Integrating multidimensional geophysical data

Surveys that utilize multiple geophysical methods offer greater insights about the subsurface because each one generally yields different information. Common approaches to integrating or ‘fusing’ multidimensional geophysical data are investigated utilizing computer graphics, geographical information system (GIS), mathematical and statistical solutions. These approaches are synthesized into graphical, discrete and continuous domains. It is shown that graphical approaches allow complex visualizations of the subsurface, but only images are generated and their dimensionality tends to be low. Discrete methods incorporate any number of geophysical dimensions, allow application of powerful Boolean operations, and produce unambiguous maps of anomaly presence or absence, but many of these methods rely on arbitrary thresholds that define only robust anomalies. Continuous data integrations offer capabilities beyond other methods because robust and subtle anomalies are simultaneously expressed, new quantitative information is generated, and interpretive data are derived in the form of regression weights, factor loadings, and the like, that reveal interrelationships and underlying dimensionality. All approaches are applied to a common data set obtained at Army City, Kansas, a World War I era commercial complex that serviced troops in nearby Camp Funston (now Fort Riley). Utilizing data from six geophysical surveys (magnetic gradiometry, electrical resistivity, ground‐penetrating radar, magnetic susceptibility, soil conductivity, aerial thermography), various data integrations reveal the structure of this nearly forgotten town.