Flux footprint models are important tools which help one evaluate the "field-of-view" of an eddy-covariance tower and are hence crucial for flux measurement siting. There are many different types of footprint models developed using different methodologies, including Lagrangian techniques, Large Eddy Simulation, using analytical solutions of the advection-diffusion equation, etc. These models have been evolving since they were first introduced in the late 20th century. They initially started out as one-dimensional (along-wind), but parameterizations were further incorporated to include the effect of the cross-wind component as well. Since these models are generally idealized, they have certain limitations and often are not consistent with real-world field settings. For instance, there are several cases of non-homogeneous above-ground source contributors to the flux measurements at the tower, and to our knowledge, this effect has not been accounted for in any existing model. There have been few studies, if any, to have utilized particle trajectory simulations in two-dimensions, and the cross-wind component is often used as an idealized Gaussian spread. For this reason, there is very little understanding of how the presence of directional wind shear and vertical source/sink distributions may impact scalar transport and footprints.
In this study, we are working with Lagrangian simulations to model concentration and flux footprints to account for both these effects, i.e., above-ground point source strengths that may vary with height, and also incorporating the effect of potential directional wind shear by computing particle trajectories in two-dimensions.
Using similar procedures, we are also working on modeling concentration footprints (in addition to the more commonly used flux footprints) so that they can be applied to sites not measuring eddy covariance fluxes but estimating fluxes using different scalar measurements used in equations such as the Penman-Monteith equation, for instance the CIMIS stations in California.
In our Lagrangian particle trajectory simulations, we are also trying to understand how the results might differ by designing a meteorological field using Eulerian data, but temporally lagging it for each adjacent point in space, to achieve a mock Lagrangian velocity field in the atmosphere. We are also testing the effects of the commonly used Monin-Obukhov surface layer parameterizations to study the sensitivity of our model to them. In particular, we are interested in the effect of σw under stable conditions.
The project is currently underway, and the preliminary results are promising. We are aiming to achieve relatively simple parameterizations for our computed footprints so that these may be realistically applicable in field settings.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner