Evaluation of energy balance closure adjustment methods by independent evapotranspiration estimates from lysimeters and hydrological simulations

Abstract

Non‐closure of the surface energy balance is a frequently observed phenomenon of hydrometeorological field measurements, when using the eddy‐covariance method, which can be ascribed to an underestimation of the turbulent fluxes. Several approaches have been proposed in order to adjust the measured fluxes for this apparent systematic error. However, there are uncertainties about partitioning of the energy balance residual between the sensible and latent heat flux and whether such a correction should be applied on 30‐min data or longer time scales. The data for this study originate from two grassland sites in southern Germany, where measurements from weighable lysimeters are available as reference. The adjusted evapotranspiration rates are also compared with joint energy and water balance simulations using a physically based distributed hydrological model. We evaluate two adjustment methods: the first one preserves the Bowen ratio and the correction factor is determined on a daily basis. The second one attributes a smaller portion of the residual energy to the latent heat flux than to the sensible heat flux for closing the energy balance for every 30‐min flux integration interval. Both methods lead to an improved agreement of the eddy‐covariance based fluxes with the independent lysimeter estimates and the physically based model simulations. The first method results in a better comparability of evapotranspiration rates, and the second method leads to a smaller overall bias. These results are similar between both sites despite considerable differences in terrain complexity and grassland management. Moreover, we found that a daily adjustment factor leads to less scatter than a complete partitioning of the residual for every half‐hour time interval. The vertical temperature gradient in the surface layer and friction velocity were identified as important predictors for a potential future parameterisation of the energy balance residual.

Eddy covariance based surface‐atmosphere exchange and crop coefficient determination in a mountainous peatland

“…For all FLUXNET site data, an overall mean energy imbalance of about 20% was observed (Wilson et al, ), signifying that in contrast to the theoretical basis of energy conservation implicit in Equation , the nonturbulent fluxes exceed the turbulent fluxes. This energy balance gap is often associated with energy transport of large scale eddies, which is not quantified accurately with the EC method and a 30‐min averaging interval (Foken, ; Mauder et al, ). Foken () described difficulties of model calibration associated with the energy balance gap.…”

“…Ingwersen, Imukova, Högy, & Streck, ); however, we decided to correct Q E by partitioning the energy balance residual (Δ Q ) according to the Bowen ratio β (Twine et al, ; Equation ) in order to adjust ET a . Using this closure procedure, we assume the scalar similarity of gas and heat exchange not only for small scale but also for large‐scale eddies (Foken, ; Mauder et al, ). The median diurnal course of β was determined from the flux data from July and August and used for the correction for every time of the day as it allows to represent the variation of β through the course of the day and it is robust against measurement uncertainties of individual flux measurements.…”

“…The parameter estimation results for the crop coefficient, K c* (−), and coefficients of determination (R 2 ) as an indication for the goodness are given scalar similarity(Ruppert, Thomas, & Foken, 2006). Nonetheless,Mauder et al (2018) could achieve a better agreement of ET rates from EC experiments with lysimeter ET rates by applying Bowen ratio based energy balance gap correction. Without a correction, the uncorrected K c* estimates had a range of 0.68-0.76, which is implausibly low and its interannual variability implausibly large.…”

The TERENO Pre‐Alpine Observatory: Integrating Meteorological, Hydrological, and Biogeochemical Measurements and Modeling

“…Comprehensive long‐term integrated observations in different compartments (atmosphere, hydrosphere, biosphere) of the terrestrial system are used to set up, further develop, and apply various process models for site‐ and regional‐scale or catchment‐scale simulations. The evaluation of the energy balance closure problem for evapotranspiration estimates (Mauder et al, 2018) and the methodology development to use commercial microwave links for precipitation quantification and improved discharge modeling (Smiatek et al, 2017) are prominent examples. TERENO Pre‐Alpine observations have been used in conjunction with physically based process models to examine the impacts of land cover–management and climate change on ecosystem‐atmosphere cycling of energy (e.g., large‐eddy simulation model PALM; Maronga et al [2015]), water (e.g., WaSiM and GEOtop; Kunstmann et al, 2006), as well as C and N (e.g., LandscapeDNDC; Haas et al [2013]).…”

Observed Key Surface Parameters for Characterizing Land–Atmospheric Interactions in the Northern Marginal Zone of the Taklimakan Desert, China

“…Studies have shown that the surface temperature, outgoing longwave radiation, cooling rates, and frozen surface extent are sensitive to far-infrared surface emissivity [8,9]. z 0m and z 0h are defined as the height at which the wind speed becomes zero and at which the extrapolated air temperature is identical to the surface temperature over a homogeneous surface under neutral and thermally stratified conditions [10][11][12], respectively, and they are very important parameters for estimating the momentum, heat, and mass exchange between the surface and atmosphere [13][14][15][16][17][18][19]. They are generally estimated directly using the eddy covariance method and satellite data inversion technique [20,21].…”