SUCCESSFUL INTEGRATION OF REMOTE SENSING AND GROUND BASED EXPLORATION TECHNIQUES IN AN ARID ENVIRONMENT

Robert F.E. Jones
Clyde Expro plc
Ledbury, U.K.

Michael Oehlers
Nigel Press & Associates Ltd.
Edenbridge, U.K.

Block 10B Seismic Base Camp, Hadramaut Region, Yemen (182922 bytes)

Exploration programs can benefit in geological understanding, operational efficiency, and safety while reducing costs, by carefully integrating remote sensing into the overall plan. Arid zones, by virtue of their lack of . vegetation, provide the ideal locations to develop these techniques.

Clyde Expro plc is operating in the Hadramaut region of Yemen on behalf of the Block 32 and Block 10b groups (Fig. 1) (52254 bytes). Exploration work has proceeded rapidly in these concessions for the last 2 years. Prior to this group's involvement, even the limits of the Seiyoun basin in this area were uncertain.

The terrain in this area consists of an arid, stepped plateau intricately dissected by precipitous wadis as much as 300 m deep. This topography greatly hinders ground transportation and VHF radio communications; in addition the lack of maps and the similarity of every djebel and wadi increase the risk of getting lost. The climate is similarly inhospitable, with midday temperatures reaching the mid 50 C.

REMOTE SENSED DATA

The remote sensed data found most useful for these projects comprises SPOT Panchromatic, 10 m resolution, "vertical looks" and "stereo pairs" for the whole area of operations. The SPOT vertical data were registered and combined with multi-spectral LandSat Thematic Mapper data of 30 m resolution to create false-color images with the apparent resolution of 10 m (Fig. 2A) (162128 bytes). The LandSat bands and image processing routine were chosen to maximize the discrimination between different rock types. The images were selected to be as recent and dust-haze free as possible; all the required data processing and enhancement of the images for the project were carried out using 486 series PCs. Scales of 1:200,000 and 1:50,000 were used for hard copies.

Ground control points were then established in a few days' field work with a hand-held Global Positioning System (GPS) device that allowed the images to be accurately tied to known points on the ground; thus correcting the images for any uncertainty in the satellite's orbit or direction of view.

INTERPRETATION, GEOLOGICAL APPLICATIONS

The images were systematically interpreted using standard stereoscopic techniques, with constant reference to the vertical look false-color images, to define the outcrops of each stratigraphic unit and visible structural features (Fig. 2B) (162128 bytes). The coordinates of doubtful features were noted and subsequently checked using helicopter reconnaissance.

This approach is a revival, using satellite images, of the aerial photographic technique that was popular for the exploration of remote areas 30-40 years ago.

The three times vertical exaggeration provided by the stereoscope brought out some subtle features not visible on the ground or from the helicopter. A monoclinal flexure was interpreted in Block 1 Ob which could not be confirmed by observation on the ground or from the helicopter. This flexure was finally confirmed to be present, with a dip reversal of only 0.2, when a seismic line was acquired through the feature (Fig. 3) (55676 bytes).

Individual satellite images are a two dimensional view of the earth's surface; however, using the stereo pairs of SPOT images it is possible to obtain, by standard photogrammetric calculations, a ground level elevation value for each 10 m by 10 m pixel common to the stereo pair. This procedure was carried out using a 486 series computer, with minimal manual editing and control. The derived elevation data can then be compiled into a digital elevation model (DEM) of the area. This model consists of a regular grid of x, y, and z coordinates describing the surface topography and can be presented as a contour map or in a color rendition (Fig. 2C) (162128 bytes).

