, Agricultural and Forest MeteorologyCitation Excerpt :Comprising a mixture of air, ice, liquid water, and water vapour, snow is a relatively unstable solid material. With a thermal conductivity ranging from <0.1 W/m/°C for less dense, fresh snow to 0.5 W•W/m/°C for a denser, compacted snowpack (Zhang, 2005), snow cover acts as an insulator (Goodrich, 1982; Zhang, 2005), restricting the exchange of heat between the atmosphere and ground (Jafarov et al., 2014; Stieglitz et al., 2003). The high snow surface albedo and considerable latent heat of fusion also influence the energy budget of the surface layer.
Infrared attitude measurement method for spinning projectile under snow background
2020, Infrared Physics and TechnologyCitation Excerpt :Cheng and Liang [4,5] simulated the emission spectrum of snow by Mie theory, and took the influence of snow particle size on emissivity into account. Jafarov et al. [6] put forward an inversion modelling method to get the snow surface temperature more precisely. Pomeroy et al. [7] proposed a relatively simple model to estimate snow surface temperature that is not reliant on prognostic state variables, which provides the necessary parameter basis for calculating the radiation brightness of the snow surface.
Improving snow simulation with more realistic vegetation parameters in a regional climate model in the Tianshan Mountains, Central Asia
2020, Journal of HydrologyCitation Excerpt :Most RCM LSMs neglect this process, which might cause overestimated SWE and a cold bias, particularly in the high-altitude areas. The role of wind compaction and blowing snow also should be considered in the RCM LSMs, which could cause the large snow density and redistribution of snowpack (Fang and Pomeroy, 2009; Jafarov et al., 2014). The accurate forcing data could reduce the bias in snow simulation in the RCM LSMs (Liu et al., 2019; Terzago et al., 2020), but the WRF model downscaling with the ERA-Interim data has a large deviation in this study (Fig. 5).
Numerical modeling of two-dimensional temperature field dynamics across non-deforming ice-wedge polygons
2019, Cold Regions Science and TechnologyCitation Excerpt :Moreover, since deformation of the soil skeleton is assumed to be negligibly small, η is assumed to be constant that depends on the soil type. For the sake of simplicity we approximate the thermal conductivity and the volumetric heat capacity for the snow cover by constants (Goodrich, 1982; Romanovsky and Osterkamp, 2000; Zhuang et al., 2001; Mölders and Romanovsky, 2006) and assume that the volumetric unfrozen water content for snow is zero (Jafarov et al., 2014). To solve the heat conduction Eq. (1), following Zienkiewicz and Taylor (1991) we employ the finite element method (FEM) with a fixed computational grid.