Use of knowledge acquisition to build wildfire representation in Geographical Information Systems
Abstract. Wildfire provides a good example to study spatio-temporal representations for GIS applications because of its spatio-temporal variability. One of the salient features ofa GIS is itsbuilt-in capability to calculate and maintain centroid, length, area, and topological relations among spatial objects. Likewise, a competent temporal GIS should be able to support queries about temporal measures, such as rate, frequency, life expectancy, and temporal relations. In doing so, temporal GIS need well-designed representations to structure spatiotemporal constructs and their relations. While temporal information is often incorporated into GIS by time-stamping attributes or layers, this research applies knowledge acquisition techniques of written materials analysis and unstructured interviews with wildfire professionals to elicit conceptual models for wildfire representation. Four elicited conceptual models form a wildfire information cycle which presents a continuum in wildfire conceptualization across different types of wildfire studies. Current GIS layer models can support three of the four conceptual models: locational snapshots (as raster layers), entity snapshots (as features on vector layers), and fire mosaics (as vector layers). The support of layer-based models for representing fire processes is, however, inadequate. With the wildfire information cycle and its four conceptual models, this study shows (1) spatio-temporal information dependency among human's conceptual models, (2) the missing process-oriented data model (fire entities) in GIS, and (3) a three-domain model of semantics, time, and space as a generic GIS representation to support all the four conceptual models. Follow-up studies are undertaken in formulating the three-domain model to implement the wildfire information cycle.
No Reference information available - sign in for access.
No Citation information available - sign in for access.
No Supplementary Data.
No Article Media