![[Go to Home page]](http://www.science.org.au/images/smallseg.gif){kind=small} |
Australian Academy of Science |
| About the Academy |
Home > Events > Past conferences and workshops > Fenner Conference on the Environment, 2007
Water, population and Australia's urban future
The knowledge-base on Australia’s climate and water: How good is it, what do we know, and what do we need to know?
What I want to talk about today refers to a report which is very close to completion, Climate change and Australian water resources: First risk assessment and gap analysis. It has been put together for the Australian Greenhouse Office and the National Water Commission by CSIRO, and I will be showing you some results.
What drives change? Is it knowledge? What we have here is a matrix of some things we know we know (visible assets), and things we know we don’t know, down to those we don’t know that we don’t know. And we can think about Rumsfeldian uncertainty, where we are absolutely sure we know but we really don’t know! We are generating knowledge all the time. We have different qualities of knowledge, different standards of knowledge. But do we actually respond to knowledge, or do we respond to stress? What gets things going? As we look at the whole water issue, we have been trying to generate knowledge on water resources and climate change for some years, yet to be honest it has got very little traction. But we get into a stressful situation and all of a sudden see a lot of action.
So what did we think we knew? Well, a decade ago, when we were trying to get some traction for this kind of research, we got the following reactions from people. They said we are climate proof. And Australia has been very successful – we have the highest per capita water storage of any nation in the world. But there was also a perception about how climate change science was proceeding. People had the idea, because the first output the models produced was a gradual trend in global climate, that climate change was an incremental, regular change to the average climate – as was produced from the climate models (at that time), because you couldn’t trust simulations of variability produced by the models. And climate variability was seen as something different to climate change. Climate variability is what we had always dealt with. So we got the response, ‘We can cope with climate variability because we can cope with droughts and floods, particularly with ENSO.’ That was certainly true, but then I think that some climate variability was perceived to be all climate variability. The other response was, ‘If we deal with such a large range of climate variability, a small average change won’t affect us much either.’ There was also the perception that rainfall changes from the climate models were too uncertain, and that it is possible to plan for the future based on the extremes of the historical climate (eg, the worst drought on record). All of those beliefs were very strongly held. And it was due to that second last belief, that the rainfall changes from climate models are too uncertain to essay the likely direction of change, that we actually came up with methods for assessing all of the changes out of models – probabilistic methods which were pioneering at the time.
A number of things were overlooked. If we look to past climate, the pre-instrumental climate and palaeoclimatic proxies, we find that Australia’s natural climate variability is much larger than contained in the instrumental record. So it might be fairly low probability, but there is a chance we can have a very large change, particularly in rainfall, which is not registered in the historical record. Coping with climate change requires a whole of climate approach, so it is not fruitful to separate average change and climate variability. There might be reasons for doing it in some circumstances, if you are trying to attribute a change and you want to know how long that might continue, but a whole of climate approach – what is happening in the air and on the ground – is extremely important. Because we could cope with some climate variability, it was interpreted that we could cope with all climate variability. John Williams talked about a downward shift we have just experienced, which looked very much like the one that occurred in 1895. So a sustained shift downwards with the current systems that we have, and a couple of droughts on top of that, shows that they produce a large amount of stress. So no, we couldn’t really cope with all variability. We often used the wrong statistical tools for analysis. We analysed time series with step changes in them with linear statistical models, and then said the change was statistically insignificant, when we were actually looking at the wrong sort of change. So looking at that aspect of change is actually very important. And we didn’t build an adequate scientific capacity to plan for a future that went beyond the historical limits of climate.
Due to time limits I can only show a very small part of what is in this much longer document which is about to be released, but one of the things that we did was to get together all of the information that we could, to produce a very cursory or preliminary assessment of catchment water yield across Australia. The information that we put into that was the development status of surface and groundwater resources, recent rainfall trends and population trends, and projected changes in runoff to 2030. (I will show you a little bit more of those projected changes later.)
The surface water development status was taken out of the National Land and Water Resources Audit. It is somewhat problematic. You can see from this figure that the status of a catchment will change when it crosses a state boundary. We have got this problem where they were self-assessed according to different criteria, and it was very preliminary. That said, another set has just been run. But there is certainly a need for a standardised set of methods, as objective and transparent as possible, to calculate the status of surface and groundwater development.
We have the same sort of thing for groundwater development. These were all scored in quintiles, with a score of 1, 2, 3, 4 or 5.
We included population trends for 2000–2005. Ben Preston, who did a lot of this work, has looked at the continuing projections at 2030 and finds them to be reasonably consistent, but obviously we would want to look further into the future, if possible.
