W11: Scenario Development And Analysis For Integrated Environmental Assessment Studies

organised by Thorsten Wagener, Yuqiong Liu, Steve Stewart, Holly Hartmann

Abstracts


Title: Imagine - Scenario Development for Environmental Impact Assessment Studies

Authors: T. Wagener, Y. Liu, S. Stewart, H. Hartman and M. Mahmoud

Abstract: Scenario analysis is a process of evaluating possible future events through the consideration of alternative possible (though not equally likely) outcomes (scenarios). The analysis is designed to enable improved decision-making and assessment through a more complete consideration of possible outcomes and their implications. The development of strategies for water resources planning and management and the assessment of impacts of potential environmental change are often guided by analyzing multiple future scenarios within an Integrated Modeling (IM) framework, usually driven by forcing derived from global climate models and/or possible future socio-economic changes. The process of scenario development involves making explicit and/or implicit assumptions about potential future conditions, such as climate change, land cover and land use changes (e.g. urbanization), population growth, economic development, and technological change. These scenarios generally surpass forcing or behavior that has been observed in the past. Realistic assessment of scenario impacts requires complex modeling frameworks that represent environmental and socio-economic systems to the best of our knowledge, including assumptions about probabilities of the occurrence of future conditions. In addition, scenarios have to be developed in a context relevant to the stakeholders involved, to facilitate transparency of scenario results, and to establish credibility and relevance of the results among them. Hence, for the IA models to be useful for policy making, appropriate scenarios have to be carefully constructed and associated uncertainties propagated into the model outputs have to be understood and quantified. This paper is a review of the state-of-the-art of scenario development.


Title: Contrasting citizen and expert-driven approaches to creating scenarios

Authors: David Hulse

Abstract: Scenario analyses are a rich set of robust, general approaches for conducting anticipatory integrated assessments. Modelers are increasingly being asked to simulate the interactions that influence complex coupled human and natural systems. In such complex systems, properties emerge that affect living organisms, including people. These emergent properties are being simulated using agent-based modeling approaches, but as with the approaches that came before, agent-based modelers must articulate an explicit story about how the future may unfold. In systems involving people, this forces strongly held but vaguely defined values out into the open, where they may be given a fair test against what is considered plausible.

This paper posits that two different approaches are used to create scenarios. The first is citizen-based and employs citizen stakeholder groups to define internally consistent narrative assumptions about how future conditions will unfold. These scenarios are then used as inputs to a set of landscape change models to produce maps of alternative futures. Alternative future maps developed through citizen-based scenarios have the advantages of easier citizen involvement, greater political plausibility, and increased likelihood of institutional acceptance. They have the disadvantages of a small number of alternatives due to the time required to create them and the inability to statistically quantify their likelihood given their small number. The second approach to creating scenarios is expert-based, in which experts in the biophysical and social sciences define a set of decision or transition rules in a simulation model that depicts future conditions. If the rules of land change functions are probabilistic, central trends and variation can be determined for the large number of simulated future landscapes. Alternative futures produced using this approach have the advantage of quantifiable statistical likelihood from the very large number of alternatives produced and the disadvantage of unclear political plausibility.


Title: Climate and Socio-economic Scenarios: Lessons Learned Through Stakeholder Engagement and Use

Authors: Roger B. Street, Richard Westaway, Michelle Colley

Abstract: Scenarios, both climate and socio-economic, are integral to the delivery of the UK Climate Impacts Programme (UKCIP). Stakeholders (central and regional governments, local authorities, other public and private organizations and agencies, consultants and researchers) are at the heart of this Programme. They are involved in the associated research efforts, including the design and development of information, tools and other resources that inform and support stakeholders when assessing their climate vulnerabilities and making decisions related to reducing those vulnerabilities. Climate and socio-economic scenarios are among the resulting assessment and decision support tools. UKCIP with the Hadley Centre (climate modelling) have made available two generations of climate scenarios (1998 and 2002) and is currently working with its stakeholder communities towards producing and making available the next generation of climate scenarios (expected delivery 2008). Socio-economic scenario development, likewise, has relied upon researcher-stakeholder engagement to deliver the required scenario package. Stakeholder involvement in the development of these two types of scenarios and in designing their delivery packages has been shown to be instrumental in their acceptance, understanding and adoption. Each generation of the climate scenarios, as well as the continued development of the socio-economic scenarios, has benefited from the lessons learned through stakeholder engagement and use. The resulting scenarios, as well as their packaging and delivery mechanisms reflect the range and evolution of needs and sophistication of the UKCIP stakeholders.


