Development of a Simulation Environment for Pathogen and Pest Spread in Vineyards

The overall goal of this research has been to develop an innovative modeling platform that can accurately simulate pathogen and pest spread in vineyards. The simulation tool will serve to help producers evaluate disease and pest management decisions using “virtual” crops. This system will allow producers to evaluate “what if” scenarios and to examine how to isolate individual management decisions that influence disease, pest, and plant development. The system can also be used to examine how row orientation, training system, etc. interact with climate and geography at new vineyard locations.

Completed research to date has focused on model development and integration. The project has successfully produced a model framework that integrates previously developed models for climate, plant growth, spore dispersion, pathogen infection, and colony growth. The system is currently able to simulate plant growth and disease progression throughout a growing season. An initial “vineyard builder” tool has been developed to rapidly build up the geometry of a particular vineyard of interest within the simulation system. Work is also underway to develop improved sub-models for meteorology and turbulence. The meteorological model will predict the three-dimensional turbulent wind field, which drives the airborne dispersion model. This work has involved comparing model outputs to field measurements, and making necessary modifications to the model to improve agreement between the two. Other work is developing improved models for airborne particle deposition to plant surfaces.

The overall modeling platform consists of a suite of coupled sub-models that represent the most important physical processes of disease spread such as plant growth, local climate, airborne dispersal by the turbulent wind, pathogen infection and colony growth. These state-of-the-art sub-models are among the most detailed simulation tools that have ever been developed for agricultural crops. Since they require substantial computational resources not provided by the processors of a standard desktop or laptop computer, we have overcome this limitation by using standard computers with a gaming graphics card. We have used the graphics card to accelerate many of the sub-models, meaning that very large simulations can be performed in a matter of minutes.