Deriving the sensible heat flux from the air temperature time-series through the flux-variance and the surface renewal methods

<p>Eddy covariance (EC) has become the standard method for determining energy fluxes at the soil-plant-atmosphere interface. However, the cost and complexity of EC often limit its widespread deployment, and therefore, alternatives need to be considered. Here, two alternative methods, flux-variance (FV) and surface renewal (SR), are evaluated in quantifying sensible heat flux at three sites representing agricultural (wheat field, straw and bare soil), agroforestry (pine-switchgrass intercroping) and natural forested wetland (mixed conifer-deciduous wetland forest) systems that span a broad range of canopy height and structural complexity. By considering the position of the sensors with respect to canopy, the measurements at these three sites were carried out in the atmospheric surface layer, roughness layer, and roughness to surface transitional layer, respectively. Since the introduction of FV and SR, several versions of these methods have been proposed, with significantly differing perspectives and assumptions. Until now, the differences between the methods have not been fully standardized or clarified. In principle, both methods require the monitoring of high frequency (e.g. 10 Hz) air temperature variation while some approaches require additional wind velocity measurements. This presentation provides an overview of the FV and SR approaches, including new perspectives as well as identifies the common framework of the methods rather than carrying out their mere comparison. We show that the frequently reported need for the calibration (e.g. against EC) cannot be fully overcome. However, it can be put in a more universal framework where the parameters of both methods requiring calibration are represented by joint physically based parameters such as surface aerodynamic properties rather than similarity constants in the case of FV or the mean volume over the area of the air parcels in the case of SR. After the selection of the most reliable approaches, regression analyses against EC shows that both methods can estimate sensible heat flux with slopes within ±10 % from unity and R<sup>2</sup> >0.9 across all the three sites. The best performance of both FV and SR was at the agricultural field, where the measurements are well in the surface layer while the worst in the case of the tall forest where the measurements are still in the roughness sublayer and the roughness layer depth (with its inherent uncertainty) needs to be taken into account in the calculations. We conclude there may be opportunities to fill gaps in knowledge of ecosystem energy balance at substantial cost-savings in specialized circumstances where EC may not be appropriate using both FV and SR methods.</p><p><strong>Acknowledgement:</strong> This study was conducted with support of SustES - Adaptation strategies for sustainable ecosystem services and food security under adverse environmental conditions (CZ.02.1.01/0.0/0.0/16_019/0000797) and USDA NIFA-AFRI Sustainable Bioenergy Program, 2011-67009-20089, Loblolly pine-switch grass intercropping for sustainable timber and biofuels production in the Southeastern United States.  Funding for AmeriFlux core site US-NC4 (natural forested wetland) was provided by the U.S. Department of Energy’s Office of Science.</p>

Duke Scholars

Author Gabriel G. Katul Civil and Environmental Engineering