Location: Hydrology and Remote Sensing LaboratoryTitle: Evaluation of a two source snow-vegetation energy balance model for estimating surface energy fluxes in a rangeland ecosystem Author
|Kongoli, C - University Of Maryland|
|Kustas, William - Bill|
|Norman, J - University Of Wisconsin|
|Marks, Danny - Danny|
Submitted to: Journal of Hydrometeorology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/18/2013
Publication Date: 2/4/2014
Publication URL: http://handle.nal.usda.gov/10113/60240
Citation: Kongoli, C., Kustas, W.P., Anderson, M.C., Norman, J.M., Alfieri, J.G., Flerchinger, G.N., Marks, D.G. 2014. Evaluation of a two source snow-vegetation energy balance model for estimating surface energy fluxes in a rangeland ecosystem. Journal of Hydrometeorology. 15(1):143-158.
Interpretive Summary: Snow cover is an important Earth surface characteristic because it influences partitioning of the surface radiation, energy, and hydrologic budgets and also is a critical water source for crop production and streamflow sustaining ecosystem health in many water limited regions. A remote sensing-based land surface scheme routinely applied over snow-free surfaces using geostationary weather satellite data over the continental U.S. for monitoring evapotranspiration and drought is modified to account for surface snow melt energy flux and snow masking of vegetation. The modified model for snow is evaluated with field measurements at two sites, a sagebush rangeland and aspen forest ecosystem in southwestern Idaho during the winter. The model is found to be robust in capturing the evolution of surface energy fluxes for both ecosystems during active melt periods. Incorporation of this snow model will allow application to snow covered areas, which currently are masked or removed from the continental product, and permit annual water balance estimates for more accurate water resource assessments, particularly in snow-dominated regions.
Technical Abstract: The utility of a two source snow-vegetation energy balance model for estimating surface energy fluxes is evaluated with field measurements at two sites in a rangeland ecosystem in southwestern Idaho during the winter of 2007: one site dominated by aspen vegetation and the other by sagebrush. Model parameterizations are adopted from the Two-Source Energy Balance (TSEB) modeling scheme, which estimates fluxes from the vegetation and surface substrate separately using remotely sensed measurements of land-surface temperature. Modifications include development of routines to account for surface snow melt energy flux and snow masking of vegetation. The land surface is treated as a composite of snow and vegetation elements with different temperatures, fluxes, and atmospheric coupling. This results in a modified TSEB model for Snow (TSEBS) that can be applied to a wide range of snow-canopy conditions, including partially vegetated and snow-covered surfaces. Directional composite radiometric temperature and fractional vegetation cover are used as diagnostic inputs, allowing inverse application of this energy balance modeling framework for interpretation of remote sensing data across a wide range of spatial scales. Comparisons between modeled and measured surface energy fluxes of net radiation and turbulent heat showed good agreement at both sites, but with better performance over the aspen field site in comparison with the sagebrush site. The TSEBS model was robust in capturing the evolution of surface energy fluxes above aspen and sagebrush canopies during active melt periods. The model behavior was also consistent with previous studies that indicate the occurrence of upward sensible heat fluxes during day-time due to solar heating of vegetation limbs and branches, which often exceeds the downward sensible heat flux driving the snowmelt. However, model simulations over aspen trees showed that the upward sensible heat flux could be reversed for a lower canopy fraction due to the dominance of downward sensible heat flux over snow. This indicates that reliable vegetation or snow cover fraction inputs to the model are needed for estimating fluxes over snow-covered landscapes.