Sensitivity of Modeled Wind-Canopy Interactions to Numerical Schemes in Higrad-Firetec

The validation of fundamental hydrodynamics and explicit drag formulation due to vegetation/canopy presence is a critical component to robust confidence in numerical modeling of numerous near-surface atmospheric and environmental dynamics scenarios. For example, investigations undertaken with Los Alamos National Laboratory's (LANL's) Higrad-Firetec atmospheric large-eddy-simulation (LES) and wildfire dynamics model are one such application scenario. The current work provides new validation of Higrad-Firetec for wind-canopy interactions in addition to a sensitivity analysis of modeled results under particular choice of numerical methodology. Four canopy configurations are studied using four different numerical schemes, and results are compared against published observational data in each case. Results of this sensitivity analysis show that computationally efficient schemes such as the Quadratic Upstream Interpolation for Convective Kinematics (QUICK) are well-suited for accurate numerical modeling of wind-canopy interactions and performance costs of more sophisticated Total Variation Diminishing (TVD) schemes such as Flux Corrected-Transport (FCT), or Weighted-Essentially-Non-oscillatory (WENO) are unnecessarily expensive in the computational sense when monotonic advection of prognostic quantities is not required. Moreover, these results motivate the need for field and laboratory experiments that quantify more comprehensive metrics beyond long-term time-averaged velocity profiles and turbulence statistics in order to validate unsteady dynamics (e.g. wind-canopy interactions that drive fire behavior).