Search

A simple Bayesian network to interpret the accuracy of armyworm outbreak forecasts

A simple Bayesian network to interpret the accuracy of armyworm outbreak forecasts

This document describes the relationship between rainfall and moth catch and how it can be used to forecast an outbreak of fall armyworm, using conditional probabilities to predict armyworm larvae outbreaks. The data were from a new initiative – a community-based way to forecast an outbreak of armyworm in Tanzania, in which the outbreak risk for a village is determined locally from a single moth trap and rain gauge located within the village. It was found that, following a positive forecast, an armyworm outbreak was approximately twice as likely to occur as would be expected by chance.


Region: Tanzania
Date published: 2006
Published by: Annals of Applied Biology
Type of resource: Journal article
Resource topic: crop protection, Pests


Project/Programme: Not specific
Pest/Disease: Fall armyworm
Pages: 6
File type: PDF (124 KB)

Open resource Download resource