Representing Reality:

Imagery in the Cognitive, Social and Natural Sciences

 

A Conference presented by the University at Buffalo IGERT in GIScience

We are pleased to announce the keynote speaker will be Dr. Mark Fonstad (Department of Geography, Texas State University). Dr. Fonstad’s most recent work has focused on the high resolution remote sensing of river habitats and hydrological work on the upcoming NASA SWOT satellite.  Among his many accolades, Dr. Fonstad was recently appointed Environmental Science associate editor o fthe Annals of the Asssociation of American Geographers. The abstract for his talk is posted below.

The riverscape is a microcosm of more general issues in GIScience, including representational, modeling, and measurement debates. The past ten years have seen a complete revolution in spatial river science, specifically in remote sensing and spatially distributed river models and simulations. Nevertheless, many of our standard theoretical models of river behavior remain centered on spatially scattered point and section measurements that may miss entire levels of observation. As one example of this growing revolution, this presentation will show how remotely sensed data can be used to extract three dimensional below-water environmental data for large area riverscapes at scales from a few centimeters to that of the entire basin. Examples of this “hyperscale” utility, based on calibrated high-resolution imagery show systematic deviations from the classic theory of river organization as well as an unprecedented level of unexplained randomness at most scales. Further theoretical advances need to be rooted in this type of spatially distributed process study. Spatially-distributed process models, such as those based on cellular automata, require such hyperscale representations if they are to be both spatially accurate and precise. Debates about the need for more realistic vs. more spatially useful models in river science are very much evident in this age of multiple modes of representation. Clearly riverscape studies are in a mode where revolutions in observational method translate to theoretical advance.

The 2010 Representing Reality Conference will be held

May 12-15, 2010 in Buffalo, NY

To register now for the Representing Reality 2010 Conference in Buffalo, NY, please CLICK HERE

The Light Fantastic: GIScience and the Representation of Riverscapes

Finally, in extending river studies from hydrology and geomorphology to ecology and management, several conceptual issues remain. The common assumption that areas of high habitat quality will have high animal densities is overly-simplistic. This conceptual issue is a by-product of the predominantly Eulerian frame of reference used in river science. By invoking an object-oriented, Lagrangian frame of reference, it may be possible to combine remotely-sensed habitat suitability maps with agent-based models of organism population dynamics. In this view, modeled agents (such as individual fish) interact to produce spatial and temporal patterns such as schooling and clustering, while the per-pixel habitat maps supply probability and energy surfaces upon which the agents can act. Other riverscape dynamics, such as woody debris transport and the interactions of rivers with humanity, might also benefit from this alternate frame of reference. Our major difficulty is in the observation and measurement of mobile individuals in a Lagrangian frame of reference; the few such methods that exist (such as RFID’s) are incredibly time-consuming, invasive, and treat only a small sample of the riverscape. We have, as yet, no general principles about when or where our representations should be object-based (Lagrangian) or field-based (Eulerian). This issue is very old, being well-defined even in the ancient differences between Heraclitus and Plato, and is extremely relevant to GIScientists today.

Keynote Speaker: Dr. Mark Fonstad