Naturnet newsletter05 1


Published on

  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Naturnet newsletter05 1

  1. 1. Naturnet - Redime NEWSLETTER No. 5 April 2007 ContentNew Release of the Garp3 Collaborative Danube Delta Biosphere Reserve (DDBR) isWorkbench one of the five case studies considered in the FP6 NaturNet-Redime project to develop QualitativeAnders Bouwer, Jochem Liem, Floris Linnebank Reasoning (QR) models…and Bert Bredeweg (University of Amsterodam) Page 7 A new version (1.3.8) of the Garp3 CollaborativeWorkbench has been released for general use, as of Danube Delta Biosphere Reserve Case StudyApril 18th, 2007. Garp3 is a tool for Qualitative Qualitative Reasoning Model - an EnvironmentalReasoning and Modelling that supports domain experts System Sustainable Development Decision Makingand novices in building conceptual models of system Supportbehaviour … Eugenia Cioaca (Danube Delta National Institute Page 2 Tulcea, Romania), Bert Bredeweg (University of Amsterdam), Paulo Salles (University of Brasilia)Troubleshooting Qualitative Models Qualitative Reasoning (QR) models may serveJochem Liem and Bert Bredeweg (University of as mediators in decision making process onAmsterodam) Sustainable Development. For that purpose, a QR This paper describes the issues modellers model can be said to capture (to document) anencounter when developing qualitative models, and explanation of how a system behaves...initial work started in the NaturNet-Redime (NNR) Page 9project to automatically detect and solve these issues.Ecologists working on different case studies in the Curriculum for Learning about SustainableNNR project are developing qualitative models in which Development Using Qualitative Reasoningthey capture their knowledge about how river ecology Tim Nuttle (University of Jena, Germany), Paulorelates to sustainable development... Salles (University of Brasilia, Brazil), Bert Page 4 Bredeweg (University of Amsterdam, The Netherlands), Eugenia Cioaca (Danube DeltaA qualitative model about sustainability in the National Institute, Romania), Elena Nakova (CentralBrazilian Riacho Fundo water basin Laboratory of General Ecology, Bulgaria), RichardPaulo Salles (University of Brasilia) Noble (University of Hull, England), Andreas Zitek (Vienna University of Applied Life Sciences, A qualitative model exploring sustainability Austria).issues in the Riacho Fundo water basin (Brasilia,Brazil) was developed by the Naturnet – Redime Work Package 6 partners have been hard atproject and in the next months this model will be work producing Qualitative Reasoning (QR) modelspresented to stakeholders for evaluation. The model and simulations of sustainability issues pertaining togoal is to support stakeholders in understanding and case studies in different river catchment areas...decision-making about water resources management... Page 6 Page 13Qualitative Reasoning Model – the modelled Events of interestsystem functioning study: Danube Delta BiosphereReserve case study Page 15Eugenia CIOACA (Danube Delta National Institute ContactTulcea, Romania), Bert BREDEWEG (University of Page 16Amsterdam), Paulo SALLES (University of Brasilia),Tim Nuttle (University of Jena, Germany) České centrum pro vědu a společnost, Radlická 28/263, 150 00 Praha 5, Czech Republic ISSN 1801-6480 <META NAME="DC.Identifier" CONTENT="(SCHEME=ISSN) 18016480">
  2. 2. New Release of the Garp3 Collaborative Workbench Anders Bouwer, Jochem Liem, Floris Linnebank and Bert Bredeweg (University of Amsterodam) A new version (1.3.8) of the Garp3 model fragments in the target model whereverCollaborative Workbench has been released for possible, instead of creating extra copies of eachgeneral use, as of April 18th, 2007. Garp3 is a tool for conditional model fragment. It is also possible to copyQualitative Reasoning and Modelling that supports more complex modeling concepts (i.e., refinements)domain experts and novices in building conceptual than was previously possible.models of system behaviour [1]. A model in Garp3consists of model fragments (specifying situations orprocesses) and scenarios, which can be simulated to The Sketch environment has been improved ingenerate predictions. The result of a simulation is a several respects, based on requests from Garp3graph consisting of qualitative states and transitions users in the NNR project. The new version allowsinvolving changes in quantity values and/or saving and loading sketches, so that users can edit(in)equality statements, representing behavioural and store multiple versions of each type of sketchchange. Garp3 consists of several screens and (and export them as postscript figures). The interfaceeditors that are organized into three environments: for editing inequalities, quantities and values in the(1) the Build environment, which provides editors to sketch state-transition graph editor has beenbuild the basic model ingredients (e.g., entities, improved to better handle the different kinds ofquantities, values), model fragments, and scenarios; (in)equality statements. Other improvements relate(2) the Sketch environment, which provides editors to to the import functionality, which allows importingcreate preliminary sketches (concept map, structural parts of one sketch to another sketch, following themodel, causal model, and expected state-transition modeling framework described in [2].graph); and (3) the Simulate environment, whichprovides control over the simulation and views on thesimulation results (e.g., state-transition graph, values The Simulate environment (see figure 1) hasand dependencies per state, value and inequality been improved in several ways to address twohistories over multiple states). Compared to the issues: (1) in complex simulations, users had troubleprevious release version, a number of substantial finding the possible behaviours; and (2) usersimprovements have been made to each of the three sometimes need more control over the simulationenvironments, as well as to the simulation tracer, and engine to manage the complexity of the simulation.the qualitative reasoning portal, as described in the To address the first issue, the functionality forparagraphs below. searching behaviour paths has been improved by using a more efficient search algorithm, and adding options for searching for end states, cyclic paths or In the Build environment, the copy functionality complete paths (from a begin state to an end state, orhas been improved to facilitate sharing parts of an end loop) in a state-transition graph. To addressmodels more efficiently, by reusing existing the second issue, the inequality reasoning has beenknowledge definitions. For example, when pasting a improved, and the set of preferences has beenmodel fragment that includes other model fragments revised, to provide better, and more intuitive controlinto another model, Garp3 will now reuse existing over the reasoning of the simulation engine. 2
  3. 3. Figure 1. A screenshot of Garp3, illustrating the new functionality related to simulation control and search in the state-transition graph. The tracer functionality provides insight into that whenever the user clicks on the Help-icon (thethe reasoning of the Garp3 simulation engine. The owl) in Garp3, the relevant help page for the currentlytracer presents, among other things, information active screen will appear. To support novice users inabout model fragments that are being considered, acquiring the vocabulary of qualitative reasoning andinequality reasoning being performed, and modeling, the terms used in the Garp3 interface areterminations, which possibly lead to state transitions. described in a glossary.The tracer has been completely revised to providemore information, in a better readable format, andalso includes a new general/critical category to Visit the QRM portal website athighlight the most important feedback from the for further details and tosimulation engine, which makes it easier for users to download the most recent version of the Garp3diagnose problems with their model. Furthermore, the software.tracer works significantly faster than before. The QRM (Qualitative Reasoning and ReferencesModelling) portal website provides an overview of the [1] J. Liem, A. Bouwer, and B. Bredeweg, 2006.Garp3 software, documentation, and related research Collaborative QR model building and simulationpublications, as well as example models and results workbench, Naturnet-Redime, STREP project co-from the NNR case studies described elsewhere in funded by the European Commission within the Sixththis newsletter. The major new addition is an Framework Programme (2002-2006), Project no.extensive collection of help pages about the Garp3 004074, Project Deliverable Report D4.3.collaborative workbench, explaining all of itsfunctionality. Each help page contains, among other [2] B. Bredeweg, Salles, P., Bouwer, A., and Liem L,things, a description of the task at hand, the context, 2005. Framework for conceptual QR description ofrelated tasks, and, in most cases, an example with a case studies, Naturnet-Redime, STREP project co-screenshot. The help pages can be accessed directly funded by the European Commission within the Sixthfrom Garp3 (provided an internet connection and Framework Programme (2002-2006), Project no.browser) in a context-sensitive manner. This means 004074, Project Deliverable Report D6.1. 3
  4. 4. Troubleshooting Qualitative Models Jochem Liem and Bert Bredeweg (University of Amsterdam) This paper describes the issues modellers One of the main issues encountered byencounter when developing qualitative models, and modellers developing qualitative models is dealinginitial work started in the NaturNet-Redime (NNR) with undesired simulation results. For example,project to automatically detect and solve these simulating a scenario may result in having no states,issues. Ecologists working on different case studies an end state could contain a changing quantity thatin the NNR project are developing qualitative models has not yet reached its final value, and magnitudesin which they capture their knowledge about how and derivatives could be missing. Furthermore,river ecology relates to sustainable development. The simulations may result in too many states, somecaptured knowledge is organized in model fragments behaviour could be missing, and other behaviourthat describe small parts of the structure and might be undesired. It is often difficult to determinebehaviour of a system. The main purpose of model which model ingredients are responsible for an issue,fragments is the introduction of causal relations and equally difficult to figure out how and whichbetween quantities given a certain system structure. model ingredients should be changed to solve the issue. Currently, modellers ask