Frontier Exploration: Avoiding Interpretation Bias

We welcome our new Senior Geologist Kirsty as she presents her first Cognitive Whiteboard.

Click on the image below to see the whiteboard in more detail.

whiteboard 1.png


Hello and welcome back to the Cognitive Whiteboard. I'm Kirsty, one of the new Geo's here at Cognitive Geology. My background is as an explorationist and I've joined the company to bring some of those exploration workflows into the products that we're developing. In preparation for today's video, I was watching some of Luke's old blogs, and one of the things that struck me was Luke talking about the difference between accuracy and precision, and that really struck a chord with me and some of the learnings that I had to take on board as a young geologist learning my trade in exploration.

So, early on, when I was a youthful geologist at the cliff face with my hand lens, I was looking at the really fine detail, getting excited about the bedforms that I could see, walking along the whole outcrop, seeing all the contacts and mapping them really accurately. When I joined BG group in 2009, and started working in exploration, I discovered that we really rarely have that kind of luxury of data density. In fact, if we start working in frontier basins, for example we've got a basin here maybe 4 or 500 kilometres in diameter with a handful of wells that we can you use to learn about the basin from. So, we've got very fine vertical detail with big horizontal spacing, tens, hundreds of kilometres between them, and from those, we're very comfortable with our vertical detail and it allows us to build depositional models. We can use those depositional models to show the depositional environments in map form, and this really looks like the kind of map that I would have drawn early in my career at BG Group. It looks really nice and geological, it looks like a snapshot of something we might see today in a depositional system.

 It's very appealing. But what it also is, is a precise view of one single possible interpretation of the data that we're seeing in my depositional model, and it doesn't encompass the full range of time and geological processes that are happening in that basin. So, if you draw something like this, you risk becoming really anchored to that one interpretation very early on with very little data to tie it back to. I would go back and tell myself 10 years ago that what I really need to be is accurate but encompassing the full range of possibilities, and therefore, what I should be drawing is not lovely geological facies maps like this which look great. But things that really represent what I actually know right now and that would be a play fairway style map where I'm showing where I believe the sand dominated facies to be. Where I believe the mud dominated facies to be. That still allows me to focus my exploration, recommend which blocks we should be bidding on in a license round. But it also doesn't tie me to an interpretation that I can't really back up at the moment with my data. And we also risk, because this looks so nice, it tends to get used again and again and again within the company and we've all seen this happen. And if this gets persisted 10 years down the line, when I'm not involved in that asset anymore, it can really bias people's decision-making and we don't have that information when this was drawn, and therefore, we could miss some really good prospectivity as a result of tying ourselves to one specific interpretation in time. Well, that's all I've got time for right now. But I look forward to seeing you back at the Cognitive Whiteboard soon.