Two-source energy balance model implementation in the Alaska Arctic tundra

Abstract:

Evaporation and transpiration are the two main processes involved in water transfer from vegetated and non-vegetated areas to the atmosphere. Evapotranspiration (ET) from the Earth’s vegetation constitutes 88% of the total terrestrial ET, and returns more than 50% of terrestrial precipitation to the atmosphere (Oki and Kane, 2006); therefore it plays a key role in both the hydrological cycle and the energy balance of the land surface. In Arctic regions, surface-atmosphere exchanges due to ET are estimated from water balance computations to be about 74% of summer precipitation or 50% of annual precipitation. Even though ET is a significant component of the hydrologic cycle in this region, these bulk estimates do not accurately account for spatial and temporal variability due to vegetation type, topography, etc. (Kane and Yang, 2004).

In this work we present the implementation of the Two-Source Energy Balance method, TSEB (Norman et al., 1995), in two Alaska Arctic tundra settings, as a base-line to retrieve energy fluxes at the regional scale from remote sensing imagery. In order to calibrate and validate the model, four flux towers located at the Imanvait Creek and the Anaktuvuk river were used. The TSEB model mainly requires meteorological inputs as well as land surface temperature (LST) and leaf area index (LAI) data. In this study, TSEB was run from late May to early September from 2008 to 2011 in all sky conditions using half hour intervals of meteorological data from the flux tower, and the LST derived from the four component net radiation instrument. TERRA/AQUA MODIS LAI daily product (MOD15/MYD15) was used as LAI input data. Results show an acceptable agreement between the TSEB model and flux tower data. RMSE obtained in the case of net radiation, latent heat, sensible heat and soil heat fluxes was 12, 51, 60 and 27 W/m2. Further efforts will be focused on the daily energy flux integration through implementation of the DTD model (Norman et al., 2000).