作者: Nele Schuwirth , Florian Borgwardt , Sami Domisch , Martin Friedrichs , Mira Kattwinkel
DOI: 10.1016/J.ECOLMODEL.2019.108784
关键词: Management support 、 Statistical model 、 Uncertainty quantification 、 Feature selection 、 Ecosystem model 、 Ecology 、 Environmental resource management 、 Computer science 、 Stakeholder 、 Decision-making 、 Causality
摘要: Abstract Understanding and predicting the ecological consequences of different management alternatives is becoming increasingly important to support environmental decisions. Ecological models could contribute such predictions, but in past this was often not case. are developed within research projects rarely used for practical applications. In synthesis paper, we discuss how strengthen role modeling supporting decisions with a focus on methodological aspects. We address mainly modellers also potential users results. Various approaches can be predict response ecosystems anthropogenic interventions, including mechanistic models, statistical machine learning approaches. Regardless chosen approach, outline better align decision making process, identify six requirements that believe increase usefulness support, especially if need justified public. These cover: (i) understanding regarding causality, (ii) alignment model input output decision, (iii) appropriate spatial temporal resolutions, (iv) uncertainty quantification, (v) sufficient predictive performance, (vi) transparent communication. challenges synthesize suggestions addressing these points.