Modelling alteration as numerical data

YirimamboBonyaah1
YirimamboBonyaah1 Posts: 4 Calcite Rank Badge
in Leapfrog

What is the best way to model alteration? Is it by modeling it by categorical data or changing the categorical to numeric or still use the alteration intensity for the modeling?

Comments

  • FrançoisBuscail
    FrançoisBuscail Posts: 12 Fluorite Rank Badge

    Hello,

    It depends! What do you want to do, geological model or numerical model?

  • ScottBriscoe
    ScottBriscoe Unconfirmed Posts: 2 Calcite Rank Badge
    edited September 15

    What type of deposit is it? Some systems are favorable to a zonation style model which could be good to do numeric models for each alteration type. You could then create a combined model that classifies the type of alteration in a given block by assemblage types. So for example numerical model for quartz intensity, pyrite intensity, and sericite intensity could be used to generate a QSP alteration score. Likewise if you get more involved and you also model potassium, biotite, magnetite, chlorite, epidote, calcite, albite you could then create very involved model categories with if statements to classify phyllic, potassic, propylitic alteration scores. It's a big topic to ask about as it also needs to feedback to how logging is done. Core logging should be directly usable by geomodelers but it rarely is. If your logging is intensity of phyllic, potassic etc then you might be better off creating categories weak, medium strong of each.

  • YirimamboBonyaah1
    YirimamboBonyaah1 Posts: 4 Calcite Rank Badge

    @FrançoisBuscail I want to do both but which onw is the best practice

  • YirimamboBonyaah1
    YirimamboBonyaah1 Posts: 4 Calcite Rank Badge

    @ScottBriscoe

    The deposit is orogenic gold deposit.

  • ScottBriscoe
    ScottBriscoe Unconfirmed Posts: 2 Calcite Rank Badge

    So orogenic systems you could numerically model chlorite, calcite, ferroan dolomite (sometimes simplified to ankerite), quartz, muscovite, pyrite, pyrrhotite, k-spar, amphiboles and then create categories of alteration by the ratios of those different things to eachother and then create a geological model of those categories. If you had hyperspectral data you could look at changes in white mica, changes in carbonates, and chlorite as many orogenic systems show wonderful zonation in hyperspectral that's not always apparent to the eye.

    Best practice is a whole other question- really it depends on what the resource estimators you work with will actually use. You could create something that everyone agrees is truly best practice only to find that the resource team will never use it. So needs to be built for the purpose of creating good resource estimation, good decision making and good mine design or whatever your end use case for the model is. That's the challenge- there is no easy single answer of how you should do this.

  • FrançoisBuscail
    FrançoisBuscail Posts: 12 Fluorite Rank Badge

    As this is an orogenic deposit, you need to produce both models.

    You need to know about the alteration intensity and alteration minerals if you are proximal or distal from your hydrothermal system. And with the geological model, you need to model your lithologies and the deformation to determine your domains if there are several.

    Before building a model, go back to the core. You absolutely need to know what is guiding your mineralization. Making resources is not just linking colour points on a section.

Categories