To improve our understanding of aquatic ecosystems and support societal decision making in water management, we contribute to the following three research fields:
We develop mathematical models of aquatic systems to (1) formalize our quantitative understanding of system structure and functioning, (2) test alternative hypotheses, and (3) predict future behavior of these systems under changing external driving forces. The growing importance of mathematical models in various applications is widely recognized within Eawag. Siam is active in the development of models for complex hydrological and ecological systems, and provides modeling know-how to other departments.
- Development: We develop model building techniques and integrated models to quantitatively represent relevant system structures and functions. Our emphasis is on aquatic systems such as small to large scale catchments, rivers and lakes.
- Applications: We apply these models to improve our understanding of the response of large scale catchments and crop growth to climate change, and of river and lake ecosystems to management interventions, e.g. rehabilitation projects and measures to improve water quality.
A careful analysis of uncertainty is essential to (1) assess the degree of confidence in a given model, (2) compare the relative performance of alternative models, and (3) generate reliable model predictions. The development of rigorous approaches for statistical inference and uncertainty analysis is therefore a key research field to support our other activities.
- Development: We focus on the development of realistic likelihood functions for environmental models that correctly represent the various sources of uncertainty and stochasticity. We develop Bayesian methods to infer parameters and states of these models, to quantify their uncertainty and to make probabilistic predictions. In particular, to appropriately account for input and model structural errors, we emphasize approaches that can deal with stochastic models and consider input in the form of stochastic processes.
- Applications: We apply these techniques to increase our understanding of environmental systems and improve our predictive capabilities, mainly within the scope of the other two research fields.
Supporting environmental management decisions requires the consideration of multiple objectives, multiple societal perspectives, and the prediction of consequences of alternative management strategies of complex environmental systems. We evolve and apply approaches of Multi-Criteria Decision Analysis (MCDA) to formally combine scientific predictions with quantified societal preferences for the support of rational decision making in environmental management.
- Development: We develop methodologies and procedures for structuring the societal decision-making process, for involving stakeholders and for eliciting societal values under consideration of the possibility of trade-offs that depend on the system state (non-additive aggregation). We focus on quantitative analysis including the quantification and communication of uncertainty.
- Applications: We apply these methodologies to create ecological assessment procedures for aquatic ecosystems, to support decision making in surface water management in Switzerland (e.g for river rehabilitation planning and water quality management) and to water management and food production in selected cases worldwide.