qualitative reasoning Qualitative models allow these ecologists not experts to help them troubleshoot their models.just to formalise their knowledge, but also to discover However, at the end of this project, the softwarethe implications of that knowledge through simulation. should support modellers well enough for them toAn initial situation of the system (called a scenario) solve issues without the help of qualitative reasoningcreated by the modeller serves as input for the experts anymore.simulation engine that generates the behaviour of themodelled system. The behaviour has the form of a Consider the following example from a casestate graph, and shows how the quantities in the study describing how river planning affects thesystem change in time. The simulator works by recruitment of salmon through different life states.determining which model fragments apply to the Figure 1 shows a scenario that represents a riversystem described in the scenario and augments the containing a population of salmon divided intoscenario with the causal relations introduced by these different life stages. Each part of the population livesmodel fragments. These causal relations make it in the same river (according to this model), and eachpossible to calculate the value of the derivatives of life stage migrates into the next life stage. The part ofthe quantities, and thus the successor states of the the population in the first life stage has a high numberinitial state can be determined. For each of the of individuals, while the other life stages have asuccessor states the same procedure is performed to normal amount of individuals.create a complete state graph. Figure 1: A scenario describing the number of individuals of a population in different life stages living in a river. 4
  5. 5. An expert created a model fragment capturing stage is bigger than the number of individuals in thehis knowledge about how different parts of the second stage, the recruitment rate is positive. If theypopulation recruit from one life stage into the next are equal, the rate is zero. And if the second stage is(see Figure 2). The model fragment requires two bigger, the rate is negative. The influence (I+) setsparts of a population to inhabit a river. The first life the derivative of the second number of individuals tostage should migrate into a second life stage. When positive if the recruitment rate is Plus, negative if it isthese conditional model ingredients are present, new Min, and Zero if it is zero. The proportionality (P+)knowledge is added to the scenario. A recruitment sets the derivative of the recruitment rate torate quantity is introduced, and its value is calculated increasing if the number of individuals of the firstby subtracting the number of individuals in the first stage increases, to decreasing if it decreases, and tostage from the number of individuals in the second stable if it is stable.stage. When the number of individuals in the first Figure 2: The recruitment model fragment representing the causal relations involved in the recruitment process. When simulating the scenario, the model Associating the recruitment rate to one of thefragment fits on the scenario 3 times (on the eggs population stages instead of the river can solve theand the juveniles, on the juveniles and the smolts, issue, as it would result in a recruitment rate quantityand on the smolts and the adults). However, the to be introduced for each interaction.simulator generates no states. This is a typicalexample of a modelling issue that is difficult to solve In the last half-year of the NNR project, we aimfor the modeller. When a scenario is simulated and to develop model-troubleshooting support within 1no model fragments are active, there is only a single Garp3 that will automatically detect these kind ofstate in the state graph, i.e. the state describing the issues within models and suggest possible solutions.scenario itself. When there are no states, the We will analyse the intermediate modelling resultsintroduced knowledge causes a contradiction, e.g. created by the modellers working on case studies totwo incompatible inequalities are true, or a value has determine what the most frequently occurring issuesmultiple values. are, and how they could be solved. Furthermore, we are looking at existing model-based diagnosis In this case, the issue with the model fragment literature to see how existing methods can contributeis that the recruitment rate is associated to the river. to a solution. Model troubleshooting support shouldSince there is only a single river, only a single make it easier for domain experts to capture theirrecruitment quantity is introduced, even though there knowledge about sustainable development.are three interactions. Calculating the recruitmentrate for the juveniles to the smolts and the smolts tothe adults yields a value zero. However, for the eggsto the juveniles the recruitment rate yields the valuePlus.Therefore, the single recruitment rate quantity is setto both the value Plus and the value Zero, which isimpossible. Therefore no states are generated. 1 5