Title: Comparing Scenarios Through Model Output

Authors: Larry Larry Winter

Abstract: Simulation of the states of an environmental system depends on i) a formal description of the state, ii) the system’s external properties, and iii) a model for computing future states from the system’s present, its past, and its external properties. External properties may include initial and boundary conditions, as well as system parameters. More formally, a model, M, maps a system’s previous states, St-, and properties, P, into its future states, St+,

M: St- x P -> St+. (1)

A scenario is produced when M', a particular computational model is applied to P', a specific set of system properties, i.e., M': S't- x P' -> S't+. Engineers and scientists simulate environmental systems for specific purposes or to answer specific questions, and those questions or purposes influence the design of M, as well as the definitions of P and St. Of course, simulations may differ on any of these dimensions, but in this talk I will concentrate on differences between simulations that are due to different models. A related set of concerns revolves around “model error”, or the differences between a simulation and reality that are due to the choice of underlying model. In many cases we do not have enough information about the detailed behavior of an environmental system to compare simulations to reality, so I will address the more limited problem of comparing two models of the same system. To be precise, the system will be a small watershed and the system state will be spatially distributed soil moisture, and the problem will be to develop metrics for comparing the performance of different models on a given scenario; that is, I will compare S't to S"t where the former arises from one model, M': S't- x P -> S't+ and the latter from another, M": S"t- x P -> S"t+ . Along the way, I will discuss other uses of models than scenario evaluation, and I will consider the role of other forms uncertainty, especially parametric uncertainty.


Title: Multi-Agent systems to simulate scenarios for decision-making in river basin systems

Authors: Thania Rendon, Miquel Sanchez-Marre

Abstract: The complexity of environmental problems makes the management of environmental systems especially difficult to be undertaken by traditional software systems. Particularly river catchment systems are very intricate to manage in order to achieve a good quality and quantity of water at the river. River catchments are important social, economical and environmental units. They sustain ecosystems, which are the main source of water for households, agriculture and industry. Therefore the protection of all surface waters and groundwaters must be assured in their quality and quantity. The best way to fulfill these requirements is with a management system at catchment scale that integrates all the water systems involved (sewer system, Waste Water Treatment Plants and River). Multi-Agent systems (MAS) are able to cope with this complexity by integrating several agents who model real environmental situations. We propose the design of a MAS for simulating multiple scenarios in a river catchment system in order to support in the decision-making of river basins management. The aim of the MAS is to simulate various scenarios in order to draw conclusions and help in the decision-making for the river basin management. There are other objectives to be fulfilled like: to manage critical episodes; to minimize discharge of poorly treated wastewater; to maximize the use of the installations treatment capacity; to minimize the economical costs of new investments and daily management; and to maintain a minimum flow in the river guaranteeing an acceptable ecological state. The River Basin MAS has been designed and their functionalities have been studied. Moreover, we have selected the agent platform Jadex as a starting software platform for our new software development. With this proposal we intend to provide feasible solutions at catchment scale throughout modelling and simulation of different scenarios in a river basin system.


Title: A Modular Modeling Approach to Integrating Adaptive Modeling Systems with Resource Management in the FRAME Project

Authors: George Leavesley, Jimmie Chew, Roland Viger, Christine Turner, William Romme, Mark Miller, George San Miguel, Neil Cobb, Lisa Floyd-Hanna

