@karinnadrowski

Best practices in prediction for decision-making: lessons from the atmospheric and earth sciences

, and . Ecology, 84 (6): 1351--1358 (June 2003)

Abstract

Predictions that result from scientific research hold --- great appeal for decision-makers who are grappling with complex and controversial environmental issues, by promising to enhance their ability to --- determine a need for and --- determine outcomes of alternative decisions. A problem exists in that decision-makers and scientists in the public and private sectors -- solicit (erbeten, erbitten, erbetteln), -- produce, and -- use such predictions with --- little understanding of their accuracy or utility, and often --- without systematic evaluation or mechanisms? of accountability. In order to contribute to a more effective role for ecological science in support of decision-making, this paper discusses three "best practices" for quantitative ecosystem modeling and prediction gleaned from research on modeling, prediction, and decision-making in the atmospheric and earth sciences. The lessons are distilled from a series of case studies and placed into the specific context of examples from ecological science.Corresponding Editor: J. S. Clark use for prediction in the energy data forecasting a flood quite accurately but being accused: no understanding of uncertainty both forecasters and decision-makers failed to understand the uncertainty associated with the predictions both forecasters and decision-makers failed to understand the implications of uncertainty for decision-making thus, a "good" prediction product can contribute to a "bad" decision three lessons: 1 -- effective use of predictions results from focusing on prediction as one component in theprecess of decision-making sarewitz (2000) identifies three processes associated to the process connecting scientific predictions and policy: a research process, a communication process and a choice process 2 -- don't conflate (zusammenfügen, verschmelzen) prediction for science and prediction for policy 3 -- prediction products are difficult to evaluate and easy to misuse

Links and resources

Tags

community