Spatial data acquisition and system modeling: notes from the field and the lab.
Name of Lecturer: Philip Graniero Department of Lecture: Earth Sciences Date and
Time of Lecture: January 13th, 2000 at 4:30pm Purpose of Research Project: The
primary purpose of the project is to use model simulations to forecast spatial
patterns among various species in the environment. By comparing current
situations with test results, Graniero hopes to have the ability to predict
spatial patterns for species in the environment. This will give
environmentalists and scientists alike the ability to prevent specie disaster
and to study such areas as future habitat. Description of Research/Technology
used: Graniero’s first step involved measuring the earth’s topography, under
the bedrock of the surface. This experiment took place in Newfoundland, Canada.
To do this he took a random sampling scheme. These schemes were tested at a
density of 40 points per hectare. In order to bring the most precise and
comprehensive data to the table, such technologies as mobile computers and GPS
systems were used. The field in which was being tested proved to be very
difficult to measure due to the changing system and the high demand of physical
resource. His objective still remained the same though, to take this data and
run a model that would enable him forecast spatial data on various species. The
model he used was known as Cellular Automation (CA). The models properties were
as follows: a finite set of discrete states and a state transition rule where
the next state is determined by; current cell state, states of the nearest
neighbours, and the state of other layers. The model worked in specific steps.
First, a spatial structure was built. Second, data was collected from it. Third,
the simulation of different collection agencies were put forth. Fourth, the
model information was compared to the behaviour of actual systems. Fifth, the
model was repeated with random initial conditions. Thousands of trials were done
at this point. This model is often referred to as a “virtual lab”. When the
information was taken at the conclusion of each test, it was sent to processing
units where it was studied in the form of a grid. These grids were then used to
study the spatial patterns of various species. Such future models will be more
complex and more specific, thus showing species habitats and migratory trends.
Adjusting the variables in the model can allow scientists to measure such
activities as the population density of a species. Through the experiment there
were three experiment sets. These included populations, disturbances, and
resource mapping. The resource spatial structure also varied from uniform,
smooth, and “patchy” environments (soil and forest types). Conclusion: This
information is very valuable to environmentalists and society in general due to
the fact that it “looks-out” for species that may be in danger and monitors
the move from one territory to another over a given time frame. Allowing
scientists to predict the habitat and density of species in given areas with
such models keeps humans aware of the impact they may have. This helps protect
the future of species and insures that humans don’t interfere with its habitat
as well. In conclusion, the model is very useful and as it grows and becomes
more sophisticated it should prove to be a valuable resource to environmental
scientists.