Dolerite sills of the Ferrar magmatic system in the Dry Valleys, Antarctica, are very well exposed on the slopes of deep valleys, providing a large amount of data for detailed geological mapping in accessible areas as well as remote sensing observations in inaccessible places. Collection of those data obtained from archival geological maps and digital sources were used for the construction of a three-dimensional geologic model of the Basement, Peneplain, Asgard, and Mt. Fleming sills. This model constitutes a spatial framework for the continued detailed structural, petrological, mineralogical and geochemical studies of perhaps one of the best-exposed magmatic systems on Earth. For construction of the model the technique of building contiguous rock volumes was used. In the case of horizontal sills the most convenient method of creating volumes is the building top and bottom surfaces using thickness constraints. The resulting surfaces were later merged into the volumes. To construct component surfaces large amounts of data were obtained from multiple sources including maps from the New Zealand Geological Survey and the USGS Antarctica program. In the first stage, the lowermost sill, the Basement Sill, was modeled in its original sheet-like form. In the next stage its eroded parts were removed from the model by tracing the intersection of the original sill form with the present terrain surface without unconsolidated sediment cover. A digital elevation model (DEM) was generated from digitized contours at 50 m intervals using the USGS topographic maps (1970 survey) and from SRTM and Lidar data. Next, maps and cross-sections were scanned, geo-referenced using XYZ coordinates system, and registered in 3-D space. Scanned maps were then draped over the DEM. All of the above operations were done for interactive and intuitive digitizing of points in 3-D space. This process has allowed intricate field relations of single sills and between touching sills to be understood in intimate detail. Visual depiction of whole rock chemical compositional variations (e.g., MgO and CaO) within the Basement Sill has also been initiated, which has allowed the 3-D structure of the pervasive ultramafic tongue of orthopyroxene phenocrysts to be traced across the region. This was done using nine detailed sampling profiles in the Bull Pass and Wright Valley region. The spatial mapping of this unusual diagnostic feature allows the 3-D pattern of magmatic flow during sill emplacement to be determined in some detail, which is central to understanding the dynamic establishment and workings of this magmatic system. Construction of the overall model has also required the extensive reviewing and refinement of the subsurface interpretations presented on vertical cross-sections of the existing geologic maps especially in places where limited data is available. This general working digital model may also be updated and continuously refined as new field observations become available. And, more important, it is a valuable resource in the design and implementation of future fieldwork.
