Modeling the Dispersal Kernel Across Scales: Local Validation and Regional Insights into Seasonal Patterns and Spatial Variability

MANU NIMMALA, Hope Gruszewski, Regina Hanlon, Landon Bilyeu, Tyler Newton, David Schmale, Shane Ross, Hosein Foroutan, Virginia Tech

     Abstract Number: 384
     Working Group: Aerosol-Ecosystem Interactions

Abstract
Dispersal patterns of bio-aerosols like spores, seeds, and pollen are essential for studying the spread of plant diseases, gene flow, and cross-pollination. The dispersal kernel gives the probability of particle deposition as a function of distance from the source, and is highly dependent on meteorological conditions. For example, the movement of particles from local to regional scales is influenced by the balance between shear and convective turbulence. Lagrangian Stochastic (LS) models effectively capture this balance by using random walks to simulate particle trajectories. An ensemble of these trajectories forms a dispersal kernel. In this study, we use LS models to simulate dispersal kernels spanning local and regional scales, ranging from a few meters to 100 km away from the source. We first evaluate two LS model formulations – one for the surface layer and the other for convective conditions – against near-source, ground-level concentration measurements from a recent switchgrass field experiment in Tennessee. Following this local-scale validation study, we extend the model domain for a regional-scale application. Inspired by cross-pollination concerns in the emerging US hemp industry, we use LS modeling to simulate the dispersal kernel of hemp pollen for each US county. We drive the model using hourly data from a 2016 Weather Research and Forecasting (WRF) simulation over the contiguous US. The resulting dispersal kernels show seasonal and spatial patterns, revealing seasonal shifts in areas of the country more prone to long-distance dispersal. To the best of our knowledge, this is the first simulation study investigating the inhomogeneity of pollen dispersal across regions and seasons. Although we discuss pollen specifically, the methods used are general and could apply to dispersal of any light particle. Overall, this work contributes experimental data and model validation at local scales, and furthers understanding of how regional-scale dispersal kernels change with weather conditions.