François Marillier, Urs Eichenberger and Anna Sommaruga
Institut de Géophysique
1015 - Lausanne
Work in conducted in 2005 included signing several contracts that give us access to seismic data, reports or digital data not publicly available, interpreting seismic lines in the Canton de Vaud and in Fribourg, setting up a GIS data base and developing special routines for it. Also, our project and preliminary results were presented at meetings and in a publication, and contacts with several scientists were established.
Durant l’année 2005, des contrats ont été signés qui nous donnent accès à des données sismiques, des rapports ou des données numériques qui ne sont pas disponibles publiquement ; l’interprétation des lignes sismiques a été menée dans les cantons de Vaud et Fribourg ; une base de données sur SIG a été mise sur pied et des routines ont été spécialement développées pour cette base de données. De plus, notre projet a été présenté à différentes réunions scientifiques ainsi que dans une publication ; nous avons également rencontré plusieurs chercheurs.
The project to interpret seismic and borehole data mainly from the petroleum industry in the entire Swiss Plateau is now in its second year of activity. The aim of the project is to provide a synthesis to address fundamental geological issues related to the Molasse Basin, to the Mesozoic fill and to the occurrence of Permo-Carboniferous basins. Also, this work will help address large scale scientific problems such as the evolution of the alpine foreland basin and the development of the Jura mountain fold belt. Besides its scientific interest, this project will provide regional information that may help address other issues such as the search for commercial hydrocarbon, storage of nuclear waste, better understanding of large scale groundwater circulation, heat-flow distribution and safety of main urban and industrial sites.
One of the final products of the project is an atlas that will contain a series of major transects across the Molasse Basin, a summary of well data, as well as subcrop and isopach maps of major structures that involve Cenozoic (Tertiary), Mesozoic (Secondary) and Pre-Mesozoic strata. The output will be available in digital form through a GIS data bank that will not only allow on screen visualization of the various products, but it will also provide cross-referenced access to the results with all data handling possibilities offered by a GIS based system.
During 2005, activities included gaining access to data from government offices and other offices or companies through special agreements, interpreting the data in western Switzerland, setting up a GIS based system that can be used both for data interpretation and for data visualization, and reaching out to other scientists through individual discussions, scientific meetings, and a publication in ACTUEL, GeoforumCH.
The bulk of the seismic data became available to the project after signing the contract with SEAG (see 2.2). Besides of line trace data, copies of several shot point and horizon maps of the basin were gathered. The public well data (which we got in 2004 from Bern) also was completed by documents from the SEAG archives.
From current thesis work at the Geneva University, we got 7 TWT maps of key horizons from Geneva, Vaud and Fribourg Cantons which were interpreted on a workstation. For the area of the canton of Vaud, seismic and well data is available to the public and was partly copied from the collection of the geological museum of Lausanne University. The Geneva University helped with missing copies of the Vaud area. For areas in Fribourg, we got data from the canton directory for environment and construction and some data was made available from the Fribourg University.
The Nagra data includes key seismic lines, public seismic interpretation reports, two internal reports and horizon data (Opalinuston) at the northern edge of the basin in digital format (see 2.2).
The surface geological data from the former BWG was completed and all maps available in GIS format from the Molasse basin were integrated in the project.
Several contracts were signed with companies or institutions during this year:
SEAG: As the SGPK has no legal status, the general secretary of SCNAT agreed to sign a contract with SEAG on our behalf. To establish a contract between SEAG (Schweizerisches Erdöl AG) and ourselves negotiations took place. On request of SEAG and SCNAT, data handling procedures at the Institute of Geophysics at Lausanne University were clearly formulated and made an integrant part of the contract. The contract was signed on 1st of February 2005.
Fribourg Canton (directory for environment and construction) gave us permission to consult and work on seismic lines from the Fribourg Canton.
NAGRA: In June 2005, a contract was signed with NAGRA. It gives us access to some specific seismic data, internal reports and regulates exchanges of digital data.
