🚀 New Group Alert: Estimation and Advanced Geostatistics
We’ve just launched the Estimation and Advanced Geostatistics Group in the Seequent Community!
This group is your go-to space if you’re working with Seequent’s geostatistics solutions such as Leapfrog Edge, Evo, BlockSync, and more. Here, you can:
- Connect with peers across the industry
- Share knowledge, workflows, and practical tips
- Stay informed on product updates and new innovations
- Discuss challenges and collaborate on solutions
Whether you’re deep into resource modelling or just exploring best practices, this group offers a place to learn, exchange ideas, and grow together.
👉 Join today and help shape the discussions from the start.
Let’s make this a thriving hub for all things geostatistics!
Comments
-
Hello,
I’ve been reviewing the topic of resource classification, and in my opinion, there are many subjective variants on how to classify properly. Some approaches rely on percentages of the variogram range, others on drill spacing, the creation of envelopes based on average drillhole spacing, and finally on drillhole spacing itself.
In particular, there’s a methodology that applies the latter in a highly effective and reliable way. It not only requires a minimum of 3 drillholes but also includes an additional one at shorter distance, ensuring that the blocks are classified more appropriately. This can be achieved quite simply using AvD in the outputs of any estimator: by applying 1 sample per drillhole, with a minimum of 3, performing a geometric adjustment by multiplying by the square root of 2, and then applying the classification based on drill spacing—or whichever criterion ensures greater confidence, since it also depends on the quality of the geological model.
I can provide the script for this if you get in touch through this channel.
Best regards!
1 -
Hi Fernando,
Thanks for your comment.
Classification is a topic guaranteed to spark discussion and cause dissent in any group of resource geologists.
In this year's Parker Challenge event (held in conjunction with AusIMM's Mineral Resource Estimation Conference), classification was again the single largest cause of difference between entries.
This is not actually surprising - classification is deliberately not prescriptive in any of the reporting codes because there are so many potential factors that need to be incorporated. All codes are very clear however, about the need to be open and transparent about exactly HOW resources are classified i.e the method and decisions applied.
A few comments:
The 3 hole method is widely used as an input to classification - exactly as you describe (with our without the root2 correction).
As a possible alternative take a look at Leapfrogs 'declustering object'. While the main purpose of the declustering tool is to derive declustering weights, if this object is evaluated onto a block model, it produces a map of sample density. That can be a quick and useful way to quantify/separate areas of differing sample density. The underlying maths is described here https://www.seequent.com/understanding-declustering-in-leapfrog-edge/
Conditional simulation is increasingly used for 'classification support' (if not directly for classifying).
Recommend this recent (open access) paper for a very thorough review of existing approaches supported by an extensive reference list, and a proposed new method.
https://link.springer.com/article/10.1007/s11053-025-10470-52
