A fast-response model of turbulence and passive scalar transport in row-organized canopies

Authors: ULMER, LUCAS - University Of Utah; MARGAIRAZ, FABIEN - University Of Utah; Mahaffee, Walter; STOLL, ROB - University Of Utah

Submitted to: Agriculture and Forest Meterology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/4/2024
Publication Date: 2/28/2024
Citation: Ulmer, L., Margairaz, F., Mahaffee, W.F., Stoll, R. 2024. A fast-response model of turbulence and passive scalar transport in row-organized canopies. Agriculture and Forest Meterology. 349. Article 109919. https://doi.org/10.1016/j.agrformet.2024.109919.
DOI: https://doi.org/10.1016/j.agrformet.2024.109919

Interpretive Summary: Many food crops are organized into intermittent rows due to trellising or the need to move equipment along rows. This creates channels that significantly influence pathogen spread and the potential of long distance dispersion. An improved understanding of how these differences impact disease development and effectiveness of management practices is needed. Methods for simplifying the physics based air turbulence models and accounting for the interaction of row and wind direction were developed and compared to air turbulence and particle dispersion data collected from commercial vineyards. The new models represented the observed particle dispersion patterns more accurately than previous models were able to run. This research will aid in the development of simulation environment that accurately shows growers how disease management practices impact disease development over a season and allows them to ask "what if" questions to improve their skill at disease management.

Technical Abstract: Accurate dispersion models are required to study the aerial transport of crop diseases and the dynamics of disease outbreaks. For many future applications of interest, such as spatially resolved ensemble-based disease forecasting or locating the epicenter of an outbreak using inverse modeling, the models must be quite fast. For crops grown in spatially homogeneous canopies, Gaussian plume models are often sufficient. However, in crops planted in rows with sparse aisles like vineyards and some orchards (so-called “row-organized canopies”), the impact of the vegetation structure on the flow limits the validity of Gaussian plume models. Previously, a simplified-physics model of the mean velocities in an ROC was developed and implemented in QES (Quick Environmental Simulation), a suite of fast-response modeling tools that includes a GPU-accelerated mass-consistent 3D wind solver, an eddy viscosity-based turbulence module, and a Lagrangian dispersion model. This work introduces a model for the Reynolds stresses near an ROC, which is validated against sonic anemometer data from a vineyard. The overall QES ROC model is then validated against a series of particle release experiments performed in the same vineyard. The turbulence model captures the shape and general behavior of shear stresses as the forcing wind direction changes. It accurately represents anisotropy in the normal stresses, where the row-parallel component of the Reynolds stress tends to dominate the row-perpendicular component regardless of wind direction. Both traditional paired-in-space metrics (e.g., FAC2, FAC10, false-positive and false-negative rates) and more detailed measures of 3D plume shape are used in the plume validation. For most wind directions, the model achieves FAC10 > 70% and FAC2 > 40% with false-positive and false-negative rates of less than 15%, meeting proposed acceptance criteria for dispersion models in complex terrain. Modelled plumes spread wider in row-perpendicular winds, as is seen in the field data, and veer in the direction of the row during row-diagonal winds due to mean wind channeling. Vertical spread in the QES model appears to be insufficient, and possible explanations are explored.