Remote sensing methods in offshore exploration

Remote sensing methods in offshore exploration

413 Remote sensing methods in offshore exploration Bj0rn M. Saether, Hakon G. Rueslatten, Egil Rundhovde, Christine Fichler, Tormod H. Henningsen and...

3MB Sizes 11 Downloads 110 Views


Remote sensing methods in offshore exploration Bj0rn M. Saether, Hakon G. Rueslatten, Egil Rundhovde, Christine Fichler, Tormod H. Henningsen and Stale Johansen

The aim of modern remote sensing methods is to enhance variations in digital data sets by applying various numerical filtering techniques and statistical calculations in order to enhance features of interest in the available data. However, a necessary condition for this approach is the availability of an advanced and user-friendly remote sensing system with the possibility to perform interactive processing and "on-the-screen" interpretation. In the present work, a remote sensing system from International Imaging Systems (I2S) was used to interpret geophysical data sets from the Barents Sea region (both offshore and onshore). Features interpreted from the data sets are integrated with other data sets in order to obtain information about the co-variation of features having geological significance. These gravimetric data were used to identify Palaeozoic structural elements in the Barents Sea. Gravity data based on satellite altimetry is also presented and compared with conventional gravity data. The present results indicate that main structural elements can be defined more precisely from the processed data than from ordinary gravity contour maps. An additional benefit of such processing is the detection of subtle trends in the gravimetric field.


Remote sensing may be defined as the computerassisted processing, integration, classification and interpretation of remotely sensed data in a digital format. Remotely sensed data is defined as the acquisition of data on any phenomena using sensors that are not in direct contact with them, e.g. radar data, multispectral data, potential field data. Modern remote sensing methods are applicable to all types of raster-formatted digital data sets. Data sets such as airborne magnetometry and marine gravimetry are now available in large offshore areas, and the integration and extraction of information from such diverse data sets will eventually become an important activity for the explorationist. The main benefit of using remote sensing methods on these data in exploration is the potential of screening vast areas cheaply and rapidly in search for petroleum, and also to integrate potential field data with interpreted seismic data for a comparison and final interpretation. Traditionally, remote sensing is used in the search for petroleum "onshore", and the most frequently used data are satelhte imageries. The processing of these data has been a task for the computer specialists, and the geological interpretation was performed on transparent overlays on hard copies. Modern remote sensing technology allows the in-

terpreter (a) to participate in the processing and integration procedures, (b) to perform interactive interpretation directly "on-the-screen", and (c) to carry out automatic classification of phenomena. This is made possible thanks to user-friendly menu-driven computer systems. Case study area and geological setting

To illustrate the use of modern remote sensing methods an area in the southwestern Barents Sea and onshore northern Norway was chosen as a case study area. The study area covers the Caledonian Orogen in Troms and Finnmark as well as the continental margin offshore these areas and the southern Barents Sea. For location of the study area see Fig. 1. The dominating orientations of the structural elements in the studied area of the Barents Sea are a NE-SW trend and a less pronounced NW-SE trend (Gabrielsen et al., 1990). In this region there exists a good correlation between the Palaeozoic structures and the gravimetric field (Gudlaugsson et al., 1990). These structural trends coincide, respectively, with the main fold axis of the Caledonides of West Finmark (Roberts, 1985), and the Trollfjord-Komagelv (T-K) FZ trend in East Finmark (Siedlecka and Siedlecki, 1967). Tectonic activity during the Devonian-Permian period, characterized by extensional movements

Petroleum Exploration and Exploitation in Norway edited by S. Hanslien NPF Special Publication 4, pp. 413-420, Elsevier, Amsterdam. © Norwegian Petroleum Society (NPF), 1995.


B.M. Scether, KG. Ruesldtten, E. Rundhovde, C. Fichler, T.H. Henningsen and S. Johansen




15° 20°



(^ ''V'^''"'*-%/

Sweden \




V ^

Bnland v^

Fig. 1. Location map; the study area is hatched.

(Roufosse, 1987), has caused a reactivation of these structural trends in the offshore areas which influence the configurations of the basins (Johansen et al., 1993). In the same period as the basins developed a transgression took place in the area with a subsequent narrowing/closing and evaporation period, leading to large deposits of evaporites (Jensen and S0rensen, 1992). The Palaeozoic reUef controlled the deposition of evaporites, and it is assumed that these low-density deposits increase the contrasts observed in the gravimetric field data sets in this part of the Barents Sea.


