Simplicity, model fit, complexity and uncertainty in spatial prediction models applied over time: We are quite sure, aren't we?

Abstract

There is a strong need to assess impacts on wildlife and their habitats before they occur, and to act proactively to avoid costs (e.g., loss of species, wilderness, ecological-services, human lives or money; see Nielsen et al. 2008 for an applied example). Although the concept of being proactive has been known for decades, the global climate change discussion has brought these concepts to the forefront. Proactive action also has use in impact studies of stochastic catastrophes such as floods or hurricanes. Simulations and predictions across time can help to mitigate or even resolve current problems (Fig. 10.1). Such techniques are widely used to assess risks and they have evolved into industrial standards elsewhere, such as in operations research (Fuller et al. 2008), within the pharmaceutical, insurance, and car industries, and in public health (Herrick et al. in press). Weather forecasts have already shown the value and power of such models, -driving many day-to-day decisions. © 2011 Springer Science+Business Media, LLC.