FREAG: During the summer 2005, Roger Multone, representing the FREAG company gave us access to the seismic lines under his concession, a small area located in the Canton of Fribourg.
BWG: Contracts were signed with the Federal Office for Water and Geology that give us access to several digital data of geological maps at different scale.
Interpretation work in 2005 focussed on the western most part of the Swiss Molasse basin (Fig. 1). It involved nearly 2000 km of 2D seismic. In the canton of Vaud, the acquisition between 1963 and 1986 resulted in a rather dense seismic grid with about 400 line crossings. 7 deep wells with geological stratigraphy and velocity data were available and 12 geological surface maps at 1:25’000 scale.
Figure 1: Seismic data coverage in Switzerland. In red, progress of interpretation until end 2005
As stated earlier, most of the data is available only on paper. Interpretation work therefore is done on big tables, using lots of colour pencils and rubber. A first step is the calibration of the lines by deep well data (Fig. 2).
Figure 2: The stratigraphic and geophysical information of deep wells is used to identify seismic reflectors. Seismic data are from a report deposited at the Musée géologique du Canton de Vaud, Lausanne.
Useful well data for this purpose are the ground elevation, the well track geometry, the stratigraphic interpretation of log data and the well shoots or vertical velocity surveys and the sonic logs. The TZ curves derived from such data allow carrying a stratigraphic horizon from the depth domain of the well into the time domain of the seismic line display. Synthetic seismic curves calculated from good well data help considerably to make the link between the high frequent well strata and the seismic reflector images (Fig. 3). Unfortunately, most wells have incomplete log records.
Figure 3: Seismic synthetic sections are used to correlate well stratigraphy and seismic reflection sections. These data were measured and calculated for the Essertines-1 well, drilled in 1991. Data are from a report deposited at the Musée géologique du Canton de Vaud, Lausanne.
Once the ground elevation of the well is fixed on the seismic line and the time intervals which correspond to the thickness of stratigraphic units is put down on the trace display, the main seismic reflectors can be identified.
The geological maps were printed with the overlay of the line positions and their shot point annotation (Fig. 4).
Figure 4: An example of geological maps with location of the seismic lines referenced for the Yverdon-Les-Bains (1203) geological map. Seismic lines are in red and shot-points are in black. Well locations are represented with a cross. The combination of surface geological information with seismic shot-point locations allows for more accurate near-surface interpretation of the seismic lines.
Relevant geological surface features such as faults and stratigraphic dips and limits were then put on top of the seismic line. Line crossings on the maps were checked against the annotation of shot points and intersections on each of the lines. The sections were interpreted by carrying the horizons from the well location into un-calibrated parts of the basin and from one line to the next while folding the paper at the line intersections.
Once the interpretation is consistent, we draw the key horizons, fault and thrust data on transparent paper. These sections are then digitized and the horizon and fault files put into the GIS system (Fig. 5).
Figure 5: Correlation between a digitized seismic section and table of attributes for horizons and faults in our GIS data base. Column “Horizon” contains the names of specific horizons or faults. Each horizon or fault has a different name.
The correction of miss-ties is done in several steps: The observed bulk shift at a single intersection of two lines is noted in a spread sheet. These values are needed for the correction of the calculated miss-ties from all intersections of the digital horizon data. A rather simple line hierarchy is established on well crossings and line quality in order to progress the miss-tie corrections.
The in 2006 established link to interpretation software allows for more elegant horizon gridding, fault handling and depth conversion.
By the end of 2005, we completed the area of the canton Vaud and advanced well in the two surveys of Fribourg south and north.
Seismic interpretation looks at the bulk of the data and distinguishes seismic facies which are characterized by a specific range of amplitude and frequency values and of more or less continuous horizons. The identification of seismic sequences bounded by stratigraphic unconformities and sometimes by faults is the starting point of the analysis of a sedimentary basin.
