Comparison of the Sensible Heat Flux Determined by Large-Aperture Scintillometer and Eddy Covariance Measurements with Respect to the Energy Balance Problem in the Taklimakan Desert.

The large-aperture scintillometer (LAS) has been widely used in the estimation of sensible heat fluxes (H) on various land surfaces, but they are seldom used in deserts with dune topographies. Here, an LAS was deployed in the Taklimakan Desert for about two weeks to investigate its ability to estimate H over the dune topography in the desert and its impact on the surface energy balance. We estimated the LAS and eddy covariance (EC) fluxes using standard methods, as well as H directly calculated by the EC system via a spectral fitting method. We conducted a series of sensitivity tests to investigate the scintillometer-based heat flux estimation uncertainties by considering surface and meteorological parameters and surface-layer similarity structure functions. In addition, the imbalance problem in the surface energy budget was examined using the measurements. Desert areas are different from other land surfaces in that they are characterized by a high sensible heat flux. It was found that the LAS had good performance in the Taklimakan Desert and could capture more low-frequency turbulence. Also, the spectral fitting method used on the data obtained with the EC system could capture more low-frequency turbulence compared to the direct calculations made by the EC system. The H values obtained by the LAS (HLAS) were overestimated by about 11.6% relative to those measured by the EC system (HEC), and the HEC values obtained by the spectral fitting method were overestimated by about 12.1% relative to them being directly calculated by the EC system. Large and dispersed values of HLAS/HEC appeared with smaller values of rwT (the correlation coefficient) (i.e., 0.1–0.3), and the large ratio of HLAS/HEC might be related to a predominance of low-frequency motions in stable stratification. However, the H values estimated by the LAS had great uncertainties in the case of high wind speeds and weak winds at night. The main source of uncertainty in the estimation of H by the LAS was the use of the different surface parameters (e.g., the Bowen ratio, roughness length, and LAS effective height) and the different forms of fT(ζ) from the Monin–Obukhov similarity theory (MOST). Different from the flat surfaces in desert areas, dune topography affected the use of fT(ζ) of MOST by 37%. MOST can improve the estimation of H, and then improve the calculation of the surface energy balance ratio (EBR). [ABSTRACT FROM AUTHOR]