W8: TIAS workshop on Complexity and uncertainty and a new role for models

Organised by Claudia Pahl-Wostl, Marcela Brugnach


Title: Complexity and uncertainty and a new role for models

Authors: Marcela Brugnach and Claudia Pahl-Wostl

Abstract: Models play a central role in understanding and managing complex systems. They constitute an important tool to experiment with complex and uncertain phenomena. In this paper we investigate how complexity and uncertainty affect the way in which we approach modeling: its development, its use and the societal role models play to understand and manage environmental problems. We aim at revising the assumptions upon which models are being built and used, claiming that the modeling activity needs to broaden its goals of predictions to encompass communication, exploration and understanding. Particularly our focus is on models and methods that allow a pluralistic representation of reality, embracing different types of uncertainties and that advance the modeling activity by combining with other methods of knowledge integration.

Title: A validated population dynamics model for Scrobicularia plana (Molusca, Bivalvia) in the Mondego Estuary (Portugal)

Authors: Pedro Anastacio, Tiago Verdelhos, Jojo Carlos Marques, Miguel Pardal

Abstract: A Scrobicularia plana (Molusca, Bivalvia) population dynamics model is proposed and was built using Stella v. 8.1 software. The model simulates the number of individuals on three different sampling stations located on the Mondego estuary (Portugal), before and after the management of the system. Field locations are a highly-eutrophic, a medium-eutrophic and a non-eutrophic area. The state variables are the larvae and age classes zero to five. Each age class is in practice a sub model with several months in sequence, due to modelling constraints. Forcing functions are the water temperature and the salinity. The model regulates mortality and recruitment using density, temperature and salinity. Two different sensitivity analysis methods are used: the traditional method with parameter modification and the effect on the final value of the state variable, and a modification of this employing the cumulative values of the state variable. Calibration was performed using data from the non-eutrophic area under improved estuary conditions. Model validation was performed with 5 different data series from 3 field stations and both pre and post-management situations. There was no need for model recalibration when validation was attempted using data from the same area (non-eutrophic) or the same post-management period but different areas. Nevertheless some recalibration, including a shift on the recruitment period, was needed for accurate simulations of pre-management scenarios on the highly-eutrophic and medium-eutrophic areas.

Title: Questions and methods to model emergence of land use patterns in coastal zone area

Authors: Tatiana Filatova, Anne van der Veen

Abstract: Interactions between economic and natural systems in the coastal zone area (CZA) result in complex systems, where nonlinearity and cross-scale interactions give rise to emergent properties, which cannot be predicted from the knowledge about elements and their interactions. This makes it difficult to model complex systems with traditional mathematical tools and begs for alternative methods of modelling complex systems such as computer simulations. The processes and interactions within economic and natural systems in CZA appear in a variety of temporal, spatial and analytical scales. Different factors and laws operate at each level and the nature of relationships changes across scales. It is a challenge to learn how the development of the whole system is defined by the emergent behaviour of micro agents. The neglect of this connection can cause major problems in CZA management and spatial planning policy along the coast. Dutch government guarantees a certain safety level for the territory in the CZA. But recently it was realized that part of the developed area, including coastal cities, is situated beyond the legally protected line. Nevertheless, there are different types of land use developing in this area. In that case we would like to explore the following questions: 1. What factors (economic and geophysical conditions, policy regulations) are crucial for the micro level land-use decision in CZA? 2. Do stakeholders understand the risk of flooding, and how do they perceive it? 3. How does the emergent behaviour of many land users lead to aggregated land use patterns in CZA? 4. What methods and simulation tools are appropriate to represent development of a complex system through the individual stakeholders’ decisions? We focus on the modeling of individual stakeholder behaviour. We introduce a notion of risk perception in the micro model and investigate how aggregated simulation outcome is sensitive to it. Agent based modeling is most likely to be the right tool for simulations. There are still open questions about what assumptions might be made and how data about stakeholder behaviour can be obtained.

