Sessions

 

Stream: Model Development, Analysis and Application: Methodological Aspects

 

D1.1 Uncertainty analysis - environmental model application
James C. Ascough II, Holger Maier, Olaf David

The purpose of this session is to provide a forum for presentations focusing on communicating state-of-the-art information on advances in uncertainty analysis methodologies for integrated environmental models.

Suitable uncertainty analysis (UA) topics for this session include, but are not limited to:

  • Evaluation of uncertainty in model outputs with respect to decision making or risk management objectives
  • Uncertainty propagation in complex, environmental models with large parameter sets or high computational requirements
  • Assessing and quantifying information requirements (e.g. theories, data, models) to reduce predictive uncertainty in environmental models
  • Scale effects in UA of integrated environmental models
  • Methods for identifying and managing structural uncertainty and bias in integrated environmental models.

Link to full papers

 

D1.2 Uncertainty analysis - general frameworks, data and decision support
Jack Carlson, James C. Ascough II, Olaf David, Holger Maier

The purpose of this session is to provide a forum for presentations focusing on communicating state-of-the-art information on advances in uncertainty analysis methodologies for general frameworks, data and decision support.

Suitable uncertainty analysis (UA) topics for this session include, but are not limited to:

  • Evaluation of uncertainty with respect to decision making or risk management objectives
  • Incorporation of uncertainty in decision support methods, such as multi-criteria decision analysis
  • Development and evaluation of UA methods that appropriately consider subjective and qualitative factors
  • Assessing and quantifying information requirements (e.g. theories, data) to reduce predictive uncertainty in environmental models
  • Assessment of uncertainty in socio-economic models.

Link to full papers

 

D1.3 Sensitivity analysis
Holger Maier, James C. Ascough II, Olaf David

The purpose of this session is to provide a forum for presentations focusing on communicating state-of-the-art information on advances in sensitivity analysis methodologies for integrated environmental models

Suitable sensitivity analysis (SA) topics for this session include, but are not limited to:

  • Practical strategies for local/global SA given models with large parameter sets or high computational requirements
  • The use of SA to gain insights into key sources of uncertainty in order to prioritize additional data collection or research efforts
  • Key criteria in selecting SA methods for different Modelling structures and problems
  • SA in the context of probabilistic risk assessment (PRA) and
  • Limitations and promising new advances/directions for SA methodologies in environmental models.

Link to full papers

 

D2. Distributed environmental modelling
Ari Jolma, Karl Aberer, Phillip C. Dibner, Kostas Karatzas, Mike Sips

Modelling is a process that involves articulation of the goals, understanding the domain reality, selection of methods and tools, integration, and eventually testing of the model and interpretation of the results. A lot of modelling work is moving into the Internet in the sense that baseline data, real-time measurements, and computational resources are becoming available over the Internet, collaboration is carried out by email and increasingly by interactive websites, and many tools are being exposed to users on the Internet. Service-orientation is becoming the leading paradigm for developing information systems and infrastructures. Models, many steps of the modelling process, and the modelling process itself can be implemented as or aided by services provided by software on the Internet. These services can be targeted to programs or to humans and they can be based on almost any computational method. New generic web browser based technologies including free libraries for many kind of data make highly interactive and visual tools relatively easy to implement and publish. There are many general distributed systems issues that need to be considered from the point of view of environmental modelling: search, trust, linking of services, semantic interoperability, and so on. For this session we seek case studies, descriptions of novel service concepts, theoretical considerations, and other contributions, which demonstrate or give new insights into how modelers can be better interact with each other, use and create services in the Internet, and help decision makers.

Link to full papers

 

D3.1 Advances in software engineering for IEM - methods and approaches
Olaf David, Sven Kralisch, James C. Ascough II, Peter Krause

The purpose of this session is to provide a forum for presentation and discussing approaches, methods, techniques, platforms, and general advances in software engineering beneficial for integrated environmental modelling. Therefore, this session is specifically dedicated to software methods and approaches for integrated modelling.

Session topics include (but are not limited to): design patterns, environmental modelling framework design approaches, design and use of domain specific languages (DSLs), workflow engines, model testing, model/framework integration and coupling approaches, advanced methods for integrated models such as parallelization techniques, services integration methods or cloud computing.

Link to full papers

 

D3.2 Advances in software engineering for IEM - applications and use cases
Olaf David, Sven Kralisch, James C. Ascough II, Peter Krause

The purpose of this session is to provide a forum for presentation and discussing approaches, methods, techniques, platforms, and general advances in software engineering beneficial for integrated environmental modelling. Therefore, this session is specifically dedicated to software methods and approaches for integrated modelling.

Session topics include (but are not limited to): design patterns, environmental modelling framework design approaches, design and use of domain specific languages (DSLs), workflow engines, model testing, model/framework integration and coupling approaches, advanced methods for integrated models such as parallelization techniques, services integration methods or cloud computing.