  6. 6. A qualitative model about sustainability in the Brazilian Riacho Fundo water basin Paulo Salles (University of Brasilia) Qualitative Reasoning models may be useful in this A qualitative model exploring sustainability case. An evaluation study with Latvian students,issues in the Riacho Fundo water basin (Brasilia, reported in the Naturnet-Redime Newsletter lastBrazil) was developed by the Naturnet – Redime November (Bredeweg et al. 2006) confirms theproject and in the next months this model will be potential of such models to communicatepresented to stakeholders for evaluation. The model sustainability concepts. The qualitative modelgoal is to support stakeholders in understanding and developed by the Naturnet – Redime team for thedecision-making about water resources Riacho Fundo focuses on the following aspects: (a)management. Relevant concepts were selected the influence of urban drainage systems on soilaccording to stakeholders opinions, and are conditions, economic activities and human well being;expressed in the model with everyday vocabulary (b) the dynamics of erosion and water infiltration inand diagrams, in order to make them accessible to the soil, and their influences on springs, streams, andthe general public. This article explains some economic activities in semi urbanized areas; and (c)features of the Riacho Fundo basin addressed by the the effects of soil protection on springs and rivermodel and how the evaluation process will be done. conditions, biodiversity and agricultural production. The Riacho Fundo is a small water basin in The Riacho Fundo model will be evaluated in aBrasília, Brazil. Since the new capital was built in the series of meetings with members and peoplelate 50’s, it has been impacted by changes of natural involved in the WBC – Paranoá river. A generallandscapes into rural and urban areas. The most workshop is being organized to present the results ofimportant economic activities in the basin are those the Naturnet-Redime project. Next to that, therelated to services, including business offices, Naturnet – Redime team will run a number ofcommerce and car repair garages. Industrial activities evaluation sessions with volunteers, in which differentinclude food, dyed fabric and clothes production. aspects of the model will be actually evaluated. GivenVegetables and corn production are the main its potential for educational purposes, some of theseactivities for most of the farmers, and cattle, pork and meetings will include teachers and students. As insheep are the most important herds in the Riacho similar activities previously done, qualitative modelFundo region. Agriculture represents a small part of evaluation starts with discussions aboutthe regional GDP, but it is important as it contributes representations of concepts and simulation keep green areas and parts of natural vegetation. Using diagrammatic representations of cause-effectAs a consequence of these economic activities, soil relations encoded in the model as the basis forand water contamination with pesticides, industrial representing knowledge, evaluators are asked toeffluents and organic matter have been reported. discuss how to solve conflicting situations, and toBesides that, urbanization has caused deforestation draw (using paper and pencil) alternative diagrams toand loss of biodiversity and erosion resulted in the express their own views. Finally, the evaluators givedisappearance of springs and small streams. their opinion in questionnaires and interviews about According to Brazilian legislation, decisions on different aspects of the model, such as the clarity ofwater management are made in Water Basin concepts represented in the model, significance forCommittees (WBC), institutions that gather improving understanding and usability of the modelrepresentatives from the government, productive as a tool to support decision-making. The results ofsectors and civil society. Riacho Fundo water the evaluation will be described in the next issue ofresources are now managed by the Paranoá river the Naturnet-Redime Newsletter!WBC, recently created after five years of Reference:stakeholders and community mobilization. Since thepreparatory meetings, the WBC has been the locus Bredeweg, B.; Salles, P.; Bouwer, A.; Bakker,for discussions about what is required for the Riacho E. and Liem, J. (2006) Successful use of Garp3 byFundo basin to be sustainable. From these activities, stakeholders. Naturnet-Redime Newsletter, numberit became clear that there is a need for tools to 4, November 2006, pp.: 10 – 12. ISSN 1801-6480.improve understanding of the complex problems theWBC has to decide upon, and to express people’sideas and proposals using a formal language. 6
  7. 7. Qualitative Reasoning Model – the modelled system functioningstudy: Danube Delta Biosphere Reserve case study Eugenia CIOACA (Danube Delta National Institute Tulcea, Romania), Bert BREDEWEG (University of Amsterdam), Paulo SALLES (University of Brasilia), Tim Nuttle (University of Jena, Germany) Danube Delta Biosphere Reserve (DDBR) graphics. They are focused on describing theis one of the five case studies considered in the system’s features related to those causes and theirFP6 NaturNet-Redime project to develop effects which delimit the sustainable developmentQualitative Reasoning (QR) models. actions. The three conceptual structures (ConceptModel System: DDBR, located at the mouth of the map, Causal model, and Structural model) focus onDanube River before it reaches the Black Sea, has water pollution negative effect propagation froman area of 5,800 sq. km, making it one of the water to aquatic biota (plants, zooplankton,greates t wetlands