The study area onshore is covered by a gravimetric (Bouguer anomahes) data set provided by the Norwegian Geological Survey (Olesen et al., 1990), and these data are merged with a marine gravimetric data set, processed by Amarok Inc. The resolution is 1x1 km, and the values are ranging from —88 to 99 mGal, scaled to pixel values 0-255 and geometrically corrected to UTM. These "raw" data set is presented in Fig. 2A. A minor part of the study area in the Southwest-


Remote sensing methods in offshore exploration


500 km

Fig. 2. (A) "Raw" data; coast contour is indicated by black. (B) Result of the "Wallis filter"; large window size. (C) Result of the "Wallis filter"; small window size. (D) Principal Component-1 of the processed gravity data set, based on gradient filtering from NW-N-NE-E (4 files).

ern Barents Sea is selected for testing the use of GEOSAT satellite altimeter data. These data of the GEOSAT exact repeat mission No. 11 (4/27-5/13/87) have been converted to free-air gravity anomalies. The processing was based on programs by Johansen et al. (1991). The altimeter height has been corrected for electro-magnetic (e-m) bias, sohd and ocean tides, the influence of water vapour on the speed of e-m waves, the weight of air masses on sealevel, the varying content of free electrons in the ionosphere and for barometric pressure. The long

periodic orbit errors have been removed by crossadjusting the tracks. Gridding has been performed by means of biquadratic functions and a gridsize of 5x5 km has been chosen. The geoid has been converted to free-air gravity anomahes by a fast Fourier transform method (Schwartz et al., 1990; Farrelly, 1990). Comparison of these two types of gravity derived data is discussed below. In addition to the gravimetric data we also had access to seismic data offshore and Landsat-TM images and geological maps onshore.


B.M. Scether, KG. Ruesldtten, E. Rundhovde, C. Fichler, T.H. Henningsen and S. Johansen

Methods An optimal processing of gravimetric data sets based on remote sensing methods enhances largescale structural trends and local variations, which is important for the understanding of the regional geological development. The processing techniques used in this work are described below, and the effects of the various processing methods are illustrated in Fig. 2. Adaptive filtering In order to enhance contrast locally in a data set an adaptive filtering technique is proven to be useful. The apphed "Wallis-filter" (International Imaging Systems, 1990) is based on a "moving window" processing: within each window the pixel values are adjusted to a specified mean and standard deviation. The result of this processing is dependent on the specified mean value and variance as well as the window size and the step length. Figures 2B and 2C show the effect of a large and a small window size, respectively; the smallest window size gives the best "local resolution".

Fig. 3. Flow-diagram for the creation of FCC.

Directional filtering Interactive Interpretation Directional filtering enhances hnear trends and reveals subtle anomalies in the data sets. Such trends may have structural geological significance and the shape and orientation of these features may give valuable information regarding the processes which were responsible for the formation of the geological structures (Moore, 1983; Sabins, 1987). There are numerous approaches to this kind of processing. In this study the following procedure was selected: directional filtering (3x3) from North West, North, North East and East was preformed, and the information in the four files were extracted by running a Principal Component (PC) analysis, and the PC's with most structural information was selected (see Fig. 2D which displays PC-1). These processing techniques are also effective for identifying artifacts in the data sets (Saether et al., 1991). Integration of processed data A Red-Green-Blue (RGB) False Colour Composite (FCC) image was produced based on (R) the "raw" data, (G) Principal Component 1, and (B) Principal Component 2 (see Figs. 3 and 4a). This FCC-imagery was used as the basis for further interpretations as described below.

A novel program-module, unique to Statoil, allows the interpretation of linear structural features to be performed directly on images displayed on the IVAS/I2S monitor. The most important advantages can be summarised as follows. (1) The IVAS system allows the interpreter to enhance interesting features in the image under investigation during the interpretation phase by using IVAS commands such as "colour stretching", "zoom up", "pan", "pseudocolour", "pipe-select" and "level slice". These enhancements are performed without changing the interpreted graphic elements which are displayed on the screen, and without significantly interrupting the interpretation. (2) Various types of features can be colour-coded. (3) The interpreted lineaments can easily be compared "on the screen" with other data sets including digitized geological and topographical maps. Graphic files such as coast contours or lineaments interpreted from other data sets can be displayed as overlays in various colours for a visual comparison with the data set under investigation. (4) Interpreted elements can easily be corrected, updated and integrated with other graphic data files. The system is menu-driven and user-friendly.

Remote sensing methods in ojfshore



Fig. 4. (a) FCC images (RGB) based on "raw" data (R), PCI (G) and PC2 (B). (b) FCC images (RGB) based on "raw" data (B), PCI (G) and PC2 (R), integrated with interpretations of linear features, (c) FCC images (RGB) based on "raw" data (B), PCI (G) and PC2 (R). Subareas framed in various colours defining two ROI's in red and blue, covering the offshore and onshore area, respectively. "Rose diagram" for each ROI showing Uneaments orientations, (d) FCC images based on "raw" data (B), PCI (G) and PC2 (R), showing lineaments with specified direction (coded blue).

