On the use of field measurements of energy fluxes to evaluate land surface models

Eddy covariance (EC) is the generally preferred technique today for measuring energy and mass fluxes of vegetated surfaces, with many experimental sites now established in the global Fluxnet network. A fundamental problem with EC, violating the principle of energy conservation, is that energy balances determined using EC are generally “unclosed”, with combined sensible and latent heat fluxes commonly underestimating available energy by 20% or more. Despite this lack of energy closure, however, recently published evaluations of land surface models (LSM) indicate that modelers generally use available EC measurements for model validation. It is, however, almost impossible to conduct consistent analyses of LSM performance when the models, which assume energy budget closure, are evaluated against measurements that do not close the energy budget. Our study suggests that measurements of energy fluxes must satisfy the energy budget closure prior to their use in LSM evaluations. Using long-term measurements collected at four sites in North America, we show that the closure issue of the measured energy fluxes must be resolved, before it is possible to test adequately LSM simulations of seasonal and interannual variability in energy and water exchanges. The ever-increasing application of LSM to a wide range of problems makes it imperative to recognize that model validation must be carried-out using appropriate data.