OR/14/024 Modelled faults
| TERRINGTON R L, WILLIAMSON J P. 2014. Metadata report for the Craven Basin 1:250 000 resolution geological model. British Geological Survey Internal Report, OR/14/024. |
All of the faults used in the construction of the model were sourced from the Craven Basin Subsurface Memoir (Kirby et al., 2000). These were digitised per surface as ESRI polyline shapefiles as part of the Regional UK Lithoframe programme. As the model was regional in scale, an initial filtering of the faults was applied to ensure that only those that had significant throw/displacement (>100 m) were used for the surface construction.
To do this, vertical faults were selected that had a length of greater than 10 000 m, as it is recognised that the greater the length of the fault (in 2D space) the greater the throw/displacement (Young-Seog and Sanderson, 2005). By selecting faults over 10 000 m, the majority of the faults with throws of greater than 100 m would be selected for use in the modelling phase. These fault polyline shapefiles were imported and grouped by the surface in which they displaced.
All of the ‘non-vertical’ faults classified in the original Craven Basin EarthVision project were selected as there were relatively few of these, and these were interpreted as having a significant heave in the original model.
The total distribution of faults modelled can be seen in Figure 6.
Two types of faults were produced using the Structural Modelling Workflow in GOCAD®. Some of the faults were produced using the fault sticks method and the rest used the Fault Centre Line method. For both types of faults generated, the Structural Modelling Workflow in GOCAD® was used to develop the contacts between the faults.
The fault traces digitised in ArcGIS have no Z-Value (elevation) associated with them, being in effect lines on a map. Therefore it was necessary to attribute them with an elevation for use in the 3D model construction. The method used was to construct an unfaulted surface model for each horizon and then drape the fault traces relating to each horizon on its surface. This gives an approximate fault elevation per model horizon.
The Z-attributed fault traces were then displayed together in 3D, and given a colour relating to the horizon it is associated with. These fault traces were then visually inspected to find their related fault trace at different levels which indicates a coherent structure cutting several horizons. When such a spatial correlation was observed, a new object was constructed from several different GOCAD curve objects that make up the structure. The Fault Sticks section (3.1) describes this workflow.
When a significant fault was observed that is unique to a single horizon, or whose relationship to fault intersections on horizons lying above and/or below is unclear, then the fault was modelled as locally vertical. The Fault Centre Line section (3.2) describes this workflow.
Fault sticks
- Each surface was calculated using the contours digitised from the Craven Basin memoir to give a raw unfaulted surface.
- The fault traces generated were draped onto their respective surface. For example, the Permo-Trias fault dataset was draped onto the Permian-Triassic raw surface. This was repeated for each of the fault datasets per horizon.
- Faults were grouped by their fault location. For example, each of the fault traces for the F1016 fault was grouped into one curve object in GOCAD® (Figure 7).
- The grouped fault traces were allocated a Fault Stick data type in GOCAD® and put through the Structural Modelling Workflow for generating a fault surface. Manual editing of the fault surface was sometimes necessary to smooth out any spikes or anomalous data.
Fault centerlines
If the fault trace only occurred on a single surface, the fault centre line data type was used in GOCAD® Structural Modelling workflow. The fault centre line allows the user to specify the elevations of the top and the base of the fault and the dip angle. The fault generated is a simple extrusion of the fault centre line.