Salinity conference

Mapping deeper salinity in intermediate and regional groundwater flow systems
Dr Tim Munday

Tim Munday is a Senior Research Geologist with CSIRO Exploration and Mining and is currently at the CRC for Landscapes, Environment and Mineral Exploration. Educated in the UK at the University of Reading, where he received his BSc and PhD degrees, he held positions at the University of Durham, England and Curtin in Western Australia before joining CSIRO.  With over 15 years research experience in the application of remote sensing and geophysical technologies to exploration and the environment, his recent interests have concerned the role and application of geophysical and remote sensing methods in providing an improved biophysical foundation for natural resource management. He has led various government-funded research projects since 1983; and been a leader and participant of industry-funded research projects since 1992. He is a regular contributor to meetings and workshops on the relevance and effectiveness of geoscientific spatial data in mapping salinity specifically, and their use in natural resource management more generally.

I am going to take up a slightly different approach from the one that has been taken by the previous speakers and actually look at just some issues that relate directly to the report that Brian and Peter have put together. In fact, Glen [Walker] provided a pretty good lead-in because it addresses some of the issues that have been raised in the previous discussion.

In the context of the review I think we need to make explicit mention of the role of some technologies that really are designed to map subsurface elements of the landscape, and specifically materials. We tend to be talking quite extensively about techniques for mapping salt stores, salt, its distribution, and I think as Glen hinted a few moments ago some of the important issues are not necessarily linked directly to actually mapping salt per se but rather the materials in a regional context that are very relevant to understanding what salt does and how it moves in the landscape.

I think the critical thing is that if we understand how the material controls the movement of water – and salt, by definition – perhaps we have a better method or a mechanism for managing dryland salinity and particularly for protecting assets. So those are the sorts of things I would like to flag as being important.

The other thing that keeps coming up quite repeatedly, from a variety of speakers, has been matters of scale. Glen also has flagged a very important issue, and that is that whatever technique we are looking at, be it in either a local flow system or a regional flow system, a good appreciation, at least at a first level, of process is a critical driver of the technique you choose, adopt and apply.

Click on image for a larger version of figure 1

But we are talking about regional flow systems. I am going to jump to the Riverland in South Australia as a sort of example of a procedural approach that might be taken in the application of a mapping technique to understanding how we might manage salinity in that sort of setting.

We have got a regional flow system, groundwater coming from Victoria up into this region up here [an area on the southern bank of the River Murray]. This is important, an area which we have studied as part of the South Australian project where we have flown some helicopter electromagnetics with a view to understanding the subsurface materials, not salt directly. And this is a critical, important issue.

Click on image for a larger version of figure 2

The reason why we are looking at materials rather than salt in this landscape is that at the regional scale, and indeed at the intermediate scale, what the salt is doing and what the water is doing are reasonably well understood – at least, that's what we believed.

Let's take Peter Cook's model – and this comes back to the issue of understanding process – to understand how, if you clear a landscape, what happens to drainage and what are the implications in regard to an asset, in this case the River Murray being a very critical asset. So what happens? If we clear a landscape, we change the land use and land practice, do we move saline groundwater into the Murray more rapidly as a consequence of rising water tables?

Click on image for a larger version of figure 3

These are some of the drivers that influence the way we look at mapping technologies to manage the problem, in this particular case the damage to an asset, a critical asset in the context of Australia. Our target in this case was: can we understand materials – the variability of those materials – in that landscape to perhaps slow down drainage into the groundwater, prevent that rise and so on and so forth?

Click on image for a larger version of figure 4

The way we tend to take a mapping technology forward is to take a very considered approach, and again I flag that this has come up in the context of submissions by LEME and others. We have to have a very considered, staged approach to the way we look at the application of a technology in addressing a problem; that is a critical issue as well and I flag it here. We took an understanding of the materials, their petrophysical properties, tried to determine whether it was detectable by a mapping system – in this particular case, an electromagnetic system – which system might be best, and then how much it would cost to actually do it. And in the context of this study we were trying to potentially reduce the effects of salinity on a very significant asset, and that is perhaps where it is most economic.

Click on image for a larger version of figure 5

From that process we determined a system and we chose a helicopter, this thing here [on the slide] which is flying, a bird. We didn't fly the whole area. The question we then asked was whether it would map this near-surface clay layer, because that is the critical thing in reducing recharge to a conductive groundwater table. We want to reduce recharge or drainage down to there [on slide] so that this [groundwater] doesn't rise up and flow into the Murray or wherever. So these are the drivers of the way we went.

Click on image for a larger version of figure 6

But at the end of the day, through some interesting procedures that Ross Brodie and Andy Green put together, we came up with a map like this, which is a map of clay thickness. So this is the output that we have got from our mapping system. This is where we stop. We hand this product over to Peter Cook, who takes this process understanding. He knows, if he has a certain thickness of clay in a locality wherever, with a particular landscape unit, how rapidly drainage will occur at that particular point in the landscape. He can then determine, with an understanding of how deep the groundwater is, how quickly recharge will occur to that. And we can do predictive modelling, risk modelling, in that sort of landscape. Peter Cook and Fred and others have produced predictive maps showing effective recharge over time, using this as an input.

