Introduction
Recent oil and gas discoveries in western Kentucky prompted a review of the oil & gas potential of the east end of the Rough Creek Graben (Figure 1). The area was the focus for a project that combined digital remote sensing data with GIS technology to explore for oil and gas. This project was completed on behalf of an American client by Patrick Lengyel, an exploration and mining consultant based in Winnipeg, Manitoba, and Geomatics International Inc. of Burlington, Ontario, Canada.
      Recent advances in GIS software have made it possible to integrate the available satellite and aircraft-acquired data with geological, geophysical and topographic digital data sets, to better define and correlate geological structural features at surface and at depth. This progressive approach, where remotely sensed data has been combined with a variety of other data sets within a GIS environment to correlate surface expression with buried geological structural features, has proven to be a superior oil and gas exploration tool.
      Existing geological data in the project area contained gaps in terms of bedrock structural features due both to a 'masking' effect imposed by the overlying Pennsylvanian-aged sandstone and a lack of subsurface exploration data. Previous remote sensing work has shown that hydrocarbon-related features at depth, which are often related to 'basement' features beneath producing sedimentary basins, can be recognized at the surface. GIS technology, which allows the user to combine surface data with subsurface data, was used to prioritize the surface features.
      A rapid, low-cost evaluation of the 2000 km2 area was possible by using digital geophysical data provided by the National Oceanic and Atmospheric Administration (NOAA), using peak-season archived Landsat TM data and the strong support provided by the United States Geological Survey, the Geological Survey of Canada and the Kentucky Geological Survey.

Geological Background
The Illinois Basin is one of many sedimentary basins found throughout the North American continent. The base of the Illinois basin is filled with rift sediments that infilled the Reelfoot Rift Complex, a late Proterozoic/early Cambrian continental rift valley system. One arm of the Reelfoot Rift Complex, essentially a series of valleys not unlike the East African Rift Complex, extended off the main north-south valley into present-day west Kentucky. This arm, the Rough Creek Graben, was filled initially with similar marine sediments. Sizeable gas deposits have been discovered in rift-related sediments of the Michigan Basin to the north, and in the Rome Trough to the east.
      Tectonic deformation intensified inherent weaknesses in the "basement" rocks underlying the basins. Block faults developed in regular orthogonal sets throughout the relatively homogenous and brittle basement. Major lithotectonic boundaries also developed into, or were reactivated as, faults. Movement along these basement faults affected the deposition of the sediments in the overlying basin and also affected the formation, migration and deposition of hydrocarbon deposits.
      Significant oil and gas accumulations have been discovered and drilled in the uppermost units of the basin, specifically in rocks of Mississippian and Pennsylvanian age. The best producing areas are found along drape folds which occur in the sediments overlying the margins of tilted basement blocks. Significant gas fields also occur within fault zones in Devonian-aged shales.
      Continental-scale faults, some due to reactivated rift faults, and others apparently resulting from even larger crustal scale weaknesses, also resulted in deformation of the overlying basin sediments. Less voluminous but still significant hydrocarbon accumulations occur throughout the basin along these fault zones.

Exploration Activity
Exploration and recovery in the Illinois Basin peaked in the 1940s, resurging in the 1960s with the advent of secondary and tertiary recovery systems; however, annual production has declined since that time. Exploration in the Rough Creek Graben, located in the southeast end of the Illinois Basin, followed a similar history with a slight modification. The folded sequence above the Rough Creek Graben, known as the Moorman Syncline, contains younger Pennsylvanian rocks in the core and progressively older Mississippian and Devonian rocks towards the north and south. The Pennsylvanian rocks, mainly marine sandstones and siltstones, contain considerable coal deposits. Coal mining companies, controlling the mineral rights beneath the deposits, prevented oil and gas exploration in the rocks below the Moorman Syncline.
      Furthermore, during the period of peak exploration activity, the exploration costs associated with testing the deeper parts of the Rough Creek Graben were prohibitive. The development of better equipment and techniques and improvements in the understanding of hydrocarbon deposits has resulted in the recent resurgence of activity along the margins (Rhoda Pool, Edmonson County) and within the deeper areas (Conoco No. 1 Mark Turner Well, McLean County, and No. 1 Isaac Shain Well, Grayson County) of the Rough Creek Graben.