in the world. macroinvertebrates, fish, birds) and ultimately to human health. In order to construct a model of a real systemusing the Qualitative Reasoning concept [1], it is Concept map is presented in Figure 1. Itnecessary to understand this system functions. It contains the modelled system components and therequires a good knowledge of the system functional relationships among them, in thecomponents, their structural and cause-effect framework of basic physical, physical-chemical andrelationships, which leads to a better understanding biological processes. The DDBR main componentsof the overall system’s behaviour. This is achieved are aquatic ecosystems. Their behaviour dependsas a preliminary study necessary in preparing the on water quality. The concept map stresses theQualitative Reasoning (QR) model ingredients. It Water pollution process. This process “negativelyfollows the Framework for conceptual QR affects” both aquatic biological components statusdescription of case studies [2], through Concept and human health. The influences can be direct ormap, Global causal model, and Structural model indirect. Figure 1. Concept map for Danube Delta Biosphere Reserve water pollution.Full Structural model, for the two types ofecosystems terrestrial and aquatic, is shown inFigure 2. It contains the system entities and their Structural relationships hierarchical organised [3, 4,structural relationships hierarchical organised [3, 5], as they are necessary to be later implemented in Garp3 – QR specialised software.4,5], as they are necessary to be later implementedin Garp3 – QR specialised software. 7
  8. 8. Figure 2. Structural model of the Danube Delta Biosphere Reserve aquatic ecosystems. Global Causal model contains the modelled them are presented in Figure 3: DDBR global system quantities and the causal relationships causal model. Influence (I+, positive rate or I-, among them, governed by active processes. In negative rate) indicates a direct effect of a process total, 12 processes are active in the DDBR aquatic rate on other quantity, causing this quantity value environment, which influence the quantities that change: increase or decrease, respectively. represent the amount of each organism. Changes Proportionality indicates an indirect effect of a in these amounts propagate to other quantities, as quantity change (P+, quantity increase, or P-, effect of one organism group behaviour to another quantity decrease) that causes other quantity value one. The quantities and causal dependencies change: increase or decrease, respectively. (Influence : I+/I- or Proportionality: P+/P-) among Figure 3. Global Causal model for Danube Delta Biosphere Reserve aquatic ecosystems water pollution. References 2. B., Bredeweg, P., Salles, A., Bouwer, and J., Liem 2005. Framework for conceptual QR1. B., Bredeweg, 2005 – Qualitative Reasoning. description of case studies, NaturNet-Redime, NaturNet-Redime NEWSLETTER No. 1, pp. 3-5. Project Deliverable report D6.1. 8
  9. 9. 3. E., Cioaca, B., Bredeweg, T., Nuttle, 2006 - The basic biological, physical, and chemical Danube Delta Biosphere Reserve: a Case Study processes related to the environment for Qualitative Modelling of Sustainable th Development. QR-06 20 International section/Deliverables list Workshop on Qualitative Reasoning 5. T., Nuttle, et al, 2006 - Progress in developing PROCEEDINGS, pp.151-156. Dartmouth cognitive models of sustainability: case studies College, Hanover, New Hampshire, USA. from European and Brazilian river basins.4. E., Cioaca, S., Covaliov, L., Torok, Ibram, O., NaturNet-Redime NEWSLETTER 2, pp.12-15. 2006 - D6.2.1 Textual description of the Danube Delta Biosphere Reserve case study focusing on Danube Delta Biosphere Reserve Case Study Qualitative Reasoning Model - an Environmental System Sustainable Development Decision Making Support Eugenia Cioaca (Danube Delta National Institute Tulcea, Romania), Bert Bredeweg (University of Amsterdam), Paulo Salles (University of Brasilia) Qualitative Reasoning (QR) models may DDBR QR model consists of the followingserve as mediators in decision making process on ingredients: 18 entities, hierarchical structured asSustainable Development. For that purpose, a QR shown in Figure 1; 17 Scenarios, as shown inmodel can be said to capture (to document) an Figure 2, and 57 Model fragments. The descriptionexplanation of how a system behaves. The of the Scenarios and Fragments, and Simulationdocumentation has two distinguished parts: QR results, presented in this paper, refers to the DDBRmodel components description as they are aquatic ecosystems water pollution process.represented in Garp3 software [1] and the modelscenarios simulation results. This paper presentsa summary of the Danube Delta Biosphere Reserve Scenarios(DDBR) QR model documentation [2]. Within the To model the DDBR water pollution process,DDBR area, aquatic ecosystems are the mostextensive and important natural resource of this there 3 scenarios have been constructed. Here is presented the DD Water pollution and DD aquaticarea, both from the biodiversity and the humanbeing perspective. Due to its impacts, water biodiversity Scenario (Figure 3.). It concerns thequality is a central focus of concern from the modelled system’s Entities, their associatedStrategy of Sustainable Development point of view. quantities, the Configurations among EntitiesNevertheless, water