Title: Issues of uncertainty, complexity, scale and transferability in river water quality modelling with a focus on the Saale River, Germany

Authors: Karl-Erich Lindenschmidt

Abstract: This contribution focuses on computer modelling issues such as i) uncertainty, including uncertainty in parameters, data input and model structure, ii) model complexity and how it affects uncertainty, iii) scale, as it pertains to scaling calibrated and validated models up or down to different spatial and temporal resolutions, and iv) transferability of a model to a site of the same scale. The discussion of these issues is well established in the fields of hydrology and hydrogeology but has found less application in river water quality modelling. This is an attempt to transferring these ideas to river modelling and to discuss their utilization when simulating river water quality. In order to provide a theoretical framework for the discussion of these topics several hypotheses have been adapted and extended. The basic principle is that model error decreases and sensitivity increases as a model becomes more complex. This behaviour is modified depending if the model is being upscaled or downscaled or is being transferred to a different application site. A modelling exercise of the middle and lower reaches of the Saale River in Germany provides a case study to test these hypotheses. The Saale is ideal since it has gained much attention as a test case for river basin management. It is heavily modified and regulated, has been overly polluted in the past and contains many contaminated sites. High demands are also placed on its water resources.

Title: Identification of major sources of uncertainty in IWRM and current approaches for including them in IWRM

Authors: Peter van der Keur, Marcela Brugnach, Claudia Pahl-Wostl, Jens-Christian Refsgaard

Abstract: Different types of uncertainty are of importance in the steps of understanding the complex dynamics of a river basin and in developing a management plan to achieve management objectives. These uncertainties include uncertainties in data, in models, and perspectives. The higher the uncertainties in the knowledge base, the more important becomes the framing of the issue (ambiguity). An inventory will be developed how models and DSS tools currently used in IWRM take into account uncertainty and complexity. This task will build on work from other EU projects, e.g. HarmoniCOP (social learning and ambiguity) and HarmoniRIB (uncertainty in data). Here it is inquired what role uncertainty plays in water resource management, with particular focus on (the transition to) the adaptive management approach. The point of departure is that uncertainty is inherent to the management process, and allows determining possible solution paths and directions for change. Here,different types and sources of uncertainty are identified and related to the management activity. It is considered that taking explicitly account of uncertainty in management leads to more robust water resources management under changing environmental, societal and economic boundary conditions. Thus incorporating the analysis of uncertainties in strategies for integrated water management leads to robust strategies that remain valid even if the assumptions on which they were based change. Robust strategies are flexible towards the future and changing conditions can be accommodated The objectives of the present paper are to explore and map the role of uncertainty in the IWRM process and to discuss possible tools to support uncertainty analysis at the various places in the IWRM process.

Title: The role of models in transition research

Authors: Georg Holtz, Claudia Pahl-Wostl

Abstract: Transitions comprise long-term, fundamental change in the way a particular societal function (e.g. water management, mobility) is fulfilled. As such they describe a set of connected changes in different domains like technology, economy, institutions or ecology. Transitions influence the development of society for decades and it is thus desirable to find ways to direct them towards societal goals. Transition research is very much in the beginning and models play a minor role up to now. Since models proofed to be useful tools to improve the understanding of similarly complex topics like climate change we believe that models can also be helpful tools to enhance our understanding of transitions. The broadness of the topic goes along with a huge demand for knowledge and data. The theoretical and empirical base of transition research is however weak. Several, often contradictory, theories are common in the social sciences and knowledge gaps exist especially regarding interdisciplinary issues. Transitions are inherently uncertain as they comprise human behaviour and include novelties (e.g. technological innovations, innovative institutions) which can not be anticipated in detail. Due to knowledge gaps and uncertainty models can not be used to predict future development but still can serve as tools to improve our understanding. For this purpose it is necessary to trace simulation outcomes back to assumptions what is only possible if models are transparent and simple enough. The intrinsic complexity of transitions however reduces the possibility to subdivide the system and study parts of it in isolation. Abstraction to ever higher levels as a means to achieve simplicity and transparency of models is limited by the diversity of elements and processes included. We suggest that models should be tailored to answer specific questions in order to reduce complexity. Hence they should be embedded in a broader process like e.g. transition management.

Title: Exploring regionalisation of hydrological behaviour within a model averaging framework

Authors: Julian Reichl

Abstract: The influence of uncertainties in environmental model prediction is undoubtedly large. When the system under study is complex it is inevitable that system detail is lost and crude assumptions are made by the modeller, potentially resulting in errors and biases. In addition, measuring techniques and methods are accompanied by significant errors that contribute to the error in the model prediction. In a case study of Australian catchments a model averaging framework is utilised for ungauged streamflow prediction. The influence of uncertainties in forcing and response data and model parameters is investigated in the context of this study.