Discussion of geological significance

In the study area it is of geological interest to investigate and clarify whether the well estabhshed Unear structural trends onshore can be traced offshore in the geophysical data sets. Linear features were interpreted interactively on the FCC-imageries onshore and offshore (see Fig. 4b). The onshore lineaments are confirmed by Landsat-TM images and the Geological Map of Norway (Sigmond et al.,

1984). The linear features offshore are compared with seismic interpretations (internal Statoil). The comparison of these offshore and onshore features is carried out with the above-mentioned novel statistical program module. "Rose diagrams" showing lineament orientations in a Region-Of-Interest (ROI) offshore can be displayed on the colour monitor for visual comparison with "rose diagrams" of a ROI onshore. It is evident from Fig. 4c that the linear features onshore and offshore display similar orien-


B.M. Scether, KG. Ruesl&tten, E. Rundhovde, C. Fichler, T.H. Henningsen and S. Johansen

Fig. 5. (a) FCC images based on "raw" data (B), PCI (G) and PC2 (R), integrated with main structural elements interpreted from seismic, (b) FCC images based on "raw" data (B), PCI (G) and PC2 (R), integrated with Palaeozoic carbonate deposits interpreted from seismic and well-log information, (c) FCC images based on "raw" data (B), PCI (G) and PC2 (R), subarea showing the southern extension of the Nordkapp Basin with the Gjesvaer "low".

tations. Specified lineament directions may also be colour-coded and displayed on the screen as shown in Fig. 4d for studying directional trends onshore/offshore. Various types of thematic information can be integrated with the processed imageries to facilitate the interpretation. Figure 5a shows a processed gravimetric image integrated with the main structural elements from seismic of the near base Cretaceous unconformity, and a good correspondence of the relief-contrasts in the gravimetric data set and the structural elements is seen. This probably indicates reactivations of the Palaeozoic structural elements in

Early Cretaceous time, in connection with the late Kimmerian tectonic phase. Palaeozoic carbonate complexes, interpreted from seismic data and well information, are integrated with the gravimetric imagery of Fig. 5b, and a good correspondence is observed between the carbonate reefs and the gravimetric highs. This correspondence is explained by a transgression in the Late Permian which resulted in carbonate reefs building up along the basin margins (Jensen and S0rensen, 1992). These basin margins represent stable basement highs during a long period of time. Within the processed gravimetric image the main


Remote sensing methods in offshore exploration

Fig. 6. (a) Free-air gravity anomalies based on GEOSAT altimeter data; grid resolution 5x5 km. (b) Marin Bouguer gravity anomalies covering the same area as in (a); grid resolution 1x1 km.

lineament framing the western part of the Nordkapp Basin reveals a SW extension towards the Gjesvasr "low" on the platform (Fig. 5c). In the continuation of the T-K FZ these linear features are cross-cut and reflect a small dextral movement in the order of 2-5 km (Johansen et al., 1993); see Fig. 5c). This right-lateral displacement is not identified by seismic interpretation. The displacement is oriented parallel to the T-K FZ, which is a dominating structure in East Finnmark. This observation indicates that the basin development during Palaeozoic time is influenced only by a small lateral displacement in this area (Johansen et al., 1993). Deep-seated structural features between the Loppa High and the Nordkapp Basin are also difficult to outline with traditional seismic mapping. These structural configurations are, however, clearly revealed on the processed gravity data set (see Fig. 4a). Gravity imagery based on satellite altimeter data is compared with a subset of the conventional gravity data set covering the same area. The satellitederived free-air gravity anomalies are presented in Fig. 6a, and the marine Bouguer gravity anomalies in Fig. 6b. Since the bathymetry of the shelf area does not account for larger gravity effects, the imageries may be compared. The gravity anomahes observed in the two imageries are almost identical, and all major structural elements are clearly defined. The satellitederived gravity data set displays a high quahty, but less details are seen due to the coarser grid resolution. Conclusions The main conclusions from this work are as follows.

