| A backward Lagrangian Stochastic (bLS) model was applied to measure CH4 emissions from a commercial dairy farm. The experimental field site was representative of a typical Canadian dairy farm in terms of animal feeding and management strategies. Methane concentrations were measured using a multi-channel tunable diode laser. Seven open path lasers and reflectors were connected to a single laser beam. This procedure was expected to reduce the error associated with drift between sensors and that associated with calibration of several individual units. It is shown that the bLS model technique can diagnose the 'whole-farm' CH4 emission rate with confidence. Emission estimates made by means of a 'model-free' ratiometric computation using the tracer gas SF6 confirm the good performance of the US model in this situation.; A preliminary study tested the performance of a forward Lagrangian Stochastic (LS) model with respect to plume timing. By reference to a short-range tracer experiment in a neutral, horizontally-uniform surface layer, it has been shown that the LS model replicates well the timing of concentration onset and fade-off due to a transient (point) source. An algorithm to parameterise the 'surface delays' experienced by particles in the unresolved layer beneath the trajectory reflection height zr was applied, but there was found to be negligible advantage in doing so. |
