Abstract: 3-D Geologic Maps

ROBERTS, M.A., C.S. CHAN, D.G. HOWELL, C.M. WENTWOTRTH, and R. C. JACHENS, U.S. Geological Survey, Menlo Park CA, 94025

Introduction

Traditional geologic maps provide geologists with the means to interpret surface and subsurface information based on a two dimensional format. Whether sketched in the field or printed out from a digital database, the tools for interpreting these maps have changed little over time. Stratigraphic columns and cross sections remain the principal means to display subsurface relations from these maps. Geophysical and well data further help in providing information such as lithology, depth to basement and depth to the crust-mantle boundary. These types of data can be difficult to visualize, especially for novices. The USGS is currently experimenting with a new cartographic product, a three dimensional geologic map. The resulting model is not an artistic rendering in a graphics program, but a portrayal of the geology built by the software program using real data entered by the user. A 3-D map displays features such as sedimentary layers, basement rock relations, the surface of the Moho and the lithosphere-aesthenosphere bounding surface. Digital Elevation Models add topography while a geologic map provides the surface geology.

One of our first experimental models is a transect across northern California, from offshore Monterey Bay to Mono Lake (approximately 350 km x 80 km x 110 km, see figure 1). A small toxic plume, (measured in meters) that leaked into an aquifer in Morgan Hill, California, was added as an interactive part of the transect model (see figure 2). Adding the plume helps to demonstrate the dynamic range inherent with this modeling program.

Software

The program we used for this project, earthVision®, is able to read DEM's, Arc/Info® coverage, and text files of scattered data, as well as a number of other formats. The software earthVision® includes a module called the Geologic Structure Builder which allows the user to input structural relationships in the model. The first step involves building a “fault tree” in which fault blocks are carved by each successive generation of faults. Those closer to the “root” cut those further down in the sequence. Next, the horizons are defined, either by scattered data or a grid. The interaction of these surfaces is defined by one of three geologic rules: depositional, unconformity, or channel erosion. In the 3-D Viewer, the calculated model may be sliced, rotated, magnified, and draped with images, see figure 3. Selected zones and fault blocks may be singled out for display, see figure 4. Additional functions include displaying scattered data as points and calculating volumes. Other modules within earthVision® allow the user to create cross sections, base maps, and contour maps.

Utility of 3-D earthVision® models:

Many, if not most, geologic studies begin with an analysis of data that inherently possess a spatial context. Amassed geologic data and their geometric relationships inevitably convoy some kinematic information. For example, understanding crustal dislocations, movement of salt diapirs, the dissemination of metalliferous compounds, migrations of oil and gas in the deep subsurface or a contaminants in a shallow hydrologic aquifer all depend on 3-D geologic data. The kinematic relations in turn lead to a better understanding of the physical and chemical processes that account for the existing geometries and movement histories and finally to predictive modelling

Spatially compiled three dimensional information for the U.S. Geological Survey will enhance our ability to:

1. Characterize ground shaking resulting from earthquakes, we can do this because seismic energy amplification and focusing corresponds to the distribution and geometries of rock properties, as well as distance from an earthquake hypocenter.

2. Plate tectonic modeling. Because plate processes involve interactions of the lithosphere and aesthenosphere, future models must be at least 100 km deep, as is the accompanying illustration.

3. Crustal growth and associated process modeling. The 3-D earthVision® models make it relatively easy to calculate the volume of materials for any selected body of rock. combining this with age constraints, rates of formation can be calculated.

4. Estimate undiscovered resources

5. Analyze aquifer systems and model ground water movement.

6. Predict the dispersion of contaminants.

Data Entry

3-D crustal constructions in the petroleum industry rely principally on closely space seismic surveys. These data are not available for most areas where the U.S.Gelogical Survey is conducting scientific investigations. We do however. have a vast array of other data sets which lend themselves to three dimensional studies. To test their individual and collective efficacy with the earthVision® package, we incorporated all of the following data types in producing the central California model:

 projections from surface map relations

 published geologic cross setions

 gravity inversions (shallow relations)

 magnetic inversions (intermediate to deep relations)

 public domain well bore data

 seismic refraction information

 wide-angle seismic reflection interpretations

 earthquake tomography

 heat flow data

 digital elevation models

 high-resolution bathymetric data

 earthquake hypocenters

 chemical analyses from water wells.

Problems:

We encountered several problems in the process of building the models. Most of these were due to the lack of sufficient subsurface data in digital form. In many cases, data existed at the surface and in the near-surface, but we needed to extrapolate these relations to great depths. This presented two specific problems. First, because geologic relations are complex, we had a difficult time defining the spatial interactions of boundaries at depths greater than about 10km. Secondly, even if we had some idea of the geometries, creating data that represented these relations accurately was both time-consuming and nearly impossible for more than a simple model. Examples of features where we were not confident enough to charaterize the third dimension included: roof pendants, the bases of plutons, and complex patterns associated with faults such as flower structures.

Other problems involved the coincidence of faults and hypocenters. Major strike-slip faults such as the San Andreas were modelled as vertical surfaces, but we found that many of the hypocenters did not coincide with such a surface at depth. Either the seismic activity did not occur on the faults or our assumption of verticality is in error. If the faults are not vertical, we would have to assume that the hypocenters define the faults in order to model them, but this inference may not be correct. Furthermore, the hypocenters of the past 20 years may not correspond to the major long-term fault boundaries that are essential to our crustal partitioning.

The Future

Three dimensional geologic maps may one day be as widely used as the 2-D maps are now. Their ability to serve as an index to huge range of geoscience data will make them valuable tools for both scientists and the public. For example, the simple click of a button could provide a view of a contamination plume underneath a street map, allowing residents of an area to evaluate the safety of well water; or seismologists may be able to evaluate the potential hazards of new faults, such as blind thrusts, by indexing hypocenters that can provide the date, magnitude and other valuable information of a given earthquake. In order to make this information available to anyone, the data could be posted on the World Wide Web or released as CD-ROMs to provide these same capabilites on personal PC's.

Concluding Remarks

Spatial relationships among geologic features are rigorously displayed in the 3-D Viewer. This enables geoscientists to examine their data and inferences about the subsurface. Experts in fields such as tectonics and ground water are able to use 3-D models to predict future behavior of faults or toxic plumes. Models can also be made to be interactive. A user could find a section of the model that is of interest, click on it and bring up various types of information related to that site such as fossil localities or street maps. Students, and even those without a geology background will also find models such as these more efficient at demonstrating ideas that are traditionally difficult to visualize.

AAPG Search and Discovery Article #90937©1998 AAPG Annual Convention and Exhibition, Salt Lake City, Utah