The processing of gravity data sets using remote sensing methods provides an enhanced visual resolution which leads to (1) improved precision in the interpretation compared to ordinary contoured gravity maps, and (2) revelation of subtle trends in the potential field. The integration of potential field data with interpretations from various data sets (e.g. seismic) improves the final interpretation. In the Barents Sea region there exists a good correspondence between Palaeozoic structures and the gravimetric field. In areas where seismic mapping of deep-seated structures is uncertain, the use of processed gravity data is particularly helpful and provides additional information about basin configuration. Satellite data combine easy and fast access, low price and increasing quahty with new satellite technology and processing methods. Vast areas can be screened cheaply and rapidly by using various types of satellite-derived data. In frontier areas the planning of seismic surveys may even be based on interpretations of potential field data. The work is simplified by a user-friendly remote sensing system which permits interactive interpretation, statistical treatment and manipulation of the data. Acknowledgements We would Uke to thank Statoil for permission to pubhsh this paper, Statoil Exploration Office Harstad (STNN-LET) for providing the data; special thanks to Tore Svana and Erik Henriksen for supplying their interpretations of the Palaeozoic carbonate deposits, and Asle Str0m for excellent technical assistance. Finally we want to thank A.T. Buller for carefully


B.M. Scether, KG. Ruesl&tten, E. Rundhovde,

reading through the manuscript and correcting the EngUsh language. References Farrelly, B., 1990. The geodetic approximations in the conversion of geoid height to gravity anomaly by Fourier transform. Report Norsk Hydro Research Centre, Bergen, 21 pp. Gabrielsen, R.H., Jensen, L.N., Kalheim, J.E. and Riis, R., 1990. Structural elements of the Norwegian continental shelf, Part I. The Barents Sea Region. Nor. Pet. Direct. Bull., 6, 33 pp. Gudlaugsson, S.T, Faleide, J.I., Baglo, O.E., Breivik, A. and Stensen, L.R., 1990. Sen-paleozoisk struktur og tektonikk i det s0rvestlige Barentshav. Intern prosjektrapport, StatoilUniv. Oslo, 40 pp. International Imaging Systems, 1990. System 600 Digital Image Processing System — Command reference. Version 4.0 (Vol. 1+2+3). Jensen, L.N. and S0rensen, K., 1992. Tectonic framework and halokinesis of the Nordkapp Basin, Barents Sea. In: R.M. Larsen and B.T. Larsen (Editors), Structural and Tectonic Modelling and Its Applications to Petroleum Geology. Norwegian Petroleum Society Special Publication 1, Elsevier, Amsterdam, pp. 109-120. Johansen, S.E., Henningsen, T, Rundhovde, E., Saether, B.M., Fichler, C. and Rueslatten, H.G., 1993. Continuation of the Caledonides north of Norway: seismic reflectors within the basement beneath the southern Barents Sea. Mar. Petrol. Geol. 1994, Vol. 11, Number 2, pp. 190-201. Johansen, T.A., Haarden-Pedersen, G. and Samuel, P., 1991. Con-


Statoil Statoil Statoil Statoil Statoil Statoil

C. Fichler, T.H. Henningsen

and S.


version of satellite altimeter data to free-air gravity anomalies. Tech. Rep. No. 45, The Nansen Environmental and Remote Sensing Centre, Bergen, 56 pp. Moore, G.K., 1983. Objective procedures for hneament enhancement and extraction. Photometric Eng. Remote Sensing, 49(5): 641-647. Olesen, O., Roberts, D., Henkel, H., Lile, O.B. and Torsvik, T.H., 1990. Aeromagnetic and gravimetric interpretation of regional structural features in the Caledonides of West Finnmark and North Troms, Northern Norway. Nor. Geol. Unders. Bull., 419: 1-24. Roberts, D., 1985. The Caledonian fold belt in Finnmark: a synopsis. Nor. Geol. Unders. Bull., 403: 161-177. Roufosse, M.C., 1987. The formation and evolution of sedimentary basins in the Western Barents Sea. In: J. Brooks and K. Glennie (Editors), Petroleum Geology of North West Europe. Graham and Trotman, London, pp. 1149-1161. Sabins F.F., Jr., 1987. Remote Sensing — Principles and Interpretation. WH. Freeman, San Francisco, Calif., 449 pp. Saether, B.M., Rueslatten, H.G. and Henningsen, T.H., 1991. Advanced techniques for lineament analysis. 8th. Them. Conf. Geol. Remote Sensing, Denver, 1991. Schwartz, K.P, Sideris M.G. and Forsberg, R., 1990. The use of FFT techniques in physical geodesy. Geophys. J. Int., 100: 485-514. Siedlecka, A. and Siedlecki, S., 1967. Some new aspects of the geology of Varanger peninsula (Northern Norway). Nor. Geol. Unders. Bull., 247: 288-306. Sigmond, E.M.O., Gustavson, M. and Roberts, D., 1984. Berggrunnskart over Norge-M. 1:1 mill. Nor. Geol. Unders., 1984.

Research Centre, Postuttak, N-7005 Trondheim, Research Centre, Postuttak, N-7005 Trondheim, Research Centre, Postuttak, N-7005 Trondheim, Research Centre, Postuttak, N-7005 Trondheim, E&P, M0lnholtet 42, N-9401 Harstad, Norway Research Centre, Postuttak, N-7005 Trondheim,

Norway Norway Norway Norway Norway