Model to Enhance Site-Specific Estimation of Leaf Wetness Duration

Area of Science:

• Agricultural Meteorology
• Environmental Modeling

Background:

• Accurate leaf wetness duration (LWD) is crucial for disease prediction in agriculture.
• Existing empirical models have limitations in site-specific LWD estimation accuracy.

Purpose of the Study:

• To assess the accuracy of empirical models for site-specific leaf wetness duration (LWD) estimation.
• To evaluate a modified Classification and Regression Tree/Stepwise Linear Discriminant (CART/SLD) model incorporating wind speed correction.

Main Methods:

• Compared a CART/SLD/Wind model with a 0.3-m height wind speed correction against proprietary (SkyBit wetness) and standard CART/SLD models.
• Evaluated model performance across 15 sites in Iowa, Nebraska, and Illinois from May to September over three years (1997-1999).

Main Results:

• The CART/SLD/Wind model demonstrated enhanced LWD estimation accuracy compared to other models.
• Improved accuracy was observed during both dew-eligible (night) and dew-ineligible (day) periods, and irrespective of rain events.
• Model accuracy showed minimal variation across the tested sites.

Conclusions:

• A CART/SLD model with a 0.3-m wind speed correction significantly improves LWD estimation accuracy.
• The model's robustness across sites suggests its utility for midwestern agricultural applications.
• Hierarchical decision-making and wind speed correction contribute to the model's enhanced predictive power.