Salinity conference

Mapping salinity in the near-surface and upland landscapes
Dr Colin Pain

Colin Pain graduated MA from the University of Auckland, and PhD from the Australian National University. He has worked as an academic, consultant, and government employee on geomorphology, soils and regolith in Papua New Guinea, New Zealand, the Philippines, USA, Indonesia and Australia. He has published more than 100 papers in international journals, and is co-author of 2 books, one on Regolith, Soils and Landforms, and the other on The Origin of Mountains. He is a member of the International Society of Soil Science, and serves on the Editorial Board of the Australian Journal of Earth Sciences, and Geology. He is a member of the Working Group on Land Resource Assessment (reporting to COAG), and was recently appointed to the international Working Group "Geoscience for Land Use and Sustainable Development" (IUGS). He currently holds a senior position at Geoscience Australia, is an Adjunct Professor at the University of Canberra, and is Leader of Program 3, Environmental Applications of Regolith Geoscience, in CRCLEME.

What I intend to do is to have a look at the mapping techniques in the context of what we actually find in an upland landscape. I am going to set some of the geomorphic context and show how complex that can be; I am going to look at some of the techniques, including radiometrics, and then finally present some take-home messages.

Click on image for a larger version of figure 1

We have to remember that the upland regolith and the landscape is three-dimensional. The upper part of this diagram shows quite clearly that you can have different sorts of shapes of landscapes – the same goes for underground, of course – and basically these different-shaped landscapes will give you different flow paths and, indeed, different sorts of regolith materials.

Click on image for a larger version of figure 2

Within the soil you have boundaries between different soil types or different soil horizons. These soil horizons have different hydraulic properties. You get hydraulic discontinuities within the soil.

Click on image for a larger version of figure 3

The same thing goes for the regolith materials: there are hydraulic discontinuities there. If we look at it in plan, we find it is not simply a matter of a vertical sequencing of these things; in plan, water with any associated salt it might have tends to concentrate. And if you look at this one, you can see what we are looking at is a concentration of water as we go from the left-hand side across to the right, to a pipe and then to a gully.

Click on image for a larger version of figure 4

If we look at it in three dimensions, the whole system becomes extremely complex. (This diagram really is very simplistic.) The interesting point about it, though, is that the subsurface flow and pathways are integrated with the surface flow. This becomes very important when we are starting to look at the movement of water and whatever the water might contain, both across the top of the landscape and through the materials within it.

As I said, this is simplified, so you might imagine that where it says we have the water table, groundwater flow and so on, if you then superimpose, on top of that, fault lines, dykes, any other structural features within the geology, jointing and so on within the regolith and the soil, any hydraulic discontinuities within that are also going to impact on the movement of water through the system.

In turn, the movement of water and the places where water accumulates are going to impact on the nature of weathering of the underlying rock, so that in general where you have water accumulating and remaining for longer periods of time you are going to get more intense weathering and therefore more likely changes in the chemistry of the regolith, changes in the depth. The simplistic diagram which shows a weathering surface coincident with or parallel to the ground surface is quite misleading. In many cases the shape or the amplitude of the surface between the unweathered rock and the regolith can be two or three times the amplitude of the surface topography. So we have to bear that in mind as well. It is a very, very complex set of systems.

Click on image for a larger version of figure 5

We can then look at the kinds of techniques that we can use to come to grips with this. On the left-hand side there in the colour is a simple aeromag image of part of the Lachlan Fold Belt. What that illustrates is the complexity of the geological system – Greg Street mentioned this this morning, and I would underline that: it is a complex system in terms of its faults and its dykes and its jointing, there are various domains, some of which are reasonably quiet magnetically, others of which are very noisy. This sort of information can be used to subdivide your landscape into what we might call magnetic domains but they also are reflected in the geomorphology and in the nature of the regolith materials. So this is one way of getting at a kind of subdivision of the landscape which is of value when we are looking at techniques for mapping the actual location and the sort of processes that saline water might go through.

On the right-hand side is an image of so-called magnetic palaeo-drainage. In point of fact, much of what we can see on this image is that the finger-tip channels on this image turn out to be channels at the surface, with maghemite in them, and as you go to the north, in this case, they disappear in under an alluvial cover. This gives us that connection that I was making between the surface phenomenon as we can observe it on, for example, the digital elevation model and the sorts of processes that are going on there, and the subsurface complexity in the regolith as it extends out and as the system becomes larger.

Click on image for a larger version of figure 6

To turn to radiometrics, which again Bruce has covered in some detail: here I would make the point that radiometrics are actually giving us a distinction between geochemical domains but, as John Wilford's work shows – and I will go through it very briefly in a minute – you can actually begin in the radiometrics to make some geomorphic conclusions about thickness and nature of regolith materials. And if you drape it over a digital elevation model, which we have done here, you can make a clear distinction between this area up here, which is a low relief even though it is at quite high altitude, comparatively speaking, and the higher relief and steeper slopes down in here. These are important geomorphic features of such a landscape.

Click on image for a larger version of figure 7

I have done a terrible thing here: I have classified a radiometric image. The point I would make here is that if you confine your classification to something which is constrained by geology, by the materials, then you do get some useful information out of it. In this particular case, the black sand ridge over here can be subdivided into a number of different classes – and these are real things, we have actually been out and had a look at them. This is an unsupervised classification, it does not require any great expertise. What we are now able to say is that there is a catenary relationship between the very sandy soils at the top of that low ridge and, as you head out towards the lower parts of the landscape, the soil texture changes. This may be important information when you are trying to map salinity in such an area.

 

 

 

Click on images for larger versions

There is also a relationship, at least in some landscapes – this is John Wilford's work – between the slope of the land and the amount of potassium in the upper part of the soil, which is where the radiometrics are measuring. This is a relationship that can then be used to classify different parts of the landscape, separating the bedrock responses from soil and regolith responses. Generally speaking, the kind of conclusion you can come to with this sort of information is that the greater the potassium signal, the more likely it is that you are dealing with a bedrock signature and that you are then looking at very thin regolith. If this potassium signature is low, then you are probably dealing with deeper regolith materials and these materials are more likely to be important salt stores.

Click on image for a larger version of figure 10

You come up with residual images of this nature, so that thin soil and regolith show up as red, thick soil and regolith in blue.

Click on image for a larger version of figure 11

This is a wetness index derived from the digital elevation model and basically predicting areas that are low in their local vicinity, but also low relative to the whole study area – in other words, areas where we are likely to get water accumulation.

Click on image for a larger version of figure 12

Then you can combine that with the residual potassium image from the radiometrics to give a map of potential saline sites, which are shown there in the discharge sites in pink, and then the actual yellow is mapped salt scalds. So this is not a direct method of mapping salinity; we are using geomorphic knowledge together with a radiometric technique, in this case, to make predictions about where we might find salinity stores within this sort of system.

Click on image for a larger version of figure 13

There are other methods: on-ground systems – this is again from John Wilford, a resistivity line. It is displayed as conductivity, but the redder areas there are areas of higher conductivity and indeed John has been able to show that these are related to high salt contents. It is possible with a lot of work to do a number of lines and then generate a 3D image from this kind of technique.

Click on image for a larger version of figure 14

EM31 is another very useful way of finding out which bits of your landscape are conductive and which bits are not. And again this can be tied in to your geomorphic knowledge of that area.

Click on image for a larger version of figure 15

When you combine these sorts of techniques with digital elevation models, you come up with extremely powerful tools in terms of the distribution of your saline materials – or high conductivity materials, in this case, and then the field checking to see whether they are indeed saline – this connection between the landscape and the geomorphology and the regolith and the saline part of that landscape.

In terms of summary, then: the upland environment is complex. It is related very strongly to hill slope geomorphology. We need to know about geomorphic processes; they operate at a more detailed scale in upland areas, often, than they do in other environments. Salt stores and flow pathways reflect this complexity and scale – scale, scale, scale, extremely important – and each area is different. We may be able to say, 'Okay, we can extrapolate information from one catchment to the adjacent one,' but you may not be able to. You have got to go back and check your geomorphic and your geological environments before you make these sorts of steps in terms of extrapolation.

In terms of the mapping techniques, there are a number of highly suitable and cost effective geophysical techniques to assist with the geomorphic and the regolith work. I have just a couple of conclusions there, that aerial techniques are better suited to catchment than subcatchment scale, and that ground techniques are more suitable for farm and paddock scale, with support from regolith and geomorphic surveys.

I would make, in conclusion, the very strong point that the data scale has to be equal to or better than the scale at which the processes within your landscape are operating. If it is not, then essentially you are
in trouble.