Keynote Talk: Giorgio Guariso
Supporting Environmental and Energy Decisions Through an Open Software Structure
Environmental awareness spreads more and more among citizens and public institutions and thus local governments have to find new tools for policy development, that take into account environmental aspects, allow a wider participation, and increase the transparency of the decisions. For instance, the local Agenda 21 process calls for a wider involvement of all stakeholders in the development of a given territory as the only possible answer to the NIMBY syndrome that risks to block any new project in countries where a minimum of public debate is allowed.
These decision support tools should frame local environmental conditions and foresee how the situation will develop, according to existing trends as well as new government actions. The way in which they are developed should also allow an easy and direct access to the information by (most of) the people involved, should prevent a rapid aging through the possibility of an easy updating, and should support experimentation and testing of alternatives hypotheses by a (large) set of different users.
We derived two basic consequences from these considerations.
First, the process of developing a software tool to support environmental policy decisions is a cyclic one. The developers and the users must work together to determine the aim and scope of the software product and to define the critical aspects of the inquired phenomena and agree on the simplification, or by-pass, of features that do not seem to be basic for the result. The potential users often do not know exactly how the problem can be managed neither they can describe exhaustively how they would like to explore and control both the real system under consideration and the software that represents the system itself and the actions taken over it. Software development must thus be approached with a very flexible method that permits a continuous interaction and prototyping, as suggested by both software engineering and ergonomics.
The second consequences is that implementation software must be as simple and widely used as possible, renouncing to more efficient, application-specific programs, which are often dedicated to specialists and may require a considerable time to familiarise with. Nowadays, the most common software environment is Office-like and most people is accustomed to navigate the Internet through html files. This is why we propose a software structure based on three components: a database with all the relevant past information on the system under study, a set of spreadsheets which allow to easily analyse alternative scenarios, a hypertext detailing hypotheses and procedures to access the database and use the spreadsheets be distributed over the Internet or as a CD-ROM, in order to enhance the level of participation.
This approach has been followed in two quite different projects.
The first is the development of the Environmental-Energy-Plan (EEP) of the province of Cremona in Northern Italy. The purpose of such a plan is to outline the local energy budget, analyse energy resources (one of the most important indicator of economic trend), evaluate related environmental aspects (indicators of life quality) and define the sectors where the local government can intervene to foster sustainable development. The project has been carried out through several meetings with the representatives of local organizations and simple citizens and, while being relative simple from a technical point of view, derives its complexity from the variety of people involved and the need to find a consensus about an important energy and environment problem,
The second project concerns the development of a system to evaluate the reduction of external costs of air pollution following from a number of actions on traffic and/or domestic heating. The PECS project (PM10 External Cost System) covers the entire Lombardy region (more than 1500 municipalities and 8 million inhabitants) and is dedicated to local authorities as well as to technical staff of various offices in the local government. It evaluates the impact on human health of PM10 pollution, analysing human activities (transportation, domestic heating), calculating PM10 emissions (primary and secondary) and estimating the related concentrations. Combining data with the distribution of population (classes and activities) and exposure indexes, the system computes health effects and external costs at a municipality level, allowing the comparison of different pollution policies. The more sophisticated part of the system is the method to transform the emission of pollutants into ambient concentrations, which utilizes the results of a complex distributed air quality model used off-line to evaluate the output of a limited number of emission scenarios.
John P. Norton (iCAM and Dept. of Mathematics, Australian National University, Canberra) and
Kenneth Reckhow (Nicholas School of Environment and Earth Sciences, Duke University, Durham, North Carolina)
The aim of this presentation is to review some of the similarities and differences between adaptive management (AM) of natural resources, the use of feedback and adaptation in control engineering, and updating of uncertain information by Bayesian analysis, which may have lessons for each other. Some fundamental difficulties and limitations, successes and failures of the three approaches are considered, and areas where there seems to be scope for cross-fertilisation are suggested. The talk is essentially on the topics addressed by Workshop 1, with more historical context.
Robert Costanza (Gordon Gund Professor of Ecological Economics and Director, Gund Institute of Ecological Economics Rubenstein School of Environment and Natural Resources The University of Vermont)
Developing the ability to model humans as integrated components in complex, dynamic, ecological systems is necessary to better understand the past, present, and possible futures of humanity. This talk will review a range of past and current integrated modeling applications at multiple time and space scales, from watersheds to global, highlighting the progress that has been made to date in building truly integrated models. It will then set out a research agenda for the future of integrated modeling. This research agenda points to: (1) the need for better shared historical databases, with which to calibrate integrated models; and (2) the need to involve humans in the modeling process in several synergistic ways, including as active stakeholders and participants in the model development process, as "players" of the models, and as integrated modeled components. We cannot predict the future of systems that include humans, but we can use integrated modeling to better understand human history and to help create a better, more sustainable and desirable future for humanity embedded in ecological systems.
