We consider an aspect of "passive tracer dispersion," which we call "surface adherence." This refers to the fact that a tracer "particle," upon descending "to" the surface (an event we can provisionally define as meaning Z -> 0, where Z is the particle height), must thereafter spend some unknowable duration of time ta "at the surface", before re-emerging into the flow. Thus defined, the "adherence time" ta is a random variable, and its unknown (and until now ignored) probability distribution g(ta) reflects real physical properties of the "ground" and near-ground flow, as well as parameters of our more explicit definition of exactly what we mean by the adherence time ta.
In the context of continuous sources, surface adherence is irrelevant. But dispersion models for transient sources, to the extent that they neglect adherence, must err in respect to "contaminant arrival" and "contaminant gone" times (ie. onset and termination times for observed concentration at a given station). For these times will be biased , too early - and with a greater bias at the trailing edge of the puff.
We investigate the pdf g(ta) of adherence time, for an idealised case of transport over a tall plant canopy, by studying particle travel times using a Lagrangian stochastic trajectory model that alternately tracks trajectories into and out of the canopy (resolving "adherence"), or reflects trajectories at an arbitrary location zr above the canopy (in neither case do we address the additional adherence time ta' in the still-unresolved ground surface beneath the canopy). By comparing travel time with/without the artifice of reflection, we give statistics of ta that are of some generality, due to their having been normalised in terms of the familiar micro-meteorological scaling parameters (friction velocity u* ), canopy properties (eg. height, h) and the arbitrary LS-model parameters (esp. trajectory reflection height, zr).
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner