Is ‘Vectoring’ to Mineralisation a Claytons Jargon?

“Whenever you find yourself on the side of the majority, it is time to pause and reflect”

– Mark Twain

Recently, I’ve frequently read and heard phrases like ‘Vectors to Mineralisation’, ‘Geochemical Vectors’, or ‘Vectoring’ used when referring to a wide zone of alteration around an existing ore body. By studying these ‘vectors to mineralisation’ from known deposits, it is hoped to identify the location of an as yet undiscovered ore deposit at another site.

I’m sure you’ve heard this ‘vectoring’ talk yourself because it’s popular at the moment. It’s mentioned by geochemists, exploration geologists, and company executives alike, and, when they talk to each other, it’s as if they know something that you don’t. If you go to a geological convention, you’re bound to find at least one author talking about the latest whiz-bang technique of ‘vectoring to mineralisation’. If you read a technical report released by public companies you’ll discover that ‘vectoring’ is mentioned in many of them (Figure 1), so many geologists are happily riding this 'vector' bandwagon in 2016.

Figure 1. A selection of the use of the various uses of the term ‘vector/ing’ in seven separate JORC and NI 43-101 technical reports published in 2015 and 2016 (RSC MME, 2016).

Perhaps the use of the word ‘vectoring’ could be dismissed as a ‘weasel word’ used by mining and exploration companies, but a quick Google search reveals this term originates from academic studies published in mainstream geological journals over the last 20 years—it wouldn’t be appropriate to dismiss it casually.

Well, today I’m going to tell you what I think of ‘vectoring’ after conducting some research into the original academic literature. I conclude that for most published papers this term is pretty much a weasel word.

Weasel Word—an informal term for words and phrases aimed at creating an impression that a specific and/or meaningful statement has been made, when only a vague or ambiguous claim has been communicated, enabling the specific meaning to be denied if the statement is challenged. (Wikipedia contributors, 2016)

In Australia and New Zealand we would refer to these words as Claytons Jargon—a jargon you use when you’re not using a jargon (Figure 2).

Figure 2. Claytons, a non-alcoholic drink, was promoted in Australia and New Zealand in the 1970s and 1980s as ‘The drink you’re having when you’re not having a drink’. The name has entered into Australian and New Zealand vernacular where it means something ineffective or meaningless.

In this blog post, I’ll illustrate why these ‘vectoring to mineralisation’ terms are misused in the mining industry and provide specific reasoning for my conclusion.

Words produce pictures in your mind

When I read the word ‘vectors’, I imagine, well, vectors. You know, a bunch of arrows, which I learned in an elementary math class. For those who have long forgotten elementary school subjects, here’s a definition:

Vector—a quantity having direction as well as magnitude, especially as determining the position of one point in space relative to another.

In many earth sciences, a vector can be viewed as one component of a vector field, such as a displacement vector field around a fault, but in the case of the phrase ‘vectors to mineralisation’, what I see in my mind is a series of vectors pointing the way to the same location—a zone of high-grade mineralisation. Illustration of such vectors, of course, would be a useful way to visualise the location where exploration companies should be drilling, because it’s comparable to a street sign, pointing the way to where exploration companies would want to explore next.

But there’s a slight hitch:

When I scanned the research papers that mention ‘vectoring’, not one picture of a vector appears in any of the publications that use this word in the title or in the body of the text

Yes, you read correctly. Not ONE.

What’s more, this has been going on for more than 20 years.

Evidently, this has not bothered any of the authors or the academic reviewers of these published papers, or even the audience who read these articles, as they don’t seem to have noticed this glaring omission—a failure to illustrate the actual subject matter listed in the titles of the papers.

If you don’t believe me, here are some titles of the research papers (my emphasis):

Target VECTORING using lithogeochemistry: Applications to the exploration for volcanic-hosted massive sulphide deposits (Galley, 1995).

Lithogeochemical halos and geochemical VECTORS to stratiform sediment hosted Zn–Pb–Ag deposits, 1. Lady Loretta Deposit, Queensland (Large and McGoldrick, 1998).

Element ratios in nickel sulphide exploration: VECTORING towards ore environments (Brand, 1999).

Vertical lithogeochemical halos and zoning VECTORS at Goushfil Zn–Pb deposit, Irankuh district, southwestern Isfahan, Iran: Implications for concealed ore exploration and genetic models (Hosseini-Dinani and Aftabi, 2016).

The word ‘vector/s’ appears more than 20 times in some of these papers, yet not ONE image of a vector in 3D space pointing to the predicted location of mineralisation is shown in any of these papers. Although I haven’t exhaustively searched all the literature that discusses this subject, I’ve viewed many key papers that are cited in the literature, and this lack of imagery appears to be a pattern.

Confused definition of ‘vector’

If you look at this paragraph from Large and McGoldrick (1998), the word ‘vector’ is used in a very peculiar way (again, my emphasis):

"Whole rock analyses are used to calculate the three VECTOR QUANTITIES as follows: (1) SEDEX metal index = Zn + 100Pb + 100Tl; (2) SEDEX alteration index = (FeO + 10MnO)100/(FeO + 10MnO + MgO); (3) manganese content of dolomite: MnOd = (MnO x 30.41)/CaO. All three VECTORS increase to ore both across strike and along strike."

Astute readers will notice that the quantities described by the authors are not vectors; instead, they are scalar quantities.

I thought maybe there’s another geochemistry-specific definition of the word ‘vector’, so I looked up the term in the Glossary of Geology (2016), but there’s no geochemical definition of the word ‘vector’.

Entrenched habits

With more than two decades of precedence, it’s now a normal and accepted practice to talk about ‘vectors’ in geochemistry papers, but not to illustrate them. It’s also now normal to talk about scalar values as if they are vectors. These habits have been passed down from the most respected of senior academics. Perhaps because of this, no one questions it. This lack of illustration hasn’t seemed to bother anyone, so less-established academics (with the approval of reviewers) have repeated this process over and over again in prestigious peer-reviewed journals such as Journal of Geochemical Exploration and Ore Geology Reviews, at the same time purporting to know everything there is to know about ‘vectors to mineralisation’. I find this the strangest professional practice, and I wonder how geochemists operate in the real world when they’re asked to plot their ‘vectors’ in 3D space for an exploration company.

A few issues bother me. I want to know:

• why the 3D vectors are not actually displayed in these papers.
• why ‘vectoring to mineralisation’ has entered into the modern geological lexicon, despite the fact that no-one has shown anyone what these ‘geochemical vectors’ actually look like and how significant they may be, and
• how vectors are even computed.

I don’t know about you, but I find this a worrying pattern, so I delved into the fundamentals of how a mineral deposit could be ‘vectored’ given a set of scalar values from the rock mass. Therefore this post addresses the ‘vectoring’ process that published papers haven’t bothered to discuss for two decades.

What do the researchers mean by ‘vectoring’, exactly?

What the researchers mean by a ‘vector’ actually refers to the direction of maximum change of some scalar quantity in a particular location in 3D space. A simple way to explain this concept is a scalar quantity that changes in 2D space (F(x,y)) instead of 3D (F(x,y,z)). Topography is an example we can intuitively relate to, because ground elevation is a scalar quantity that changes as a function of the X and Y coordinates (F(x,y)). An example of mountainous terrain is shown in Figure 3; the white vectors differ in length depending on the rate of change in slope of the topography. The vectors also point along the orientation of maximum gradient (down slope).

Figure 3. Height contours and gradient vectors. Image source here.

Imagine that the high-grade mineralisation that you’re seeking is in the river position (the dashed red curve in Figure 3). Geochemists point out that many primary ore elements, such as nickel and gold, only occur in the river itself. They then state that other scalar values, such as certain trace elements, are found in anomalous values further away, well beyond the river. Their assertion is that detecting these anomalous scalar values allows them to ‘vector’ into the river (the mineralisation). You can imagine that the height value is a scalar quantity that can be measured and then contoured—two such topographic contours are shown as yellow and aqua lines in Figure 3. The ‘vectors’ are exactly orthogonal to these contours at any location (eg. Figure 4), so obtaining the vector representing a scalar gradient at any point is simple—it’s orthogonal to the interpolated contour of scattered scalar values (in this case, height).

Figure 4. Vectors pointing to the maximum gradient of change in a scalar quantity as a function of 2D position (F(x,y)). The orthogonal relationship between scalar iso-value contours (red) and the gradient vectors (blue). Note that the separation of the contours depicts the rate of change in the scalar value. Image source here.

How do we ‘vector to ore’ anyway?

The vast majority of the papers that discuss ‘vectoring to ore’ deal with mineralised bodies that are interpreted to be syngenetic in origin, from volcanogenic massive sulfide (VMS) (Galley, 1996) to sedimentary exhalative (Large and McGoldrick, 1998) to komatiitic nickel sulfide (Brand, 1999). That is, the ductile deformations seen in all of these deposits are interpreted as post-dating the mineralisation. Thus the ‘vectoring’ process must take into account the grade patterns expected in a syngenetic deposit, but also consider how ductile overprinting would affect the original primary ‘vectors to ore’.

A simple model to explain ‘vectoring to ore’

Consider a spherical deposit (ie. deposit with point symmetry), with a cross-section taken through its centre (Figure 5). The grade contours, in section, are perfectly concentric and circular about the highest grade central core. Therefore, the gradient vectors, derived from the spatial distribution of grade, are perfectly radial and pointing inward to the centre.

Figure 5. An idealised cross-section of a spherical body of mineralisation, with the highest grade in the centre. This is the starting condition prior to plane-strain shear deformation.

This deposit is then affected by a wide zone of ductile shearing that dips to the left at 45°. A picture of the body after shearing is illustrated in Figure 6.

Figure 6. Strain state after 60° of shearing is applied to the mineralisation shown in Figure 5, with the ductile shear zone boundary dipping 45°to the left.

The original vectors, which were orthogonal to the grade contours of the pre-deformational mineralisation (Figure 5) are reorientated and are now no longer orthogonal to the new grade contours. The original vectors rotated to become closer in orientation to the maximum stretch direction imposed by the shear deformation. You may remember from Figure 4 and Figure 5 that the gradient vectors are a derivative of the grade distribution and are not identified by physical markers in the rocks, so we cannot actually see these rotated vectors displayed in Figure 6. In addition, the pre-deformational vectors are no longer orthogonal to the new grade contours so the rotated original vectors cannot be derived from the new grade distribution (the red contours in Figure 6).

In other words, the original vectors that pointed to the high-grade zone are destroyed in the ductile deformation process

However, a new set of vector trajectories can be defined by computing the vectors that are orthogonal to the new grade contours (Figure 7). In fact, a new arrangement of vectors to the high-grade zone is formed continuously throughout the progressive deformation process because the original grade distribution is continuously being distorted. A new set of vectors can be computed at any stage of progressive ductile deformation, so what you see at the end (Figure 7) is simply the last frame of a cumulative result of a continuous deformation process.

The reorientated original vectors (Figure 6) and the new set of vectors from new grade contours (Figure 7) can be computed using matrix algebra at any stage of the theoretical deformation process; these equations are appended at the end of this post for those who are interested. Many geologists, including me, are not mathematicians, so a simple and intuitive way to discover what’s possible with homogeneous transformation is to play around with a graphics program such as Inkscape or Adobe Illustrator, which I used to draw Figure 5 to Figure 10.

Figure 7. Grade gradient vectors are redrawn orthogonal to the grade contours after deformation.

How can we apply this concept to a VMS deposit?

You may think that real deposits do not look like the pattern illustrated in Figure 5. That’s a fair point, but all the possible vector orientations are represented in Figure 5, so this means that mineralisation geometries of any shape can be distorted in 3D and we’ll reach the same conclusions we reached with Figure 7. The drawings in these examples are plane strain (ie. there is no strain change parallel to axis in and out of the page) to allow the concept to be simply illustrated in 2D. They apply to any general 3D distortion—vectors to mineralisation gets reset with progressive ductile deformation.

You can apply these simple principles to any 3D deposit other than an idealised spherical shape. Below, I’ll show you idealised pre-deformational VMS geometry, which, according to many research papers (eg. Shanks and Thurston, 2012, and paper quoted therein), has a vertical axial symmetry with a T-shaped cross-section (Figure 8).

Figure 8. Cross-section of an ‘idealised mound-type’ VMS deposit prior to ductile deformation within a broad ductile shear zone. Gradient vectors that point from the low-grade (white) to the high-grade zone (red) are drawn orthogonal to the grade contours. Hanging wall and foot wall lithologies are shown in orange and blue respectively.