pollution is a major problem in participating in this process. It configures the process negative effects and the ways itthe DDBR and has caused loss of biodiversity overthe last several decades. Additionally, human propagates to aquatic biotic component (Aquaticpopulation relies on DDBR resources for fresh biological entities). This Scenario contains:water and food, so they are implicitly negatively 1. The modelled systems external influencesinfluenced by water pollution. To progress and (Agents): Agriculture: Nutrient run-off (whichreduce these negative effects, it is necessary to participates in the water pollution process onlyunderstand and communicate how water pollution if in high content. For values equal or smalleraffects biodiversity and human health in DDBR. The than Medium it participates in Plant growthDDBR QR model is therefore constructed to model process as main food resource for any Plantthe behaviour of DDBR aquatic ecosystem flora species) and Industry: Heavy metals (whichand fauna populations as well as human health, have the property of bioaccumulation in anyindicator of people living in or around this area. The aquatic biological organism, leading to thatbehaviour of these components is modelled as entity pollution, even Mortality if inrelated to the basic aquatic ecosystem physical, Medium/High concentration in water).chemical and biological processes. The main 9
  10. 10. Figure 1. Entity hierarchy of the DDBR system. Figure 2. The DDBR system QR model Scenarios.2. The third water pollution component is given, 3. To reduce the simulation complexity, mainly by Cyanotoxins. They are produced in Assumptions are introduced in the Scenario water if there is a content of some poisoning construction: “Assume nutrient consumption is species of Blue-green algae (Cyanobacteria), zero and steady”, “Assume Migration is zero which contain Cyanotoxins in their cells. and steady”, and “Assume Production is medium and steady”. Figure 3. DD Water pollution and DD aquatic biodiversity ScenarioModel fragments In the Qr software that we use there are A process is presented in this paper in Figure 4. DDthree types of Model fragments: Static, Process, Water pollution and DD aquatic populationand Agent. biodiversity. 10
  11. 11. Figure 4. DD Water pollution and DD aquatic population biodiversity - Process MF. It figures the structural and behaviouralrelationships between River delta and the Aquatic 2. A part of Danube Delta: Nutrient inflow stayspopulation, related to River delta: Water pollution in the system and contributes to Nutrientinfluence on Aquatic population components available for Plant species growth whilebehaviour, as follows: positive influence (I+) on another part is lost (Nutrient net loss), eitherAquatic population: Mortality and negative through Nutrient outflow or Nutrientproportionality (P-) on Aquatic population: consumption (by aquatic Plant species only).Biodiversity. It emphasizes the causal conditionswhich have been generating the loss of DDBR 3. So happens with Danube Delta: Heavybiodiversity, in order to delimit those objectives for a metals inflow.sustainable use of natural resources and a 4. A part of Nutrient net loss and Heavy metalsSustainable Development Strategy addressing theaquatic ecosystems. net loss is recycled from dead organic matter as result of Particulate Organic matter DDBR QR Model output: Scenarios bacterial decomposition (Pom bact decomp)simulation results. process. One of the scenario simulation results is 5. Danube delta: Water pollution rate is thepresented for the DD Water pollution and DD result of three main water pollutants: Danubeaquatic population biodiversity Scenario. This is the delta: Nutrient available, Heavy metalsDependency diagram (Figure 5), which provides available and Cyanotoxins;information on structure and causality (Influence I, 6. Danube delta: Water pollution rate has aor Proportionality P), correspondence (Q, dQ) and direct positive influence (I+) on any Aquaticin/equality (=, >, <) dependency relationshipsamong the system’s water pollutants (Nutrients, biological entity: Mortality.Heavy metals and Cyanotoxins), any Aquatic 7. Aquatic biological entity: Mortality has anbiological entity, involved in Water pollution process indirect negative influence (P-) on anyand its effect on Aquatic populations, as follows: Aquatic biological entity: Biomass. 1. Danube River: Nutrient inflow and Heavy Metals inflow main resources are the two system’s external influences (Agents): Agriculture and Industry, localised “In catchment area of” the River. There is a close relationship (P+ and Q) between Nutrient run-off from Agriculture lands and Nutrient that enters the Danube River. The same relationship occurs between Heavy metals run-off from Industry and Heavy metals entering the Danube River. From the River, these two main water pollutants reach the Danube Delta aquatic ecosystems. 11
  12. 12. Figure 4: Dependency diagram of the Water pollution process.Knowledge about the aquatic ecosystems behaviour Referenceswithin the DDBR system, as it is presented in the B., Bredeweg, A., Bouwer, J., LIEM, 2006 – Garp3-DDBR QR Model, may supportdecision-making Workbench for capturing conceptual knowledge.activities on natural resources protection, as well as NaturNet-Redime NEWSLETTER No. 2, pp. 2-4.achieving the environmental SustainableDevelopment within this system. E., Cioaca, B., Bredeweg, P., Salles, 2006 - D6.2.2. Qualitative Reasoning Model and documentation for learning about sustainable development focusing on basic biological, physical, and chemical processes related to the environment in the Danube Delta Biosphere Reserve, list/. 12