Existing Database Problems
High quality, 7.5 minute quadrangle surface geology maps exist for the entire area. However, detailed subsurface geological information was lacking in the project area. The project area, centered roughly over the east end of the Rough Creek Graben/Moorman Syncline, is underlain by Pennsylvanian sandstone. This Pennsylvanian sandstone unit, known locally as the Caseyville Formation, is a monotonous sequence of fine-grained to pebbly sandstone with few of the 'marker' horizons that allow geologists to determine structural displacement. Consequently, the thorough structural mapping data that exists along the edge and outside of the Rough Creek Graben could not be extended into the center. The lack of basement structural information has also hindered the exploration for gas in pre-Knox Group sediments.
      Recent studies, using remote sensing and hydrocarbon production data, have shown that there is a distinct relationship between surface lineaments (fractures, faults and topographic alignments), subsurface hydrocarbon traps and basement geological features (Figures 2 and 3). Consequently, any new exploration program had to upgrade the structural database through surface lineament analysis.

Database Components Satellite Imagery
Remote sensing imagery, specifically Landsat TM and JERS-1 radar data, was chosen as the most cost-effective tool to determine the location of surface lineaments. An archived, cloud-free Landsat TM image obtained during early spring provided information on vegetation and groundwater-related anomalies. Several Landsat TM band plots were made, including 7-4-1, 5-4-2, 3/1, 5/4, and 7. The 7-4-1 plot proved the most useful to enhance vegetation-related lineaments and the 5-4-2 plot proved useful to distinguish vegetation versus non-vegetation features and soil anomalies.
      Grey-scale plots of the JERS-1 imagery were used alone, with edge detection filters, and in IHS transform plots with color contour aeromagnetic data.

Topographic Imagery
Digital topographic data sets, available through the USGS, were plotted as grey tone shadow plots to compare with JERS-1 data. Included with this digital data set were highway, city and drainage data sets. Drainage data sets were used alone for drainage pattern analysis.

Geophysical Imagery
Digital geophysical data sets were acquired from NOAA, the National Geophysical Data Center and the USGS. The East-central U.S. Grid, distributed by the NOAA, contained regional digital aeromagnetic data for the east-central portion of the continental U.S. The data was derived from flight lines spaced 1 mile apart (Archive File No. NOAA/NGDC). The world digital gravity database, compiled by the USGS and distributed through NOAA, was also used (1994 version).
      Color contour plots, shadow plots and IHS transform plots (shadow aeromagnetic data over color contour gravity data) of both data sets were used to interpret regional basement structures.

Geological and Production Data
The surface geology maps of the area were manually digitized using ARC/INFO. The units and features of interest to this project were isolated and added to the data base.
      Although somewhat dated, oil and gas compilation maps from the 1970s were manually digitized in SPANS. SPANS was used to compensate for poor projection information on the original maps. The resulting files were converted to VEC/PLAY, then to ARC/INFO. Up-to-date detailed production maps, available as 7.5 minute quadrangle overlays (and scheduled to be released in blocks) were used to add those new discoveries made since the release of the larger scale oil and gas compilation maps.
      Mapped fault zones that were found to be coincident with those on the 7.5 minute quadrangle geology map fault zones, were identified and plotted on acetate overlays. This information was scanned on a Tangent drum scanner then vectorized and georeferenced in ARC/INFO.
      Once in digital form, final plots were made in ARC/INFO software. Corrected data sets were then imported into EASI/PACE (PCI) software for combination viewing and final plotting.

