DisplayTitle

<div>Disease forecasting models can be an important component of climate change adaptation. However, climate change offers challenges for the construction and maintenance of disease forecasting models, which must not only address weather variability, but also underlying trends in weather. We developed a framework for evaluating the climate scenarios where forecasting models will become more or less useful. In general, forecasting models may be of little use in environments where the disease in question is extremely common or rare. Optimizing forecasting models may involve tracking where climate change is predicted to shift disease prevalence. The quality (skill) of models used for constructing weather indices for yield loss is also an important factor, along with the spatial heterogeneity of the environment in which the models are applied. Where climate change results in increased weather variability, this may also reduce the success of forecasting models. One aspect of this effect is the potential for more frequent weather conditions that fall outside those used to parameterize the model. The extent to which a forecasting model is mechanistic will help to determine whether it can be applied effectively in novel weather conditions. We discuss the types of ongoing modifications that can help to make forecasting models resilient under climate change, so that they can successfully contribute to climate change adaptation strategies.</div>