Guy Brasseur (Director, Earth and Sun Systems Laboratory, National Center for Atmospheric Research, Boulder, CO, USA); and
Will Steffen (Director, Centre for Resource and Environmental Studies, Australian National University, Canberra, Australia)
The rapid changes that are now occurring in the global environment and the fact that these changes have no direct analogue in the past are placing increasing pressure on the modelling community to improve its capability to predict future change in the global environment. The simulation tools being developed to model the behaviour of the Earth as a whole are often called Earth System models. The most complex and spatially resolved of these are derived from General Circulation Models (GCMs), which originally focussed on the dynamics of atmospheric circulation and formed the basis for weather prediction models. As GCMs evolved towards full Earth System models, they incorporated the dynamics of the ocean as well as the coupling between ocean and atmosphere. More recently many complex Earth System models are including much more sophisticated atmospheric chemistry, as well as improving their characterization of the land surface. Further improvements in the near future are focused on continuing development of the land surface modules, especially dynamic vegetation, and incorporation of an interactive carbon cycle. The latter is important for simulating potentially large positive feedbacks in the carbon cycle as the climate warms. A rapidly developing aspect of complex Earth System models is their capability to assimilate large amounts of data in real time to improve their prognostic capability. Thus, collaboration between the modelling and Earth observation communities is of considerable importance.
Parallel with and complementary to the development of complex Earth System models has been the construction of Earth System Models of Intermediate Complexity (EMICs). With their coarser spatial resolution and more highly parameterised process modules, EMICs are less complex than the GCM-derived models but significantly more realistic than simple conceptual models. In addition, they are computationally much faster to run than fully complex Earth System models; thus, one of the most common uses of EMICs is to simulate paleo-environmental records that require model runs of hundreds of thousands of years. Some EMICs have achieved considerable success in simulating, for example, the abrupt climate changes in the North Atlantic region during the last glacial period and the rapid shift in vegetation from savanna to desert in North Africa during the mid-Holocene. One the major challenges to both types of Earth System model is to improve the characterisation of biological processes, now recognised to be more important in Earth System functioning than earlier thought. To this end, much effort has been placed on developing Dynamic Global Vegetation Models and so-called Green Ocean Models. The ultimate challenge, however, is to incorporate the human dimensions of global change into Earth System models in an interactive way. At present, most attempts to accomplish this coupling are unbalanced - the models are either very complex in the biophysical characterisation with simple, crudely parameterised socio-economic modules or are complex economics models with simple, "black box" representations of the biophysical world.
Helena Mitasova (NCSU, Raleigh, NC, USA)
Markus Neteler (ITC-irst, Trento, Italy)
Over the past decade free and open source software (FOSS) has moved into mainstreem and its growth includes a wide range of geospatial software tools. The presentation will briefly describe the principles of FOSS and provide an overview of major FOSS projects that support management, processing, analysis, visualization and on-line distribution of georeferenced data. Interoperability between the different tools enabled by the FOSS concept will be highlighted using examples of applications that combine software developed by several projects, such as GRASS, Mapserver, GDAL, PROJ, R-stats, PostGIS, paraview and others. Special focus will be given to the recent new developments in Geographic Resources Analysis Support System (GRASS, http://grass.itc.it). Since its original design as a land management tool for military installationsthe system has evolved into one of the most comprehensive, general purpose GIS and support for environmental modeling has been an integral part of its development. Overview of the GRASS6 capabilities relevant to environmental modeling will be provided, including tools for working with 2D and 3D raster data, the completely redesigned topological 2D/3D vector engine, vector network analysis and SQL-based attribute management. Various graphical user interfaces, such as QGIS and JAVAGRASS, that are being used with GRASS will be introduced and enhancements to 3D dynamic visualization will be presented. Environmental modeling will be illustrated by examples in the area of coastal management and sediment pollution control using the latest lidar data and new or updated GRASS6 modules. The talk will conclude with discussion of visions for the future of FOSS geospatial modeling.
Prof. Svirezhev will not be able to attend the Conference.
The linear world of classic mathematical physics was harmonious and consistent. Almost the entire evolution of this world proceeded in small neighbourhoods of stable equilibrium where linearisation principle held true. This was a smooth differentiable world, in which there was no room for instabilities, catastrophes, and other inconvenient phenomena. Certainly, the complete harmony did not exist and nature produced now and then surprise packets with unpleasant non-linearities; nevertheless, the situation was saved due to the fact that those non-linearities could be regarded as minor ones, so that one could manage with small non-linear additions to the main linear solutions. To explain all those non-linear phenomena, viz. jumps, discontinuities, catastrophes, hysteresis, and dynamic chaos, was the task to philosophy rather than mathematics.
In contrast to many fields of mathematical physics, where linear models have been used very extensively and effectively (recall, for instance, the wave equation, the heat conduction equation, the Schrdinger equation, etc.), mathematical ecology is principally non-linear science. The fact is that almost all interactions in ecology, both competitive and trophic, are non-linear. Perhaps, the only linear model is the model of exponential growth by T. Malthus. Even the classic Volterra "prey - predator" model demonstrates a typical non-linear pattern, while in more sophisticated models, for instance, in trophic chains, we can see non-linear oscillations, "quantum" effects, and dynamic chaos.
A vast variety of non-linear problems generated by ecology is described in many books, and there is not need to repeat their descriptions here; I would like to dwell in more detail on the problems yet unsolved, trying to predict or, strictly speaking, to guess the potential results and offer their informal interpretation.