A metamodeling framework for extending the application domain of process-based ecological models

Abstract

Abstract. Process-based ecological models used to assess organisms' responses to environmental conditions often need input data at a high temporal resolution, e.g., hourly or daily weather data. Such input data may not be available at a high spatial resolution for large areas, limiting opportunities to use such models. Here we present a metamodeling framework to develop reduced form ecological models that use lower resolution input data than the original process models. We used generalized additive models to create metamodels for an existing model that uses hourly data to predict risk of potato late blight, caused by the plant pathogen Phytophthora infestans. The metamodels used daily or monthly weather data, and their predictions maintained the key features of the original model. This approach can be applied to other complex models, allowing them to be used more widely.

The effects of climate variability and the color of weather time series on agricultural diseases and pests, and on decisions for their management

“…Within-season forecasting models can also be rescaled for application at greater spatial extents, such as in climate change scenario analysis (Sparks et al, 2011).…”

Worldwide Sustainability Hotspots in Potato Cultivation. 2. Areas with Improvement Opportunities

“…The risk of incurring yield losses due to infestation with late blight was assessed based on the work of Sparks (Sparks 2009;Sparks et al 2011). In this work a detailed process-based ecological model was used to derive a metamodel which uses lowresolution input data of temperature and relative humidity and can therefore be applied worldwide.…”

“…The threshold for blight risk was adopted from (Sparks 2009;Sparks et al 2011): a value of 4.1 or higher is considered high risk. For temperature as indicator for the risk of occurrence of pests, no authoritative source could be identified.…”

Landscape-Scale Disease Risk Quantification and Prediction

“…Given that future climatic data consist of monthly averages and that the processes that are modeled in late blight require data at a much finer temporal scale (such as daily or even hourly measurements), the first step was to develop a metamodel that could be used with these monthly data (58). The metamodel was generated via generalized additive models and an existing model driven by hourly data.…”