The better the data, the higher is the stratigraphic resolution. Good data is scarce for many reasons: old seismic data vintage, poor fold of the recording, insufficient static corrections and, in general, rudimentary processing. Most of the lines (1967, 1968, 1973, 1974) show a rather low signal to noise ratio. Therefore, it is often difficult to follow a horizon over a long distance, and we are forced to apply a stratigraphic and tectonic model, to carry it through.
The main horizons we mapped in the western part of the Swiss Molasse basin are Near Base Tertiary (BTer), Near Top late Malm (TlMa), Near Top early Malm (TeMa), Near Top Dogger (TDo), Near Top Liassic (TLi), Near Top Triassic (TTr), Intra Triassic (TMuka), Near Base Mesozoic (BMes), Intra Permo-Carboniferous 1 and 2 (PC1, PC2).
Figure 6: Seismic facies changes and thickness variations are best visible on seismic lines that are oriented perpendicular to the basin axis. This line was shot in the central Vaud area.
Three different types of faults are distinguished in our interpretation: Normal faults, strike slip faults and thrust faults.
The thrust faults are detected through the doubling of seismic horizons by a low angle plane. Within the Triassic series, in the Liassic shale, the marly Dogger, the Oxfordian clay stones and the various marly series of the Tertiary, there must be many more hidden thrusts than the ones we can identify on 2D seismic.
Two large families of normal faults can be identified: the normal faults offsetting the Mesozoic strata, reaching up into the tertiary sediments and the late Palaeozoic graben bounding normal faults. Only very rarely, the two are linked. In most cases the faults which offset the Mesozoic series are listric at the base and seem to die out in the Triassic evaporits.
The steep, often close to vertical faults with ill defined vertical offset are interpreted as strike slip faults. The regional context and local observation of flower structures confirm elongated zones with sinistral or dextral offset.
Synsedimentary faults: Many of the observed thrust-, normal- and strike slip faults seem to line up with zones which were active already during sedimentation in an extensional setting. Dehydration of the sediment and salt dissolution caused differential compaction and particular geometries in the overlaying sedimentary packages. The reactivation in the alpine compressive regime caused inversion of normal faults and offset sedimentary packages of different thickness along the strike slip faults.
Detachement horizons: The most prominent detachment horizons are the Triassic evaporates and the early tertiary claystones and marls. They separate tectonic styles and seem to accommodate for most of the alpine shortening throughout the basin. Some of the Permian graben bounding normal faults are clearly inverted and late Permian sediments are pushed out of the graben in a northern and western direction. However the total shortening in such features seems not to exceed a few kilometres.
Figure 7: Extensional faults and strike slip faults are best visible on seismic lines that are oriented parallel to the basin axis. This line was shot in the southern Fribourg area.
Completed interpretations of the seismic sections are digitalized and made accessible on a GIS (geographical information system). Canton Vaud area interpretations are now available in digital format in the GIS database. GIS database includes horizons with top or base of Cenozoic, Mesozoic and Palaeozoic formations and various types of faults and localisation of seismic lines (shotpoint map) all geo-referenced in space. Well log data and geological or tectonic maps are also loaded in the database.
The GIS strategy was developed in contact with IGAR (Institut de Géomatique et d’Analyse du Risque) at Lausanne University. Many discussions with Thomas Czaká and Prof. M. Kanewsky helped us to better define our needs and to find the easiest way to realize them.
Two persons helped us in the accomplishment of the GIS database: Baptiste Dafflon, geophysicist working on a civil service project at the Institut de Géophysique in Lausanne, and Robin Engler, master degree in Biology-Arcgis. Both gave a major input in the realization of the shot point map of the seismic lines and in the development of the GIS database containing the results of the seismic line interpretation.
Three user guides have been written by Robin Engler, which explain the precise steps for using ArcGIS and Kingdom Suite software for our project (see references at the end of the report).