After homogenous shearing strain is applied to the initial model (Figure 8), the resulting geometry is shown in Figure 9.

Figure 9. Strain state after 70° of shearing is applied to the VMS mineralisation shown in Figure 8. Note that the pre-deformational vertical is no longer orthogonal to the bedding after deformation.

In Figure 9, pervasive foliation, parallel to the long axis of the strain ellipsoid, is developed and the traces are shown in yellow. The original vectors pointing to the high-grade zone that were orthogonal to the grade contours in Figure 8 are no longer orthogonal to the newly orientated grade contours. These paleo-vectors to ore cannot be determined from this deposit in Figure 9 because the ductile strain has effectively reset the grade distribution and therefore destroyed the ability to derive the original vectors from it. Note that the paleo-vertical arrow is no longer orthogonal to the bedding (Figure 9)—an effect of ductile deformation that is commonly ignored in the economic geology research papers where geologists document stratigraphic columns orthogonal to the bedding of deformed rocks, claiming that this is ‘stratigraphy’. If pervasively strained and if the foliation is non-orthogonal to the bedding, the units in the stratigraphic column could not have been vertically adjacent to each other prior to deformation.

The new gradient vectors constructed from the new grade contours are shown in Figure 10.

Figure 10. New set of vectors drawn from the deformed grade contours.

The redrawn vectors show a distinct and simple relationship with the foliation—most of the vectors are orthogonal to the foliation trace. This is because any pre-deformational mineralisation will take on the elongation and flattening characteristics of the strain ellipsoid with increasing progressive strain.

Discussion

The salient points of these findings are:

• If the deposit has not undergone deformation, then the vectors to mineralisation are simply the gradient vector of the interpolated scalar value (Figure 5 and Figure 8).
• If a deposit is overprinted by ductile deformation, then the original ‘vector to mineralisation’ is destroyed by the deformation, and cannot be derived from the new distribution of scalar values (Figure 6 and Figure 9).
• The vector to mineralisation can be determined after deformation, and the vectors are derived from the new grade distribution (Figure 7 and Figure 10), but these vectors are not the same as the original vectors.

These facts highlight an obvious flaw in many research papers that claim to ‘vector to ore’—the original vector prior to deformation cannot be derived after deformation. I would highly doubt any claims to the contrary.

What is a very useful pattern is a much simpler fact—the main pattern of gradient vectors after deformation points orthogonal to the developed foliation, and although not entirely evident from Figure 7 and Figure 10, the vectors in 3D also converge to become parallel to the maximum elongation axis of the strain ellipsoid.

The only way a geochemist can claim that they can ‘vector to ore’ in deformed deposits is for them to know and understand the strain patterns of the largely barren host rocks that surround the ore body

This effectively allows the geologist to know the strain geometry in 3D—generally, most ‘vectors’ with increasing grade would point orthogonal to foliation, and then follow parallel to the long axis of the strain ellipsoid at highest grades.

A synthetic 3D example

I am going to show real cases of ‘vectoring’ in 3D using real drilling datasets in future posts. The key to doing this is to come up with a realistic 3D geological model. This cannot be simply achieved by just letting the software interpolate the data for you isotropically. Despite the lack of examples of ‘vectoring to ore’ by academics over the last two decades, the plotting of the vectors in 3D is a trivial process (Figure 11) once the geological model is constructed. In contrast, the geological model construction process itself requires structural geological experience, and is not a trivial process.

Figure 11. A set of synthetic 3D numerical data points (A) and the related contours and vector field (B). Vector convergence seen is similar to that shown in the 2D illustration of Figure 7.

Conclusions

Research papers published in the last 20 years that mention ‘vectoring to ore’ all neglect to illustrate the actual ‘vectoring’ process, or even show a single vector in 3D. Their true value is not in the ‘vectoring’ process, but how these papers define the scalar values that may help detect an alteration envelope far more extensive than the economic deposit.

To truly ‘vector to mineralisation’ requires one of two possible conditions:

1. the deposit has not undergone deformation, and the geometry of mineralisation is understood, or
2. the deposit has undergone deformation, but the strain patterns are documented and well understood.

Many geochemical papers that discuss ‘vectoring’ do not conform to either condition. What is characteristic of many research papers relating to syngenetic deposits is the almost complete lack of structural data and ignorance of structural overprinting. Therefore it isn’t possible to understand the grade patterns in terms of strain, which leads to the conclusion that the pattern of vectors also cannot be computed.

So is ‘vectoring’ to mineralisation a Claytons Jargon?

In my opinion, it certainly is—it’s a meaningless weasel word commonly used by those who have little idea of how important knowing the state of structural deformation really is to the exploration process.

The process of actually documenting the ‘vectors to mineralisation’ process is simply not researched by geochemists, so next time you hear anyone say that they can do ‘vector to mineralisation’, you might want to ask them to outline the explicit steps involved in this process of ‘vectoring’.

If geochemists can come up with the actual scalar quantities that can be obtained by analysis, then a structural geologist with a strong background in strain theory would be far more effective at ‘vectoring to mineralisation’. My recommendation would be for collaboration between geochemists and competent structural geologists.

Appendix

Pre-deformation vector v (written in matrix notation) is transformed by the deformation matrix D to result in the new orientation of vector v’, and is determined by the following equation:

However, the orientation of vector v′ can only be determined theoretically because the grade redistribution has occurred as a result of ductile deformation. Vector v′ is no longer orthogonal to the grade contour after deformation.

The new gradient vector orientation can be determined from the reorientated grade contour surfaces after deformation. This is computed from the pole vector p of an initial tangent to the contour surface at any point. The new reorientated pole vector p′ after deformation is determined by the following equation:

Since gradient vectors are orthogonal to the grade contours, p′ is effectively the new vector orientation after deformation. Note that v = p before deformation, but v' ≠ p' after deformation because of the different transformation equations that apply. These equations and geologically permissible varieties of deformation matrix D are discussed by Flinn (1979).

Acknowledgements

Rene Sterk and Bede Morrissey of RSC MME Ltd are thanked for searching the ‘vectoring’ terminology in public technical reports using their brilliant RSC Resource Intelligence Database.

---

Jun Cowan, PhD, is a director and principal structural geologist of consulting firm Orefind, and the conceptual founder of Leapfrog geological modelling software. He holds an Adjunct Senior Research Fellow position at the School of Geosciences, Monash University.

Based in Fremantle, Western Australia, Orefind is a geological consulting company founded by structural geologists Brett Davis and Jun Cowan. Visit www.orefind.com for more information. This post, and many more like this, can be found on the Orefind website [www.orefind.com]. Constructive feedback is always gladly appreciated.

References

Brand, N.W.,1999, Element ratios in nickel sulphide exploration: vectoring towards ore environments. Journal of Geochemical Exploration 67, 145–165.

Flinn, D. 1979, The deformation matrix and the deformation ellipsoid. Journal of Structural Geology 1, 299-307.

Galley, A.G., 1995. Target vectoring using lithogeochemistry: Applications to the exploration for volcanic-hosted massive sulphide deposits. CIM Bull. May 1995, 15–27.

Glossary of Geology, 2016. American Geosciences Institute. Fifth Edition.

Hosseini-Dinani, H. and Aftabi, A. 2016, Vertical lithogeochemical halos and zoning vectors at Goushfil Zn–Pb deposit, Irankuh district, southwestern Isfahan, Iran: Implications for concealed ore exploration and genetic models. Ore Geology Reviews 72, 1004–102.

Large, R.R. and McGoldrick, P.J.,1998, Lithogeochemical halos and geochemical vectors to stratiform sediment hosted Zn–Pb–Ag deposits, 1. Lady Loretta Deposit, Queensland. Journal of Geochemical Exploration 63, 37–56.

RSC MME, 2016, Technical reports search of term ‘vectoring’ of RSC Resource Intelligence database conducted 26 June, 2016 for Orefind Pty Ltd.

Shanks, W.C. P., III, and Thurston, R., eds., 2012, Volcanogenic massive sulfide occurrence model: U.S. Geological Survey Scientific Investigations Report 2010–5070–C, 345 p.

Wikipedia contributors, 2016. Weasel word, Wikipedia, The Free Encyclopedia. Available from: <https://en.wikipedia.org/wiki/Weasel_word> [Accessed: 28 June 2016]