The DEM is immediately useful in itself for preliminary planning of road routes, wellsites, and other civil engineering purposes. However, for the explorer the DEM becomes most useful when combined with other data. The complex pattern of incised wadis meant that the outcrop line of the top of the Umm Er Radhuma was widely distributed over the whole license area (Fig. 2B) (162128 bytes)(Fig. 3) (55676 bytes). Every pixel containing a segment of the outcrop line also has an elevation. Extracting, carefully smoothing, and contouring just these pixels and the surface expressions of the faults created a structural contour map of the top of the Umm Er Radhuma formation (Fig. 2D) (162128 bytes). This map was subsequently verified by three exploration wells drilled on Block 32 where the map was found to have correctly predicted the horizon in each well, within the 15 m accuracy expected.

This map is also useful in depth-converting a suite of seismic two way time maps as it provides a key near-surface geological horizon in depth from which to start stepping into the subsurface. Similarly, in studies of structural evolution, a shallow horizon from which to isopach is critical, but is often difficult to derive from seismic.

OPERATIONAL APPLICATIONS

In all the seismic, geological, and drilling operations the satellite images provided the only reliable maps. Several sets at 1:50,000 scale were laminated with clear plastic and bonded onto heavy linen backing. These images formed easily folded, hard-wearing maps that were simple to manipulate in vehicles and helicopters.

These images were augmented by access notes for the seismic crew, prepared using the stereoscope and SPOT stereo pair. These notes describe the roads and tracks, the best way for the crew to get to each section of the seismic lines, the best places for flycamps, every source or receiver omission, and where every wadi crossing requiring helicopter assistance is located. The use of these images with the hand-held GPS receivers prevented any downtime or emergencies due to crew members getting lost.

Testing early in the seismic program showed that the routing of the seismic lines across the terrain was crucial to obtaining good data. The two most important criteria were firstly, trying to avoid the Umm Er Radhuma outcrop, which suffered from poor geophone coupling; and secondly, minimizing elevation changes, but carrying out any necessary elevation changes in the shortest possible horizontal distance. However, small gaps in source or receiver locations and horizontal line bends did not cause much degradation of the data.

To satisfy these criteria, an idealized program map was drawn, based on the Top Umm Er Radhuma structural contour map, regional trends, and later, on the results of interpreting the earlier lines. This idealized map was then overlain on the right look stereo image. Working with the stereoscope and the stereo pair the lines were re-routed and broken down into straight line segments between "way-points" to satisfy the quality criteria; optimize use of the terrain, and avoid villages and cultivated areas in the wadis. The revised lines were then checked against the geological interpretation to avoid shooting along known faults and to minimize the sections on Umm Er Radhuma outcrop.

The re-worked program map was finally checked against the original exploration objectives and the waypoints digitized. These digitized waypoints were then loaded directly into the "mission plan" on a handheld GPS receiver, which allowed the line layout crew to swiftly and accurately set out the lines and access routes on the ground, marking each waypoint with a flag on a small steel pole.

This technique increased line layout speed from about 4 km/day to over 8 km/day while simultaneously reducing the helicopter hours needed. These improvements in efficiency alone saved more than the cost of obtaining, processing, and printing the stereo SPOT data.

In the Hadramaut, the weathering layer is extraordinarily thick in the djebels as the water table is usually just below the level of the wadi floors; i.e. about 300 m below djebel ground surface. Conventional uphole rigs can drill down to about 90 m (with great difficulty, due to lost circulation), which does not provide sufficient information for calculating the statics corrections. "Big rig" upholes to the base of the weathering cost a prohibitive $100,000/hole. Thus uphole statics surveys are only carried out in the water wells drilled at each exploration well site, where the incremental cost of the survey is very small.

Static corrections determined by refraction statics methods do not seem to work well either, due to the thickness of the section above the water table, the high velocity refractors in the dolomites and limestones and the resulting velocity inversions in the shales.

The most effective static correction method is to model the near-surface, from the Tawila group up to ground level (Fig. 3) (55676 bytes) by constructing a geological cross section in depth and using velocities derived from the water-well checkshot surveys in the layers defined by the model.