Lineament Interpretation
To place the project area into perspective and to insure that all possible target types had been considered, a geological compilation was made of the Illinois Basin. The compilation focused on the evolution of the basin. Factors that affected sedimentation, particularly those related to hydrocarbon generation and emplacement, were identified and compared to features in the project area. Information from recent deep drilling in the Rough Creek Graben and new structural interpretations of recent seismic data were also assimilated.
      Regional structural interpretations were completed using 1:2 000 000 scale plots of color contour and shadow plot aeromagnetic and gravity data. The majority of the rocks found above the basement are sedimentary, homogenous, and typically non-magnetic. Consequently, the geophysical response observed in the results of regional aeromagnetic and gravity surveys was interpreted to reflect features in the underlying basement rocks.
      Major crustal structural features were recognized by applying the "illumination" option on the software used to produce the aeromagnetic shadow plots. Specific illumination orientations, which best highlighted variations in aeromagnetic intensity, were then selected and plotted. IHS transform plots of the shadowed aeromagnetic data and color contour gravity data were made for use in distinguishing specific lithotectonic units in the basement.
      New lineaments and fault-extension lineaments were interpreted using Landsat TM, JERS-1, digital terrain, drainage and the surface and subsurface data available from the 7.5 minute quadrangle sheets and other geological reports. The lineaments were combined with the digitized fault zones to make a single, comprehensive surface lineament data set.
      Recent seismic work has shown that distinct, simple faults at depth can become more complex towards the surface. Several distinct basement features in the project area (typically a truncation in aeromagnetic data) coincided with surface lineament sets. IHS transform plots of the color contour aeromagnetic data combined with grey-scale JERS-1 radar data and the surface lineament data set proved to be the most useful for associating surface lineaments with basement features (Figure 4).
      Using the Illinois Basin compilation data, derived during this project, the various lineaments were grouped into distinct fault zones and assembled on a final interpretation map. Potential plays, structural environments capable of being oil and gas reservoirs, were discussed and prioritized as exploration targets based on a variety of factors including their similarities with known producing plays and their depth.
      A detailed structural interpretation of the area was produced, which was used in determining potential targets for further oil and gas exploration. For the deep oil and gas plays, specifically, large gas deposits within rift sediments at the base of the Rough Creek Graben, this structural information identified targets for seismic exploration. The structural interpretation was also of great value for plays of an intermediate depth, providing structural information to guide wildcat drilling. Potential wrench fault-related, high priority structures were delineated along the north edge of the Rough Creek Graben. Potential productive settings were identified along northeast-trending faults, in settings similar to those of known productive fields in the west end of the Rough Creek Graben. The identification of shallow exploration plays was aided considerably by the extension of known surface faults, and the identification of similar parallel features.
      Structural interpretations also provided new data, related to the formation and control of gas deposits in the adjacent Schrewsbury Gas Field. This new data will be useful in development drilling and locating new fields.

Conclusion
The analysis of the Landsat TM, JERS-1 and digital terrain data allowed for a thorough surface lineament interpretation. Surface lineament information combined with high resolution digital data proved to be the most cost-effective method for evaluating the entire 2000 km2 area of interest.
      The ability of the GIS software to import these data sets and combine them with digitized surface, subsurface and basement geological features allowed for an interpretation that effectively related basement features at depth to features observed at various levels within the basin and at surface. The ease of digitizing and of importing additional data, such as production information and structural features, made it possible to correlate specific structural settings with specific hydrocarbon deposit types.
      This combination of remote sensing and GIS technologies has resulted in: 1) a better definition of basement-related features; 2) an extrapolation of existing surface structural features to depth, which will better define existing oil and gas plays; 3) the identification of potential locations of rift-related delta fan complexes relevant to pre-Knox gas exploration; and 4) the delineation of previously undetected, potential hydrocarbon-bearing structures throughout this part of the Illinois Basin.
      Future GIS work includes the identification and use of specific spectral combinations to detect surface tar sand occurrences.

About the Authors:
Patrick Lengyel is an exploration and mining consultant based in Canada, currently providing services to several international clients in Africa, North America, South America and the CIS. He may be reached at (204) 255-4037. John Fairs is a remote sensing specialist at Geomatics International Inc., Burlington, Ontario, Canada. Geomatics International specializes in the application and development of geographic information systems and remote sensing technologies. He may be reached at (905) 632-4259.

BACK