Creating GM with both surficial lithologic map and borehole lithology
An ongoing problem that I am trying to solve is: how can I create a geological model (GM) that uses borehole lithologic and also uses an existing surficial lithologic map (vector polygons; e.g., shapefile)? I want the resulting GM to be informed by the surficial map in areas where boreholes are sparse, and also accurately represent the surficial geology (i.e., the surface of the GM should match the surficial lithologic map). As yet, I cannot find a tidy way to do this. For example, the red unit in the attached screenshot is the bedrock unit in a model with 6 layers. In some broad areas, however, it is exposed at the surface continually and I cannot force my model to honor this. Ideally, an approach would be systemic and would also work for complex surfaces, like those in the screenshot.
So far, the workarounds I've tried include adding the following data to the models contact surfaces:
- Creating a set of 3D points assigned to the geologic unit surface, equally spaced within each surficial geologic unit's polygons
- Creating polylines along the unit contacts that match the surficial map
- Creating equally spaced points along the unit contacts that match the surficial map
No matter what I've tried, the GM's surface fails to match the surficial map. Conceptually, it seems like this functionality should be possible, given that the surficial map and boreholes should complement each other. Yes, I have watched and tried the lesson "Building a Model from a Map."
I'm open to trying anything. Thanks for any ideas.
Answers
-
Hi Jackson,
The things I would look like trying here are:
- As you've probably seen in the 'Model From A Map' you can create the surface from the drilling then 'Add —> GIS Data' to the surface as an additional data, you've likely tried this, but I just wanted to confirm that's the route you've gone down.
- Utilising GIS lines rather than polylines, if you have a topo surface GIS lines can be drawn without worrying too much about elevation, then draped onto the topography surface. Polylines can travel above and below the topo easily and often you find if the polyline is slightly above the topo, it's technically outside the model boundary and won't be honoured.
- As an extension of the previous point you could look into turning off the 'Boundary Filter', the default will only let the surfaces see within the boundary box, this does let the GM run more efficiently by not looking for everything, but this might be limiting those polylines being 'seen'
- Also have a look into the snapping rules for the GM, if the snapping is honouring the drilling only this could explain why it's not matching your lines, you can change the snapping to 'All Data' or 'Custom' and tick the surface lines.
If none of these seem to do the job please shoot us an email at support@seequent.com, for us to have a more in depth look with you.
Cheers,
Lee
Lee Evans
Project Geologist, Seequent, APAC
2 -
Hi Lee,
Thanks for the response and ideas.
- I have been adding bonus GIS data
- I did not think about that—thanks. I've switching to using the "on topo" versions of lines
- If I turn off the boundary filter, then the resulting surface of the model becomes a rectangle. One approach may be to turn off the "Use Topo" feature when model building, and then post-processing by trimming to the DEM, deleting any "Unknown" units, and then filling any gaps between the uppermost layers and the DEM by vertically extruding the surface.
- I'm using custom snapping, and snapping to the surface contacts (known), and any manually added points; boreholes remain unsnapped.
I've learned that the feature I desire isn't built-in. That said, I see a need for it and would love to see it programmed into a future version: a way for the model building to match surfaces beyond just snapping to unit contacts (lines). My workarounds—adding systematic points and lines, created outside of Leapfrog [ArcGIS Pro or Python] to mimic the contacts or surfaces—get me close, but still require manual editing to clean up. Manual editing includes adding points, points with tangents (or structural points), and polylines near contacts to better shape contact surfaces. In small study areas, it isn't bad. In large study areas with complex surfaces, it will be a big ask.
If there is anything else that you think I could try, or want to see my workarounds in action, I'm all ears. Thanks!
0 -
Hi Jackson,
Sounds like you might have turned off the topography directly in the ''boundary tab', try instead the drop down in the 'General' tab, the boundary filter there is for the data seeking not the shape of the GM itself.
It would be good to catch up over teams or share the project via our support channel if you'd like, snapping to specific data and honoring what you need to hasn't been an issue we can't help with in the past. Shoot us an email to support@seequent.com and we can tee something up.
Cheers, Lee
Lee Evans
Project Geologist, Seequent, APAC
0 -
Hi Lee,
Sounds good. I already have an ongoing support ticket that is trying to be tackled by somebody else. I posted here to see if any Community members had ideas, but I'm also happy to discuss with more/other Leapfrog/Seequent staff to better tackle my conundrum. See attached for a conceptual diagram. It is easy to force a unit/surface down so that it isn't exposed at the surface; however, it is difficult to consistently force a unit up so that it is always exposed at the surface in the desired areas. Because of varying topography, it seems that there is often a patchy, thin veneer of other units (or "unknown") that gets brought to the surface because my attempts to force the desired unit to the surface do not work over all spaces.
Thanks for your time and help.
0 -
Lee,
Good catch—I turned the boundary filter off in the GM's General table, while leaving the "Use topography" checked in the "Boundary" tab. This allowed the interpolation to use data beyond the boundary (that is, above the DEM), which absolutely improved the efficacy of my workarounds. That said, it still requires adding an extremely dense grid of "surface" points to catch everything systematically. I am closer than I was, but would still love any other programmatic approaches to achieve 100% consistency with the surface map without manual editing (save the manual editing for the more interesting geologic interpretations). If Unit G (e.g., deepest) appears at the surface at a given XY coordinate but the surface map indicates that something else is at the surface, I have to manually force Unit G down—if this was automatic somehow, that would be neat. Thanks for the help!
0 -
Yeah the issue is the data format as it appears to be simply an image, which would take that from being a simple data input issue to image recognition functionality, however if the file was exported as .dxf shapes as well (which fyi you can do in cross sections) they could in theory be brought in as vector shapes, then be used as direct GIS lines in the model.
See for more: https://help.seequent.com/Geo/2024.1/en-GB/Content/presentation/cross-section-in-scene.htm?Highlight=cross%20section#export
But I'm glad we solved that other one anyway! Good luck mate!
Lee Evans
Project Geologist, Seequent, APAC
0 -
Sorry just to follow up I didn't see the other comment, in these kinds of situation where the entire domain is meant to be pink, now that you can create a line above the topo, I would consider putting some 'anchor points' above the mountain area to essentially 'dome' the surface above it. I suspect you may have figured that out already but just in case.
Lee Evans
Project Geologist, Seequent, APAC
0