On the rainfall trends, don’t think that the deep blue scores here are increases. They are just intensity, at class 5, so that the deep blue scores are where the decreases were largest. Again, this being for 1950–2004, that is a very crude estimate and can be improved upon.
Runoff projections to 2030 were calculated using a very simple hydrological sensitivity model which was developed and tuned from a number of hydrological model assessments – step changes, sensitivity assessments – around Australia, and we developed a quick-response model for that. It used the output from about 10 different climate models, so again it is the average change from a range of climate models, with a score of 5 indicating the largest decreases and 1 representing little change or perhaps small increases.
We summed the five sets of scores and again divided that into quintiles. Again it is very indicative, but it assesses the catchments across Australia from very low, moderate and high to very high. We did play around a bit with the sensitivity to see how sensitive the risk score was to the underlying assumptions, but it produces the kinds of results which we would expect to see from the stresses we are currently experiencing and from what we might experience in the future. That is, we have got great sensitivities in south-east Queensland, in the Murray–Darling Basin, and in south-west Western Australia. Northern Australia is not so affected.
Let’s look at the results that we got from the simple hydrological model, just for the direction of mean change of runoff. The light changes are where the models are in fairly close agreement and the dark changes are where the models produce a very large response. It is also sensitive to aridity, so you will get a larger response in arid areas. And again we are getting less of a range around the wetter and coastal areas, and a large potential range inland, which is where most of the increases in runoff were registered.
This shows the median projected runoff from those models, where the biggest decreases are weighted for population. Most of the decreases are projected to occur where people are situated.
We had a look at what was happening in the various cities. The details are not too important here but, again consistent with current experience, the risk score for catchments adjacent to the major cities ranged from very high for Brisbane to high and moderate through most of the south-east and high for Perth. But I think what is important there is shown by those pies, which show the different factors at play. As pointed out by other researchers, a different mix of factors affects each of the urban centres. So again when we are looking at one particular system and its own needs, there isn’t a one size fits all assessment which will match those.
One of the things that we can look at is the current stress and potential future changes. (This is repeating some of the information a little bit.) The average level of storages for particular risk categories in August 2006 are assessed – and we have had a summer without much rain since then, so a lot of those will have receded further. In August the storages in the very high risk areas had an average capacity of about 40 per cent; it is probably closer to 30 per cent by now. And it was a little bit higher in the catchments of high risk. So we can see that where we are experiencing risk currently – and some 70 per cent of Australians are currently under Stage 3 or greater water restrictions – and storages where we expect stresses to occur in the future are pretty much in the same regions.
We can now look the status of wetlands mapped all round Australia and collated with the catchments of various risk. We have quite a few catchments in low or very low risk areas that have important wetlands in them, but we also have wetlands, especially international (by area) wetlands, that are in the very high risk catchments. One of the things we can do fairly simply, but which hasn’t yet been done, is to test the risk status of different systems, to look at some of the resources that are dependent on those systems and do some simple ranking. That will help us to find out where the spots of greatest stress are likely to be.
So what do we know? This is a very quick collation for a number of different subjects, and I don’t want to go into detail. These are certainly subjective but are based on the evidenced we gathered. Our knowledge of surface water resources is reasonably good and improving, particularly regarding a lot of the hydrological processes. Knowledge of our groundwater resources is fairly poor. There is a lot of illegal extraction which is not logged, and little knowledge on how those resources may change under climate change. We have fragmented monitoring on water quality, with not too many records. The current snapshots of river condition are okay, but our knowledge of the processes under climate change are again quite poor, partly because water quality is affected by almost anything you can do in a catchment. Our knowledge of land-use change is largely quantitative and to do it properly requires sophisticated integrated models. That includes all sorts of land uses; questions about fires and salinity come into that as well. Water accounting has already been spoken about. We have got a national water budget which is undergoing continuing development. The Australian Bureau of Statistics is improving its water use accounting, and as Barney Foran talked about earlier, the use of embodied water is very important and knowledge about that is developing. On water economics we have got some frameworks, some trading which is going on at the moment. I don’t think the practice and all of its ramifications are that well understood yet. And most water is not fully costed. That is not necessarily a monetary cost; it is also environmental and other values. When it comes to social acceptance, we have an emerging awareness and some willingness to accept new ways of using water. So there is an opportunity to put that into action. As for the urban and industrial form, it is largely business as usual, with marginal changes.