Link to full papers

 

D4. Characterising environmental modelling paradigms (Bayesian networks, system dynamics, agent based models, coupled complex models and expert systems)
Tony Jakeman, Olivier Barreteau, Mark Borsuk, Carlo Giupponi, Rebecca Kelly, Andrea Rizzoli, Barbara Robson, Sondoss El Sawah, Alexey Voinov

A range of model types now exists for such purposes as scientific understanding, prediction, communication, social learning, and environmental decision making. Each type has strengths and weaknesses in any given situation which we classify, at the very least, according to purpose, breadth and depth of issues being addressed, level of spatial and temporal detail required, type of input data available and outputs required (eg quantitative versus qualitative), capacity to address uncertainty and assess model credibility, and capacity for effective communication and social learning.
In many cases models are chosen without good reason, and often model choice is determined simply by the existing skills of the modeller involved. This becomes especially limiting in a participatory modelling effort, where stakeholders in many cases do not have sufficient knowledge and understanding to make an informed selection and may be easily persuaded by modelling results.

The aim of this linked session and workshop (D8) is to characterise and compare the various paradigms using the above 'metrics' so that better guidance is available for future model selection. Other metrics for classification can also be considered.
The Session's presentations and accompanying full papers would describe one or more modelling approaches and be structured according to the metrics above. The organisers also welcome prior application of modelling paradigms to a set of common case studies they are setting up for the purposes of comparative analysis. The session will begin with the presentation of a position paper currently in preparation and initially drafted by Rebecca Kelly and Tony Jakeman but being updated by the organisers. This conference PP will be published in the proceedings and revised after the conference for publication in EMS as a journal PP.

Link to full papers

 

D6.1 Geographic Information Systems and geoprocessing workflows for environmental modelling - GIS Based Workflows

D6.2 Geographic Information Systems and geoprocessing workflows for environmental modelling - GIS Hydro-Climate Analysis

D6.3 Geographic Information Systems and geoprocessing workflows for environmental modelling - GIS Web Services & Systems

D6.4 Geographic Information Systems and geoprocessing workflows for environmental modelling - GIS Land-Surface Analysis

D6.5 Geographic Information Systems and geoprocessing workflows for environmental modelling - GIS Spatial Temporal Analysis

all lead by Daniel P. Ames, Robert Argent, Susan Cuddy, Nigel W.T. Quinn, Raul Zurita-Milla

In September 2000, the 4th International Conference on Integrating GIS and Environmental Modelling (GIS-EM4) was held in Banff, Canada with more than 250 presentations. In the ensuing 12 years no major "GIS for Environmental Modelling" meeting has been held in spite of:

  • Significant advances in both GIS software and hardware computational capabilities
  • Advent of a veritable tidal wave of accessible geospatial data sets
  • The introduction of an entirely new breed of GIS software (characterized by Google Earth and Google Maps) and the accompanying rise of the so-called "neogeographer"
  • The establishment of major geospatial data sharing standards through OGC and related bodies (e.g. WFS, KML) and
  • The introduction of a several highly functional free and open source GIS software tools and libraries (e.g. as sponsored by OSGeo).

This session of iEMSs 2012 recognizes all of these advances as well as the natural role of iEMSs in encouraging, promoting, and facilitating continued advancements in the application of GIS software and tools to environmental Modelling. While any novel and interesting studies in the arena of GIS for environmental modelling will be considered for presentation in this session, we are particularly interested in the development and application of geoprocessing workflows, use of free and open source GIS, web-based GIS applications, tightly coupled GIS-based environmental models, GIS-based environmental decision support systems, geospatial data services, geoprocessing semantics, GIS integration technologies, and related topics.

Link to full papers

 

D7. Third session on data mining as a tool for environmental scientists (S-DMTES-2012)
Karina Gibert, Miquel Sanchez Marre, Joaquin Izquierdo, Ignasi Rodriguez-Roda, Geoff Holmes, Serena Chen, Antonio Ciampi, Ioannis Athanasiadis

This session is strongly linked with W-DMTES-2012, fourth iEMSs DMTES workshop (D12), and aims to approach and to promote the interaction between the environmental sciences community to the data mining community and related fields, such as artificial Intelligence, Statistics or other fields to discuss the contribution of data mining techniques to knowledge discovery in environmental sciences, as well as to make data mining techniques more accessible to environmental modellers and to give data miners and developers a better idea of the needs and desires of the environmental community.

The workshop will introduce interested parties to a range of data mining techniques and a selection of software packages. We also invite submissions of papers and presentations of interesting applications of data mining to environmental problems. New or improved techniques or methods are welcome, as well as innovative applications. Particularly welcome in this edition, contributions related with applications where it was not easy to find the right data mining technique for getting good results. For this particular contributions, please include details on unsuccessful models and the possible reasons for bad performance in the particular application addressed.

Link to full papers

 

D11. Ecological modelling and networks
Brian Fath, Caner Kazanci, Ursula Scharler

Ecosystems are complex adaptive systems. In order to effectively understand and manage these systems it is important to utilize models and analyses that can capture their holistic nature. This session will present papers dealing with use of process-based ecological models and networks, with specific attention given to various methodologies. Specifically, network methodologies which can be used to investigate the indirect relations, both quantitatively and qualitatively, between components of a complex system will be discussed and applied. Model applications to terrestrial, aquatic, and integrated socio-ecological systems will be considered.

Topics include, but not limited to:

  • Food web dynamics
  • Biogeochemical cycles and models
  • Agent based ecological and socio-ecological models
  • Resilience or collapse of ecological networks
  • Ecological community relations (i.e., mutualism, predation, competition, etc.)

Keywords: Ecological Modelling, Ecological Network Analysis, Socio-ecological systems, resilience.

Link to full papers