The creation of the geological cross-section commences with the elevation profile from the seismic line survey, as this is readily available and is more accurate than deriving one from the DEM. Overlaying the shotpoint basemap on the geological interpretation of the SPOT image allows the outcrops, faults, and flexures to be marked onto the elevation profile and a geological cross-section constructed.

The cross-sections and the velocities from the water wells were then entered into the seismic data processing system to determine the statics corrections for the seismic data. This has led to some significant improvements in seismic data quality (Fig. 4) (110401 bytes) without the trouble and expense of a field geologist having to map the surface outcrop along every seismic line. Moreover, large scale, subtle features which cause long-wavelength statics problems are often not visible from ground level.

These cross sections also proved a useful constraint to the subsequent interpretation of the seismic sections by accurately defining the near-surface structure along each seismic line.

FUTURE POTENTIAL

Much of the interpretive work described above was carried out on hardcopies of the data, and the development of GfS systems will allow similar projects to be handled even more efficiently and accurately in future.

Remote sensed data will be more readily integrated with conventional ground based exploration data and new applications and techniques will appear. For example: For road and pipeline planning, proposed routes could be entered into the GIS and then using the DEM, estimates could be made of gradients, radii of bends, cut and fill requirements, pump station locations, capacity, etc. Future satellites with higher resolution and better oblique view capabilities will improve the accuracy of the DEMs and our ability to interpret geological information from the data. The future availability of radar data will help extend the use of remote sensed data into cloud covered or even vegetated areas.

Looking further ahead, beyond these fairly obvious applications, it may be possible to fine-tune the statics correction velocities along the seismic lines by using the remote sensed data. Tone and reflectivity in the various spectral bands are related to lithology, as is velocity. it may thus be possible to relate statics correction velocities in some of the visible layers to variations in some of spectral bands of the remote sensed data.

CONCLUSIONS

Twenty years ago, remote sensing promised to revolutionize exploration; unfortunately, many of the early promises made were unfulfilled and remote sensing tended to drop out of mainstream exploration. Both these extremes are unrealistic, and these projects undertaken by Clyde in Yemen illustrate some of the ways remote sensing can become a successful and cost-effective part of an exploration program.

• Firstly, the remote sensed data, integrated with a minimum of ground control work, provided maps to use in subsequent fieldwork, a surface geology map, and a digital elevation model with its derived topographic contour maps. Ground based surface mapping of this area would have required man-years of work at great expense.

• Secondly, the remote sensed data enabled us to create a structural contour map of a near surface horizon at a very low cost per square kilometer. This is useful for high grading areas for seismic acquisition, and for depth conversion and structural evolution modeling.

• Thirdly, the remote sensed data became a crucial planning tool for seismic operations to optimize data quality and minimize acquisition cost without having to resort to costly and time-consuming swath shooting or similar high-effort techniques. Preliminary cost estimates could be prepared for the construction of alternative drilling sites and their access routes without spending time and money sending surveyors out to the field.

• Finally, the surface geological map derived from the image interpretation enabled us to create geological cross sections along the shot seismic lines in a matter of hours without having a field geologist mapping along the lines. This saved several weeks' work and tens of thousands of dollars and resulted in significant improvements in the quality of the seismic data and in the interpreted subsurface mapping.

Remote sensing can provide highly cost-effective benefits to an exploration program in an arid region, and many of the applications can also be developed for use in areas with vegetation cover.

To achieve satisfactory results from the integration of remote sensed data requires some imagination and a determination not to compartmentalize the exploration program; it means starting with a clear view of the range of final objectives and the willingness to try combining and using data in different and unusual ways.

ACKNOWLEDGMENTS

We would like to acknowledge the permission of Norsk Hydro, Oranje-Nassau, OMV, and Ansan Wikfs, Clyde's co-venturers in Yemen, to publish this article. We would also like to thank Excel Geophysical Services, Solid State Internacional Ingeniera, R. Chiles and S. Russell at NPA, Clyde staff, and all the various subcontractors for their outstanding contributions to the success of our projects in Yemen.

THE AUTHORS