Click on image for a larger version of figure 7

But it is a map of materials, not a map of salinity, compared with what we were trying to work with, or what Welby(?) and DWLBC were trying to work with, which is a map of the same material from borehole data, but perhaps not at a resolution that allows an understanding of local landscape effects, local land use changes on recharge particularly. So if you look at the product from the helicopter, compared with what was existing, I think you can get a feel that the helicopter was perhaps doing a good job in improving our resolution.

 

Figure 8	Figure 9	Figure 10
Click on images for larger versions

This brings me on, very quickly, to an issue of scale. We have got a regional scale issue there. We are trying to reduce recharge at a regional scale to a flow system, if you like, that is operating at large scales. This is a data rich area. We have got a lot of data about where the water table is, how salty it is, a lot of boreholes, and so on and so forth. We have got a good understanding of process. What we have thought is, 'Well, we've got enough. We can work and produce something very useful.' One of the interesting things that came out of this study, as a residual, is a better appreciation of subsurface materials but also subsurface characteristics of the aquifers within which the groundwater sits. This is a map of the present-day topography, dominated by dune systems running east-west. Underneath these dunes, and the clay that I just showed you a map of, sit these other dunes, which relate to the Loxton-Parilla sand sequence. These are massive barrier prograding beach strandline systems that run right through the area.

The significance of these became perhaps more significant when you started to understand the sedimentology, the facies variation and so on, as it relates to a problem that was encountered in the Bookpurnong area.

 

Figure 11	Figure 12
Click on images for larger versions

This is Lock 4 on the Murray, and what they are doing is putting in a salt interception scheme in this particular locality. Now, we are right down at scales of hundreds of metres, perhaps tens of metres as well. The question in this landscape, and a problem that arose when some original planning was done with the observation bores and production bores – and these [on the slide] are where they were located relative to Lock 4 – was this variable yield from the various boreholes: why were we getting this variable yield at this scale? The question then was: did the barrier systems that we were identifying in the helicopter data actually tell us something about facies variability which had a direct relationship to the yield characteristics or the yields that we were getting out? So things move at different scales when you start to look at the materials, particularly, not at salt per se in this particular case.

I think the critical drivers to the successful application of technologies of this nature for managing salinity in regional flow systems such as this are, firstly, having a well-defined target, as we had – and that was developed in close consultation with people on the ground, people who were dealing with the issues actually in the field, so to speak.

Also a very significant asset was underpinning the whole drive of why we might actually try and map this, so economics come in, in the sense of protecting an asset.

We have to accept that there may be a serendipitous return from flying airborne data, but it should not necessarily be the principal driver in the first instance to the application of the method.

We have got to consider the issues of scale, because you can work with one data set at one scale but it has also applications at others. In this particular example that was clearly demonstrated as well.

So we have to also accept the need for an improved biophysical basis. In some areas, regional airborne EM surveys have real application for understanding where we don't have much data. And we can fly regional stuff, broad line spacing, whatever; it is very relevant and in some regions that is very appropriate. We should also recognise that there may be information at scales where you have got data rich information, so to speak, from airborne data such as I have demonstrated.

Click on image for a larger version of figure 13

In summary, we have got to be fairly open-minded about the application of particular technologies. We have got to understand what assets we might be interested in protecting or whatever. We have got to understand the problem or the target, which is often a material and not necessarily salinity directly. We have got to link that directly to an understanding of process, in order to be predictive. We have also got to have options in the way we might manage, in this particular case in South Australia, the effect of recharge. We have got options – engineering, albeit, but they are options.

So we have got to match technology such that we can get to the right sort of product at the end of the day, and balance a perceived need from different parties – so from technologists, who say, 'Ah, we should fly this, or we should do this, because it'll look good. It'll be great, it'll be colourful and maybe we'll be able to get something from it' – but also from the developers and planners and the implementers, who should come at it and say, 'There may be return or value in this data set that we haven't recognised because we don't have the experience.' We have got to come at it from two directions.

Phil McFadden – Thanks very much, Tim. There is just one thing I would like to ask you. You highlighted in your penultimate sequence there that your helicopter data certainly had a lot more structure than what you had before. How do we know that that actually reflects what is happening in the ground?

Tim Munday – By drilling.

Phil McFadden – You haven't ground truthed it?

Tim Munday – That's right.

Phil McFadden – Do you have anything in particular for our authors?

Tim Munday – Well, that was right at the beginning: explicit mention of the role of technologies in material mapping, not necessarily thinking of it as a salt store map or things like that. That is a very critical issue in many areas, and one that we have demonstrated very clearly in several areas in South Australia and that I think has wider application.