|
Activity Number:
|
6
|
|
Type:
|
Invited
|
|
Date/Time:
|
Sunday, August 6, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Environmental and Ecological Statistics
|
| Abstract - #304998 |
|
Title:
|
Applications of Hotspot Detection Analysis to Large-Scale Plant Disease Forecasting: Case Study of Fusarium Head Blight
|
|
Author(s):
|
Murali Haran*+
|
|
Companies:
|
The Pennsylvania State University
|
|
Address:
|
326 Thomas Building, University Park, PA, 16802,
|
|
Keywords:
|
hotspot geoinformatics ; disease mapping ; spatial misalignment ; crop disease ; lattice data ; Markov chain Monte Carlo methods
|
|
Abstract:
|
Fusarium Head Blight (FHB) is a disease that affects wheat crops, frequently resulting in enormous financial losses for farmers. Accurate forecasts of FHB are vital for cost effective management practices. Risk predictions for FHB are now available from weather-driven models of disease biology based on experimental data. Some sparse survey information about true disease rates is also available. We propose a hierarchical model that naturally integrates the various sources of information to predict risk. The underlying true disease rates are assumed to be spatially and temporally dependent. The survey data are treated as a function of the true disease rate and measurement error while the risk predictions are assumed to be functions of the true disease risk and terms that account for bias and random error. We conclude with an application of our approach to FHB data from North Dakota.
|
