Submitted to: Water Resources Research
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/30/2007
Publication Date: 10/26/2007
Citation: Bertoldi, G., Albertson, J.D., Kustas, W.P., Li, F., Anderson, M.C. 2007. Impact of remotely sensed land surface states on variability in atmospheric forcing and fluxes using Large Eddy Simulation. Water Resources Research. 43, W10433. http://dx.doi.org/10,1029/2007WR005911. Interpretive Summary: Land surface models tend to use high resolution remote sensing data to define key inputs of surface states, but rely on coarse atmospheric inputs, which are typically assumed to be uniform over a landscape. This neglects the dynamic interactions between heterogeneous surface states and the lower atmosphere, which in turn affect the fluxes of mass, energy, and momentum through feedback mechanisms. These interactions are investigated using remotely sensed land surface data coupled to a Large Eddy Simulation (LES) model of the Atmospheric Boundary Layer (ABL). During the 1997 Southern Great Plains (SGP97) field experiment, remotely sensed and land surface flux data were collected over an area characterized by a high contrast in surface temperature, canopy cover, and roughness between grassland and dry bare soil areas. Thee relative role of the wind speed and air temperature variability on surface heat and water vapor fluxes was assessed. The results indicate that often the air temperature and wind speed variations tend to cancel their individual effects, and as a result local variations in air temperature and wind speed and resulting impact on the fluxes are not significant. However, under certain situations where local meteorological conditions might be measured (a bare soil area), these effects can cause significant deviations in heat flux estimation some areas of the landscape. Therefore, a simple conceptual scheme is proposed to evaluate the impact of atmospheric variability on the surface fluxes for other types of land surface contrast conditions.
Technical Abstract: Soil Vegetation Atmosphere Transfer Schemes used for operational land surface flux mapping from remotely sensed data typically use high resolution remote sensed data, but coarser atmospheric inputs, either specified by a model or taken from a nearby weather station. While the role of spatially variable land surface states on the variability of surface fluxes has been widely studied, the scientific problem arises from an incomplete knowledge of how heterogeneous surface states excite heterogeneity in the states of the lower atmosphere, which in turn affect the fluxes of mass, energy, and momentum through a feedback mechanism. The impact of heterogeneous surface states on the heterogeneity in the states of the lower atmosphere is investigated here through the use of the coupled model of Albertson et al. [2001a] for merging remotely sensed land surface data into a Large Eddy Simulation (LES) model of the Atmospheric Boundary Layer (ABL). Simulations have been performed during two clear-sky summer days during the SGP97 field experiment the with different wetness condition over an area characterized by a high contrast in surface temperature, canopy cover, and roughness between grassland and dry bare soil areas. Our focus is to investigate the relative role of the wind speed (U) and air temperature variability (Ta) on the variability of the surface fluxes (H and LE), for a sparsely vegetated environment. Simulations show that a coupled Ta has the effect to decrease H over bare soil areas and to increase H over the highly vegetated riparian zones, and that coupled U has an effect of a clear increase of H over bare soil, but almost no effect over the vegetated areas. It has been observed that, while the feedback mechanism between surface temperature and air temperature acts to limit the spatial variability in the surface fluxes, as found by Kustas and Albertson , a more complex interaction between surface properties and the surface wind field tends to increase the spatial variance of surface fluxes, limiting the combined effect on the total variability of the fluxes. The sensible heat flux H show a negative sensitivity to Ta, both over bare soil and vegetation, but the sensitivity to U variations is much larger over bare soil than over vegetation. Results suggest that, in a situation like this case study where the local variations than Ta and U are less of ±1K and ±1m/s, the error committed in the estimation of the spatially averaged fluxes using a local forcing observation instead of the correct average is quite limited. However, the strong sensitivity of H respect U over bare soil locations can lead to a much more significant local errors in the estimation of H. Finally, a simple conceptual scheme is proposed to evaluate the impact of atmospheric variability on the surface fluxes for other types of land surface contrast.