We also have a conflict, I think, between science and decision-making. This has already been reflected in some comments today. But one of the things that science aims to do is to discover what we don’t know, and in doing that you pull a thing apart, you have a look at it, you examine it. If you take something that we didn’t know we didn’t know, and you turn it into something we know we don’t know, you actually increase the uncertainty that we know about. It makes uncertainty more visible. So, in a lot of ways, science makes things more visible and more complex and more difficult to understand. But decision-making, as was referred to by the previous speaker, Kim Russell, requires information in the simplest form possible to make decisions without degrading its purpose. It is about what we know we know. And that is relevant to scenarios, methods and the use of language. So again I think, in terms of where we are doing the research and doing the science, we actually need to understand that process between the generation of new knowledge and its application a lot more explicitly than we have done in the past.
This is included in much more detail in the report. A lot of the sorts of things that are required are very obvious. Baseline information: in climate change everyone talks about baseline climate data, but we actually need baseline information about everything – the condition of the resource, its use, socioeconomic information, water accounts and so on. Climate change projections: the Fourth Assessment Report is about to come out, a new crop of models that Graeme Pearman talked about and that need to be put into databases for easy access; and later this year CSIRO will be producing projections with probabilities in them. So again, hopefully, that will tie down a lot of our uncertainties, particularly about the mean resource change if not the extremes. Prediction of hydrologic responses to climate change: we need a set of models from simple, quick application, rapid assessment right through to the more complex models that can do land-use change. The impacts of climate change on managed water resource systems: again that is an integrated assessment. A few have been done for particular systems – Melbourne Water is one of the most notable of those. It is extremely useful, but one aside about that Melbourne Water project is that they put a certain amount of money in a budget to do the research; what they didn’t realise was how much effort they would expend, as an institution, conducting the assessment with the researchers. They probably spent at least that much again. But the end of the process they felt they had done a huge amount of institutional learning and could move forward from that. The impacts of climate change on aquatic systems: this is probably our most vulnerable area. The whole issue of adaptation strategies and barriers to adaptation: this is the elephant in the corner, if you like. It is an area which really has hardly been looked at, but which reflects back into the discussions we have had already today on sustainability and population. So again we need to engage in that whole understanding of social process, how we introduce these ideas into systems and we can become more sustainable in the way we operate water.
So we come back to that question: is (or should) the sustainable management of water systems, of water and population, be knowledge driven or stress driven? I would think that the sustainable use of water is not likely to be achieved through undergoing stress every now and then and engaging in a blame game with 20/20 hindsight. A lot of what I have read and seen in recent months is not even using 20/20 hindsight; I think it is about 1/1 hindsight. Pretending that climate is not a driver in the current situation, or that the blame is all due to management or it is all this or it is all that, tends to obscure the issue. Our largest knowledge need is not actually the science itself. We need to learn the ability to apply knowledge: how we can generate, develop and use knowledge in advance to achieve sustainability. I don’t think we are going to get it through responding to stress; I think we are going to get it through learning – learning by simulation, learning by observation and learning by doing.
Discussion Question: If I take Melbourne as an example, there is not only the effect of the decrease or change in rainfall but the forests on that catchment are regenerated by fires. They have incredible hydrological consequences of regenerating old forest to young forest, which uses a good deal more water and can do things like halving the flow. Presumably climate change, increased temperature, more extreme weather are likely to increase the frequency of bushfires. Did you factor that into your analysis? Roger Jones: Not quantitatively. We did talk about the possibility of what happens if another ’39-type fire gets into Melbourne’s catchments. When we were trying to set the baseline for that study we had data back to 1913. We set up the hydrological models, ran them all the way through, and found that the hydrological response in the catchment beyond the 1950s was different from the way it behaved in the first half of the century, and different from like catchments that we compared it with. We interpreted that as being a completely different hydrological response following the ’39 fires. So we have evidence from the past, largely circumstantial, that that is the case, and we would fully expect that in the future. I think that is one of Melbourne Water’s next questions. That modelling was done just with simple rainfall runoff models. They need a much more complex simulation system to assess those risks in the future. It is why authorities burnt 100,000 hectares east of the Thomson River catchment this summer to stop fires from burning back into it. Question: You mentioned the work done on the Melbourne Water climate change project, and the corporate benefits that came from interactive assessment. Did that inform the thinking about the way in which the institutional structure organised itself, whether it required the preservation or change of the corporate culture you were talking about? Did that enter into those things that were learnt out of that project? Roger Jones: No, it didn’t. I think most of the knowledge gained was about the institution as it is currently structured. The biggest benefits were in the different technical areas, understanding how the impacts flowed through the system and how their particular corporation, the institution itself, could look at the way they managed and dealt with knowledge. But it certainly didn’t enter into actually investigating the institution itself.
|
© Australian Academy of Science | aas@science.org.au