  13. 13. Curriculum for Learning about Sustainable Development Using Qualitative Reasoning. Tim Nuttle (University of Jena, Germany), Paulo Salles (University of Brasilia, Brazil), Bert Bredeweg (University of Amsterdam, The Netherlands), Eugenia Cioaca (Danube Delta National Institute, Romania), Elena Nakova (Central Laboratory of General Ecology, Bulgaria), Richard Noble (University of Hull, England), Andreas Zitek (Vienna University of Applied Life Sciences, Austria). Work Package 6 partners have been hard at 1. Goal: the expected learning outcome ofwork producing Qualitative Reasoning (QR) models completing this assignment.and simulations of sustainability issues pertaining to 2. Introduction: a textual description of thecase studies in different river catchment areas. system and its key features and issues.Models from the case studies for the River Mesta(Bulgaria), Danube Delta Biosphere Reserve 3. Overview: a list of the main concepts that will(Romania), and Riacho Fundo (Brazil) are completed be covered in the exercises.and models for the River Kamp (Austria) andYorkshire River Ouse and River Trent (England) are 4. Exercises: a series of questions about theprogressing well according to plan. Papers from each model system and instructions on how to interact with the model to discover theof these cases studies are being presented in a answers.special session dedicated to the NaturNet-Redime stproject at the 21 International Workshop on Here, we provide an example of some of theQualitative Reasoning, to be held 27-29 June 2007 in questions that learners would seek to answer whileAberystwyth, Wales ( exploring the Riacho Fundo case study: These case studies focus on different parts of Sce ur01a: What happens when there is nothe sustainability domain and will be integrated into a drainage in a city?common curriculum where learners can interact withmodels to learn about cause and effect in sustainable Open scenario Sce ur01a drainage zerodevelopment scenarios. The learning endeavor will constant and run the simulation (for the initialbe supported by a curriculum and assignments that state/s only).guide learners through the learning process. Q1. (structure) Explain in your own words the Assignments for exploring QR models from system being modeled by this scenario.each case study will follow a newly developed Hint: Select state [1] and open the entitystandard assignment template, which can be relations view. Rearrange the elements asreapplied to future models as they are developed. necessary.This template follows basic principles in pedagogy,where learning progresses from basic levels focusingon understanding definitions of terms (i.e.,“knowledge” in Bloom’s Taxonomy of cognitiveunderstanding), through application of knowledge, tosynthesis and evaluation of sustainable versusunsustainable scenarios. This assignmentorganization also progresses through the hierarchicallevels of the model, from structure (What are thecomponents of the system and how are theyrelated?) to causality (How do the relations betweencomponents affect each other?) to dynamics (How dothe causal relationships cause the system to changethrough time?). As with the other NaturNet-Redime learningmaterials, assignments for learning aboutsustainability in each of the case studies will bepresented using Moodle software, an open sourcecourse management system that applies soundpedagogical principles to help educators create Figure 1. The Garp3 workbench and state grapheffective online learning communities for the Scenario (top) and the Entities Relations( Each assignment is structured as view for state [1] (bottom).follows: 13
  14. 14. After following the instructions (and the hint),the learner would see that the system being modeledis one where rainfall falls on the urban Riacho Fundoand where there is no drainage (Figure 1). Thelearner would progress through several morequestions (not shown) to explore structure andcausal relationships that result from that structure.Then learners are asked to make predictions basedon that causality: Q6. (causality) Predict what will happen throughout the system and why. Will the system evolve towards a final end point? What will that end point be? Hint: Display the causal dependencies for state [1]. Reason through the causal dependencies to see how they cause and propagate change through the system. Figure 3. Value History for the path (Figure 1 left) showing the change in the modeled scenario. Learners would see that with a medium amount of water runoff and zero drained water, there is uncontrolled flow (value plus) in the urban Riacho Fundo (three value histories on the left of Figure 3). When there is uncontrolled flow, this causes the amount of uncontrolled water to increase. As uncontrolled water increases in the catchment, flooded area increases as does the amount of economic damage. Figure 2. Causal dependencies from state [1]. This simple scenario demonstrates how the model of the Riacho Fundo case study transmits the expert knowledge contained in the model to the Learners would follow the instructions, learner. Subsequent