Environmental Forecasting
Project period: October 2019 - October 2020 This project establishes a research focus of the Freiburg Institute for Advanced Studies (FRIAS) for the academic year 2019/2020.
Research Focus: Environmental models are the main tool through which we analyse and understand long-term and large-scale dynamical human-environment systems. They help transfer academic expertise into daily practices, guide the work of decision makers and inform the general public: weather forecasts, flood warnings, carbon balances of forests, landslides, recycling budgets are all computed using environmental models along a range of complexity. Yet, the intended audience can have crucial implications for the way a model is developed and for the predictions and explanations it produces. FRIAS will bring together a team of scientists to investigate the strikingly divergent modelling cultures along with their different scientific credibility, that have evolved in environmental disciplines.
Our project has three aims: (1) To understand modelling cultures as reflecting distinct goals. (2) To (3) To formulate a research agenda for those areas where we can identify deficits without an existing solution.
These aims are achieved through a work program that comprises two complementary parts: Part 1 - Causes of disciplinary forecasting cultures: We shall analyse the arguments put forward in forecasting studies from the disciplines represented in the research focus and beyond, as well as organise a workshop inviting active environmental modellers. Also we will look at the demand-side of environmental forecasting, analysing what decision makers want and expect from models. Part 2 - Towards best practice in environmental forecasting: We shall develop a fine-grained best-practice framework for environmental forecasting, taking into account the different goals of models, current practice and cutting-edge modelling possibilities. Elements of such a framework are, among others, uncertainty quantification, comparisons among models, fit to data, comparison with machine-learning approaches, an open-source culture and strategies to identify relevant processes and scales.
Our project produces: (1) a review paper of the current state of practice across environmental disciplines, (2) a detailed framework for model forecasting as a guide for maximising credibility of environmental forecast, (3) a DFG-proposal for a Graduate School on Environmental Forecasting, training the next generation of competent and critical environmental modellers.
Principal investigators: Prof. Dr. Carsten F. Dormann, University of Freiburg, Professorship of Biometry and Environmental System Analysis Prof. Dr. Kerstin Stahl, University of Freiburg, Professorship of Environmental Hydrological Systems Prof. Dr. Stefan Baumgärtner, University of Freiburg, Professorship of Environmental Economics and Resource Mangement
Detailed project information: for download (pdf) |
|