Abstract: The FRAME (Framing Research to support Adaptive Management of Ecosystems) project is a collaborative, multi-disciplinary effort currently focusing on pinyon-juniper woodland management on the Colorado Plateau. The USGS Modular Modeling System (MMS) provides a modular framework to address a variety of pinyon-juniper management issues using a set of adaptive modeling tools. MMS is being coupled with the U.S. Forest Service model SIMulating Patterns and Processes at Landscape Scales (SIMPPLLE) to enable the assessment of the effects of alternative resource-management options on a variety of hydrologic and ecosystem processes. A variety of watershed, erosion, hydraulic, and ecosystem models in MMS will be used to evaluate the spatially explicit output of SIMPPLLE. Output from SIMPPLLE is an ensemble of potential vegetation conditions years to decades into the future. Key components of the linked MMS and SIMPPLLE models are 1) tools to estimate parameters in MMS process-based models using vegetation and ecosystem attribute data from SIMPPLLE output, and 2) a climate generator to provide time series of meteorological variables, such as precipitation and temperature, for use as input to the process-based models. The magnitude and timing of these meteorological variables must be spatially and temporally representative of possible future climate conditions. Initial application of the coupled MMS-SIMPPLLE modeling tools is to support fire-management planning at Mesa Verde National Park. A major objective of the development and application of these tools is to allow resource managers to develop more flexible management scenarios that can adjust to changing conditions, and to develop spatially explicit landscape-management scenarios that incorporate the social, economic, legal, and environmental constraints that managers face.


Title: Integrated Environmental Assessment of Urban Underground Infrastructure Development Policies

Authors: Nikolai Bobylev

Abstract: At the turn of the twenty-first century, nearly half of the world’s population (about three billion people) lives in urban areas. It is estimated that in the next twenty-five years, almost two billion more people will move to cities. Sustainable management of the urban environment represents one of the major challenges for the future. Development of the urban underground infrastructure is one of the ways to tackle transport, space, and environmental problems of urban areas. However underground developments can impose many adverse impacts on the environment, thus methodology for comprehensive assessment of various environmental impacts is needed. The presentation will discuss methodology for environmental assessment which utilizes analytic hierarchy process of multiple criteria decision analysis. Emphases will be put on discussing a hierarchy structure of criteria for assessment and analysis of assessment results for different policies.


Title: Scenario development for semi-arid regions: a multi-disciplinary example

Authors: Steven Stewart, Holly Hartmann, Yuqiong Liu, Thorsten Wagener, Mohammed Mahmoud

Abstract: We illustrate the scenario development process for SAHRA, an NSF Science and Technology Center charged with conducting stakeholder-relevant research on hydrology and water resources in semi-arid regions. Much of SAHRA’s research relies on multidisciplinary models of the Rio Grande and San Pedro river systems in the Southwestern U.S., a region which is experiencing extreme variability in water supply and rapid population growth. Developing a consistent set of scenarios allows us to forge linkages between local and regional models and and the conflicting interests of local and regional stakeholders. We begin with a n overview of the scenario development process with an emphasis on the development needs unique to semi-arid environments. We highlight the process of developing coherent scenarios that are consistent across multiple models of varying scale and scope, and built to address a variety of science and stakeholder-driven questions. Included is discussion of the the process of collecting stakeholder input, how scenarios can be used to help integrate disciplines having often-conflicting paradigms (hydrology, ecology and economics), and how understanding of scenarios, integrated models, and their limitations can be effectively related to stakeholders. We conclude with scenario definitions developed under this process.


Title: Use of High resolution Lidar Data for Targeting and Prioritization Methodology of Residential Stormwater BMPs

Authors: Helena Vladich, Alexey Voinov

Abstract: There is much concern about quality and quantity of stormwater draining from residential properties. Local and State agencies nationwide realize the importance of stormwater management and design practices that can be implemented in a variety of scales. Working with the neighborhood that already have been developed, in many cases it is expected that at least some burden of cost for these management practices will be shared by the residents, like with the utility institutionalized by the city of South Burlington in Vermont. However there are not many tools that could help residents make informed decisions about the alternative methods of distributed stormwater management. We explore how high resolution Lidar data can provide essential information to compare and prioritize scenarios for traditional and alternative stormwater best management practices (BMPs). Applying ArcGIS hydrological modeling tools and Model Builder capabilities we are able to target subcatchments for traditional big scale engineering solutions and discriminate between the most appropriate locations to install medium and small scale distributed alternative BMPs. In many cases this approach can be a cheaper alternative to full-scale engineering surveys usually conducted for stormwater planning.