Allgöwer Britta

Knowledge Based Dynamic Landscape Analysis and Simulation for Alpine Environments

Project Number: 4048-064432
Project Type: Research_Project
Project Duration: 02/01/2002 - 05/31/2006 project completed
Funding Source: SNSF ,
Project Leader: Dr. Britta Allgöwer


e-Mail: britta.allgoewer(at)academiaraetica.ch

related to this project.
for which the project has a relevance.


Research Areas:
Landscape

Disciplines:
environmental sciences
other areas of environmental sciences


Abstract:
Research Question and Concept Modeling landscape evolution is like a journey through time and space, involving different speeds and scales as well as varying thematic depth. When dealing with complex topography and (natural and man-made) processes we should be able to take a ‘bird's eye and worm's-eye view’ into perspective at the same time, to truly understand the processes we investigate. We lack, however, the methods and instruments to do this by a cross-scale and hence ‘cross-model’ approach in real time and to the benefit of the processes we want to understand. Based on the disciplines of landscape visualization/virtual reality, modeling theory/theoretical systems ecology, and geographic information science, we design and attempt a prototype system that allows the joint modeling, analysis and visualization of landscape dynamics at various scales together with the processes triggering or forcing the changes. The system to be developed has to meet the needs of researchers, decision-makers and laymen alike. In modification to the original concept we learned to give even more emphasis to the performance and analysis of real case studies. Case studies allow the derivation of so-called use cases, that ‘brake down’ complex processes into sub-processes, which are easier to understand, to formalize and finally to translate into software architecture and computer code. Hence, the more realistic our case studies are, the more we can provide the ‘hard’ proof of concept while developing IPODLAS (Interactive Process Oriented Dynamic Landscape Analysis System). Only then the necessary theoretical background can fully be approached and can a generic computer based framework be build that integrates different scales for space, time and models. Procedure We address three case studies at three scale levels. The case studies comprise larch bud moth, wildland fire and human infrastructure modeling. The scales are: 0 = ‘landscape’ level (e.g. valley level), -(-2)/-1 = one ‘level bellow’ (e.g. forest stand level), +1/(+2) one ‘level above’ (e.g. Alpine Arc). One of the major difficulties is to define the validity range of the specific scales. ‘Boundaries’ between scales are never strict but fluent, especially for natural processes. In particular, we do not feel comfortable with the often-used ‘spatial unit’ landscape. It allows a wide range of interpretations, and definitions tend to be fuzzy and highly subjective. On the other hand, it would not make sense to define the different scales by explicitly reduce them to scale indications as e.g. 1:25’000. What may seem appropriate and represent e.g. scale -1 for the larch bud moth case study, might not fit for wildland fire modeling at all. We therefore add the term level of detail or granularity and integrate this into our implementation concept. However, since we aim at developing a generic system we try to formalize interface procedures for data and model exchange within and across scales as far and as general as possible in order to finally allow the hooking up of any models and data of interest. Based on our experiences of the case studies and use cases respectively we develop the specific IPODLAS data model and system architecture. Beta-testing will be performed with selected end users during 2004. First Results Besides elaborating the necessary theoretical background our three PhD candidates performed experimental test implementations, in particular: - ‘Cross implementation’ of temporal and spatial features in order to reveal the drawbacks when aiming at their joint modeling. First, a wildland fire spread model, which had previously been implemented in a GIS environment was implemented into a Systems Modeling environment designed for modeling phenomena through time. Second, a larch bud moth dynamics model, which had previously been implemented in a Systems Modeling environment was implemented into a GIS system. For the visualization part one major result was the successful implementation of dynamic textures that are controlled by a simple fire model. Computation of a grid cell’s status and display of the corresponding texture are performed in real time on a standard PC, while modeling and visualization are tightly coupled. - ‘Intelligent tree’ implementation. It is a well-known fact that visualization ‘forgets’ about attribute data and the semantic context of an object or item to display. Once an object is ‘handed over’ to visualization it becomes ‘unintelligent’ and usually does not remember its lineage, no connections between its original spatial and/or temporal environment(s) are stored. For this we performed the seemingly trivial looking task of implementing one tree that must ‘travel’ through all systems gathering information influencing its life and appearance, i.e. local (and global) spatial conditions and growth conditions. This example demonstrates the relative ease of the one-way communication channel ‘GIS-Systems Modeling-Visualization’ but proves to be difficult when wanting to reverse it and change e.g. tree growth simulation parameters. For the time being it is only apt for ‘circular analysis processes’ where the whole chain GIS-SysMod-VIS is performed each time a change should be initiated. Expected results The expected final outcome is a prototype 4D landscape analysis and simulation system, that integrates the so far separately kept functionality of landscape visualization, geographic information science, and modeling theory. By systematically performing and analyzing real-world case studies we hope to contribute scientific innovations and backgrounds that allow the true integration of space, time and models in and over various scales. Networking The project and the three granted PhD studies is well anchored in a wide spread user community. The Swiss National Park and its affiliated associations (Scientific Council of the Swiss National Park FoK-SNP, Federal Commission of the Swiss National Park ENPK) will be a main end user, together with the Regional Planning Organization Pro Engiadina Bassa (PEB) and the Office of Economics and Tourism of the canton of Grison (Amt für Wirtschaft und Tourismus, Kanton Graubüden). (see also abstract version 1.0) On-site Research Places The main study areas for micro and meso-scale research case studies are the Upper and lower Engadine and the MŸnster Valley with the Swiss National Park in the center. To address the problems of major cross-scale and cross-model jumps, macro-scale case studies addressing the whole Alpine arc are performed as well. The research case studies are taken from the background of the main applicants and address the impact and eventual interactions of the larch bud moth on it’s host the larch, wildland fire occurrence and human infrastructures.


http://www.nfp48.ch/projekte/projectdocs/32/allgoewera.pdf


Publications:
Poster: Knowledge based dynamic landscape analysis and simulation for Alpine environments [Britta Allgöwer, (German)] Poster (pdf))

Isenegger D. (2006) IPODLAS – A Framework for Coupling Temporal Simulation Systems, Virtual Reality, and Geographic Information Sys- tems, Universität Zürich, Zürich.
Isenegger D., Price B., Wu Y., et al. (2005) IPODLAS – A software architecture for coupling temporal simulation systems, VR, and GIS, in: Isprs Journal of Photo- grammetry and Remote Sensing, 60, 34–47.

Price B. (2005) Spatio-temporal modelling and analysis of larch bud moth population dynamics in the European Alps, Sierke Verlag, Göttingen.
Price B., Allgöwer B., Fischlin A. (2006) Synchrony and travelling waves of larch bud moth? Time series analysis with changing scale, in: Ecological Model- ling, 199, 433–441.
Wu Y. (2007) Knowledge-based 4D Visualization of Amorphous Phenomena in Complex Terrain, Dissertation, Universität Zürich, Zürich.
Wu Y., Price B., Isenegger D., et al. (2006)
Real-time 4D visualization of migratory insect dynamics within an integrated spatiotemporal system, in: Ecological Informatics, 1, 179–187.

Source of Information: NF Import 2002


Last update: 12/23/16
Source of data: ProClim- Research InfoSystem (1993-2020)
Update the data of project: CH-64432

Go Back