Quantifying the compounding effects of grid resolution to improve vegetation predictions in delta models

Vegetation growth in coastal environments, such as estuaries and deltas, plays an important role in coastal morphology. Gridded hydrodynamic and morphodynamic models typically have options to incorporate constant vegetation effects via a roughness parameter, where taller and denser vegetation is associated with higher roughness that alters flow velocities and sediment transport. In this study we present DYCOVE (DYnamic COastal VEgetation): a flexible, open-source model that couples with physics-based, hydro-(morpho)dynamic models to simulate life-cycle dynamics (colonization, growth, and mortality) of multiple vegetation species in coastal environments. Via interactive coupling, changes in vegetation state provide spatial and temporal updates of friction effects in the physics-based model, creating a dynamic feedback loop. However, vegetation colonization, growth, and mortality depend on accurate modeling of local inundation, flow speeds, and bed level changes, which can be strongly dependent on model grid resolution. Finer grid sizes resolve these physical processes more accurately but are not always feasible due to computational demand, whereas coarser grids may result in topographic smoothing that distorts the hydrodynamic solution. Since vegetation depends on reasonable accuracy of physical processes, any errors in hydrodynamic predictions will affect the vegetation calculations, resulting in potentially compounding errors over several iterations. Using DYCOVE, we quantify errors caused by grid resolution effects and evaluate produced differences in vegetation distributions compared to observations from the Wax Lake Delta in Louisiana, USA. Lastly, we discuss strategies for overcoming limitations and errors related to grid resolution that allow for accurate and efficient prediction of vegetation species cover in coastal environments.