Files
Abstract
In this study, we formulate a stochastic dynamic framework for pest control over the
growing season taking into account forecasts of weather conditions and pest infestation
expectations. Using stochastic envelope theorem and stochastic comparative dynamics,
we analytically show how the stochastic correlation between the prediction errors should
affect optimal pesticide usage path. As a case study, we apply the analytical results of
the paper for pesticide use in the Palouse region of Washington where pea aphid is the
primary threat for lentil production. By stochastic dynamic programming, our simulation
shows the optimal dimethoate usage path, which illustrates our findings in the analytical
part.