questions in the assignmentproducing Figure 2, and write a paragraph describing lead the learner to evaluate whether this system istheir predictions. They would then compare their sustainable or unsustainable, and how sustainabilitypredictions to those produced by the Garp3 of the system might be better achieved. Furtherreasoning engine: assignments address other scenarios in the Riacho Q7. (dynamics) Describe what the model Fundo, including more possibly more complex predicts to happen in the system over time. scenarios that involve more processes, as well as point to other models and case studies where related Hint: Run the full simulation. Note the general concepts can be explored. nature of the state graph (cycles, branching paths, etc.). Once the assignments for each case study model are completed, they will be evaluated by Select a path. View the value history, stakeholder and student groups to assess the degree focusing on what happens to the state to which they help meet the learning goals variables. Repeat for several paths (if there established in the assignment. are more than one). Look for assignments for each of the case studies to appear on the NaturNet-Redime web portal ( in the coming months. 14
  15. 15. Events of interest9th – 11th May, Maputo 12th June – 15th June 2007, Brussels IST-Africa 2007 Conference and Exhibition is Green Week will provide a unique opportunitythe second in an Annual Conference Series which for debate, exchange of experience and best practicebrings together delegates from leading commercial, among non-governmental organisations, businesses,government & research organisations across Africa various levels of government and the public,and from Europe, takes place in Maputo, Brussels, BelgiumMosambique, South Africa ml10th – 11th May 2007, Dresden 27th June – 29th June 2007, Aberystwyth th The 6 Saxonian GI Strategy – Forum GI QR07: 21st Annual Workshop on Qualitative2007. Dresden, Germany. European week Reasoning, Aberystwyth, U.K e.html15th – 16th May 2007, Prague 2nd – 5th July 2007, Glasgow Information systems in agriculture and forestry2007, XIII. European conference, themed Living EFITA/WCCA 2007 , Environmental and ruralsLabs Prague, Czech republic. sustainability through ICT, Glasgow Caledonian University, Glasgow, Scotland May 2007, Prague 4th – 6th July 2007, Porto FP7 Information Day - will be specifically thdedicated to FP7 opportunities in 2007 in the fields of 13 GI & GIS Workshop. GISDIs forICT for the environment and ICT for energy environmental analysis, monitoring, and modelling;efficiency, Prague, Czech Republic environmental protection, natural hazards prediction and mitigation, and other applications areas , Porto, Portugal May – 24th May 2007, Prague The International Symposium on 4th – 9th November 2007, PapeeteEnvironmental Software Systems. ISESS 2007 is partof a series organized by IFIP (International MUTL 2007, International Workshop on MobileFederation for Information Processing) Working and Ubiquitous Technologies for Learning, Papeete,Group 5.11 "Computers and Environment". Crowne French Polynesia (Tahiti)Plaza Hotel Prague, Czech 15
  16. 16. SIXTH FRAMEWORK PROGRAMME Educational programmes on social, economic, and environmental tools for the implementation of the EU Strategy on Sustainable Development at both EU and international levels NATURNET-REDIME New Education and Decision Support Model for Active Behaviour in Sustainable Development Based on Innovative Web Services and Qualitative Reasoning Project no. 004074 Instrument: SPECIFIC TARGETED RESEARCH PROJECT Thematic Priority: SUSTDEV-2004-3.VIII.2.e Start date of project: 1st March 2005 Duration: 30 months Web: Project officer Patrizia Poggi Patrizia.POGGI@ec.europa.euProject co-funded by the European Commission within the Sixth Framework Programme (2002-2006)Co-odinator Contact person Country E-mailCzech Centre for Science and Society Karel Charvat CZ ccss@ccss.czPartner Contact person Country E-mailEnvironmental Network Ltd Peter Barz UK pkab@env-net.com University of Freiburg Barbara Koch DE barbara.koch@felis.uni-freiburg.de of Francavilla di Sicilia Nino Paterno IT npat@stepim.it Bozeny Nemcove Jan Sterba CZ sterba@gybon.cz Grenzüberschreitendes Netzwerk e.V Frank Hoffmann DE 320052219146@t-online.dehttp://www.IGN-SN.deInstitute of Mathematics and Comp. Science, Univ. of Latvia Maris Alberts LV alberts@acad.latnet.lv Research Alexander Almer AT alexander.almer@joanneum.at - Municipality Juris Salmins LV krimulda@rrp.lvAPIF MOVIQUITY SA Wendy Moreno ES wmp@moviquity.com TIC de la Collectivité Territoriale de Corse Jerome Granados FR jerome.granados@mitic.corse.frhttp://www.mitic.corse.frRegion Vysocina Jiri Hiess CZ Hiess.J@kr-vysocina.czUniversity of Jena, Institute of Ecology Tim Nuttle DE Tim.Nuttle@uni-jena.de Michael Neumann mn@mneum.deHuman Computer Studies Laboratory, Informatics Institute, Bert Bredeweg NL of Science, University of Amsterdam Academy of Sciences, Central Laboratory of Yordan Uzunov BG uzunov@ecolab.bas.bgGeneral Ecology (CLGE)Danube Delta National Instit. for Research and Development Eugenia Cioaca RO of Brasilia, Institute of Biological Sciences Paulo Salles BR psalles@unb.brUniversity of Hull, International Fisheries Institute Ian Cowx UK of Natural Resources and Applied Life Sciences, Stefan Schmutz AT of Hydrobiology, Fisheries and Aquaculture 16