A “sismique Plateau toolbar” in ArcMap has been developed in order to render easiest the procedures of:
- verification of digitized data
- extraction of horizon TWT from digitized seismic line files
- geo-referencing each TWT data: attribute a real X,Y,Z to each data
- merge all the horizon data in one unique shapefile
- calculate the value of “miss-tie” at the crossing of the seismic lines for each horizon
- data declustering in order to uniform the density of the points on seismic lines.
Figure 8: Extract form our GIS data base showing a digitized section and its correlation with a geo-referenced map. ID value corresponds to a shot-point geo-referenced in the Swiss coordinates system.
A precise shot point map of whole Switzerland has been realized in ArcGIS (example in Fig. 4). This map represents more than 800 seismic lines and around 40 important wells Map has been compared and corrected to the shotpoint annotation on the paper seismic sections. This shot point map is essential for a precise localization of intersections of seismic lines from various vintages and surveys and is the base map for drawing fault system. It serves also as reference data for geo-referencing the digitalized horizons.
Interpreted seismic horizons of the Canton Vaud and partially Canton Fribourg have been digitized on a table from transparent paper where faults and seismic horizons were drawn. At this stage reference data X corresponds to value of shotpoints and Y to the value of TWT. Digitalization points are not chosen regularly but in function of the change of the geometry of the line.
Digitized horizon points have been resampled in order to obtain for each shotpoint unit a TWT value which is geo-referenced. TWT values points are extracted and associated to real coordinates. A data declustering function has been created in order to uniform the density of shot points along a seismic line. This is important later for the interpolation of the data.
Digitized fault points have been maintained, because of the geometry of the faults. One unique shotpoint corresponds in some cases to two TWT values (e.g. faults associated to flower structures).
Due to different surveys and vintages, TWT miss-tie is present for a single horizon at crossing lines. It is necessary to take in account this problem, in order to avoid shift of value without geological signification. Miss-ties are identified during the interpretation phase and then integrated in an Excel file. A macro controls the coherence of the miss-ties and afterwards they are corrected on the base of a degree of priority of the line.
Public seismic data should be charged in the SIG data base during the next year and be available in formatted way.
Assemblage of tectonic map sketches from Switzerland at 100’000 scale (Fig. 9) and raster or tiff-scans of geological maps at 25’000 scale have been formatted and geo-referenced in the ArcGIS data base. Around ten geological maps with seismic line location and shotpoints from western Switzerland have already been printed.
Figure 9: Compilation of geological sketches from geological maps. Dataset Geological maps 25’000, © BWG, Bern.
Several institutions use GIS systems to compile geological datasets country wide. In order to discuss the development of the ideas, practical application of them and coordination of GIS geodata base, we organized a meeting in Bern including participants from the Federal Office of Topography, the Federal Office for Geology and Water, the NAGRA and the IGAR. With NAGRA, we signed a contract allowing an exchange of a file containing a geo-referenced horizon (Opalinuston) (see 2.2.2).
During 2005 we met several scientists interested in the project.
François Vuataz, CREGE, Centre de Recherche en géothermie, Neuchâtel ;
Laurent Tacher, EPFL ;
Thomas Kohl, Sarah Signorelli , Geowatt ;
Oliver Kempf , Molasse meeting in Gwatt;
Robin Marchant , Musée de géologie Lausanne ;
Hans-Andreas Jordi, Marc Weidmann, Scientists:
SIG (See chapter 4)
Jean-Pierre Berger, Jon Mosar, University of Fribourg;
Martin Burkahrd, University of Neuchâtel.
Our work was presented on several occasions: we gave a conference at the Sediments’05 scientific meeting (Gwatt, July 2005), a short publication was written for ACTUEL, GeoforumCH (2, 2005), and our work was presented at the seminar of the Institute of Geophysics, University of Lausanne.
We thank SEAG, University of Fribourg, Geological Museum in Lausanne, Nagra, BWG-Archives, Swiss topo, BWG for providing data for this work. Francis Perret carried out the compilation of geological sketches and helped with the drafting of many figures.