Development toward an Advanced Aerobiological Vegetable Disease Forecasting Tool

The Pest Information Platform for Extension and Education (ipmPIPE) is a service that provides national warning to farmers in an effort to protect crops from devastating diseases. The Integrated Pest Management (IPM) team uses many tools to aid in the forecast of aerobiological plant disease epidemics, in this case Cucurbit Downy Mildew, as they spread downstream from various sources. Cucurbit Downy Mildew (CDM, Peronospora cubensis) is a fungus that can affect cucurbits such as squash, cucumbers, and pumpkins leading to significant decreases in vegetable yield and quality. The fungus is very progressive and can spread rapidly under favorable weather conditions. It is very time consuming and costly to prevent and treat such diseases. Disease activity generally persists on hosts in the Deep South during winter (south of the climatological freeze line). The epidemic progressively spreads northward into the continental United States and Canada through repeated events of dispersion and deposition on downstream hosts. Therefore, a dispersion model is used by the Cucurbit ipmPIPE team at NC State University to predict the future atmospheric trajectories and dispersive patterns of airborne fungal spores to provide guidance for the risk of new fungal outbreaks.

Based on the atmospheric conditions and plant biology / pathology, risk predictions for susceptible hosts are issued dependent upon the source strength, spore survivability, and potential for deposition and infection. In an effort to increase the accuracy of these assessments, the forecaster must have a comprehensive understanding of the dispersive nature of the particles under current and future weather conditions, which often alter the deposition location and rate. Previously, the HYbrid Single Particle Integrated Trajectory (HYSPLIT) model was used to capture the transport and dispersion of airborne spores. However, use of HYPLIT can be challenging for a plant disease meteorologist given the many source locations throughout North America. Hence, automated use of another dispersion model (FLEXPART) has been introduced to combine the high resolution meteorological data with a more flexible dispersion model.

FLEXPART is a Lagrangian dispersion model that can simulate long- and short-range atmospheric transport, diffusion and deposition of trace gases both forward and backward from a point, line or area source. FLEXPART was first adapted to be driven by the European Centre for Medium-Range Weather Forecast (ECMWF) and was later modified to run off of the Weather and Research Forecast (WRF) model capturing the important mesoscale features. Used by over 30 groups across 17 countries, FLEXPART is becoming versatile in its applications. The most common applications are air quality, fire emissions, emergency response, ozone and stratospheric exchanges. Most recently, FLEXPART has been used to provide experimental guidance for plant disease forecasting.

FLEXPART is driven by meteorological fields produced by the Advanced Research WRF (ARW) version 3.1 to predict where and when these diseases will occur. FLEXPART creates dispersion patterns based on WRF meteorological fields for known areas of infection. Within the dispersion model framework, many features are taken into account such as convective parameterizations, turbulent fluctuations and boundary layer dynamics. Other minor characteristics that aid in replicating the dispersive nature of the atmosphere are particle splitting, removal methods and age class of particles. After the calculations are complete, dry and wet deposition arrays are produced in a text format where they are ingested into the GRASS geographic information system (GIS). FLEXPART is post-processed using GIS to project and interpolate accurate dispersion plumes across the specified region. As the final step, the GIS images are then placed into GoogleMaps API producing a user-friendly application for forecasters. An interactive website has been created based on these new techniques and includes a variety of information tools for assessing and predicting plant disease risk such as observed precipitation, satellite imagery, and known infection sources.

The use of an atmospheric transport model in conjunction with a dispersion platform can aid in the prediction of many different types of particulates. Long-range plant disease forecasting, as part of a larger plant disease management framework, serves as an important decision-support tool as growers consider spray programs. The use of FLEXPART and other new technologies is expected to increase both the accuracy and efficiency of plant disease epidemic forecasting. The improvements should accentuate the benefits of the forecasting effort, leading to better spray decisions and better use of fungicides. The results are improved crop quality and increased yields, decreases in negative environmental impacts from overuse of chemicals, and reductions in costs for both growers and consumers.