Testing of a Lagrangian model of dispersion in the surface layer with cattle methane emissions

Results from an experiment measuring methane emissions from a herd of cattle are used to investigate the performance of a backward-Lagrangian stochastic model (distributed under the name WindTrax). The availability of simultaneous mass-budget measurements of the emission rate, together with a unique setup geometry, allow to compare modelled and measured normalised concentration profiles and horizontal flux profiles with five sensor heights, z, and for four horizontal source-sensor distances, x. Simulated emission rates differ typically by 10-20% to those obtained from the mass-budget measurements, which is in agreement with previous tests of the accuracy of WindTrax. Thus, the idealisation of a herd of animals as a homogeneous area source at ground level does not seriously affect the model's applicability to infer emission rates. The profile comparison suggests that WindTrax may overestimate the speed of vertical dispersion. As a consequence, for this experiment an ideal z/x ratio exists where the modelled emission rate is unbiased. Its value is about 0.080 in unstable and 0.067 in stable stratification. Using concentration measurements taken above or below this z/x threshold leads to emission rates that are slightly under- or overestimated, respectively. Simultaneous measurements with an open-path methane laser are compatible with this finding. Possible causes of the apparent overestimate of vertical dispersion rates are discussed, leading to the cautious suggestion that it may stem from the choices for the Kolmogorov constant and/or the normalised dissipation rate in the model, which reflects gaps in our understanding of the atmospheric surface layer. It is argued that this notion does not contradict the earlier results from a number of controlled tracer-release experiments that had been designed to test WindTrax. (C) 2010 Elsevier B.V. All rights reserved.