Summary
Report
See the full content of this document
Extract
Geocollaborative soil boundary mapping in an experiential GIS environment.
Introduction
Soil maps are an integral part of contemporary digital information infrastructures. The use and application of soil maps has expanded well beyond their traditional role in agriculture and now underpin many domains in engineering, environmental resource management, urban planning, hazard mitigation and others. Soil map production, however, remains a time consuming process and is a challenge to providing timely, up-to-date, and reliable soil information. This paper proposes the use of an innovative Experiential GIS (EGIS) environment that draws on immersive three dimensional (3D) graphical displays coupled to a GIS to allow soil scientists to develop critical soil-landscape models and delineate soil boundary maps in a more efficient, time-saving, and collaborative mapping environment. Currently, a number of methods, ranging from traditional soil survey to digital soil mapping, are used to identify, delineate and describe soils. Regardless of the chosen method, accurate soil mapping requires field observation, laboratory measurement and the all-important soil scientist's knowledge. Traditional digital soil mapping focuses on creating reliable and replicable soil maps using numerical models to infer the spatio-temporal distribution of soil classes and soil properties (Hempel et al. 2008; Weber et al. 2008). These predictive models are often based on conventional statistics such as regression (Gessler et al. 1995; McKenzie and Ryan 1999; Bell et al. 2000), geostatistics including kriging (Odeh et al. 1994; 1995) and artificial intelligence inference engines (Skidmore et al. 1991; 1996; Zhu et al. 1997; 2001). Although predictive soil modeling can enhance the speed of soil mapping, traditional soil mapping based on the stereoscopic interpretation of aerial photographic pairs remains the predominant approach to soil mapping (Scull et al. 2003). This approach is primarily a geovisualization exercise in which soil scientists use maps, imagery, and spatial data to develop cognitive soil-landscape models. The most common cognitive modeling process involves the stereoscopic interpretation of two-dimensional (2D) analog stereo photos augmented by field investigation. This process is labor intensive, time consuming, and largely noncollaborative. Only one soil scientist at a time can effectively use a stereoscope to conceptualize the soil-landscape model. Although the soil scientist perceives a 3D terrain model, the model has to be manipulated in the mind inducing a significant cognitive load. Soil mapping in this mode is thus highly individualistic and relies heavily on the cognate and interpretive skills of the soil scientist. Surprisingly, only one soil map at any given scale is given to the user community despite several possible soil perspectives for an area. Generating multiple soil maps that reflect in...See the full content of this document
Sponsored links
