Fudging Figures on Murray River Salinity: More Shame on CSIRO

CSIRO, Australia’s largest scientific research organisation, released a two-part report* last Friday on ‘water’ in the Murray-Darling Basin, a region often referred to as the food bowl of Australia. The icon within this region is the Murray River and salt levels in the river have long been considered an indication of the region’s health and the sustainability of Australian agriculture.

The report reiterates “salinity … as one of the most serious environmental issues in the Basin” and suggests that “stream salt loads” and “stream salinity” will increase. Part 1 of the report is 48 pages and Part 2 is 29 pages but there is only one graph of Murray River salinity and it was drawn in 1988, some 16 years ago. It is computer-model generated and interestingly begins in 1920 even though first recording were not made until 1938.

Salinity Gph CSIRO Feb06.JPG

In my opinion it is both sad and deceitful that the CSIRO won’t show us what salinity levels really look like but instead keeps republishing a dated and misleading graph from a computer model.

Would you like to see what salt levels are really like?

Here’s a plot of yearly average stream salinity from when recordings where first made in 1938:

salinity Yearly Averages.JPG

This graph is based on data that I recently requested and received from the Murray Darling Basin Commission.

A plot of all the daily readings for Morgan, a key site as its just upstream from the off-shoots for Adelaide’s water supply, also shows a downward trend for the last 20 years:

Salinity all data May06.JPG

This data was also sourced from the Murray Darling Basin Commission and the last data point represents last Friday, 19th May 2006.

It’s good news that salt levels are falling. But no-one will acknowledge it!

A common ‘excuse’ given for the low stream salinity levels is that it’s been so dry, “It doesn’t rain so much in the Murray-Darling Basin anymore”. But hang-on, a plot of rainfall for the Murray-Darling Basin shows no recent drop-off. The last very dry year was 2002 and that wasn’t unusually dry in the scheme of things.


The graph is from the Australian Bureau of Meteorology, click here.

Perhaps so much funding is dependent on the perception that salinity is a continuing problem, and so many reputations have been made on the myth, that there is almost an obligation to repeat the falsehood?

I reckon it matters that CSIRO and others keep misleading the Australian public on this issue. I reckon it matters that the federal government just announced another $500 million for the Murray River on the pretext that river salinity is a continuing problem.

* The reports are titled ‘The Shared Water Resources of the Murray-Darling Basin’ by Kirby M et al. 2006 and ‘Risks to the Shared Water Resources of the Murray-Darling Basin’ by Van Dijk, A et al 2006 published by the Murray Darling Basin Commission, Canberra and prepared by CSIRO Land and Water as part of the Water for a Healthy Country National Flagship Program.

This Sunday, Channel 9’s Sunday program is planning to feature at story on the Murray River and salinity. I’m hoping that Ross Coulthart from SUNDAY will go beyond the empty rhetoric and show that the emperor has no clothes.

So if you live in Australia, watch Channel 9 from 9am on Sunday.

I’ve written a bit about the Murray River which can be accessed online including an IPA backgrounder, click here, and something for Online Opinion more recently, click here, and for ABC Radio National’s Counterpoint with Michael Duffy, click here. I will in due course do a complete critique of the two-part CSIRO report.

25 Responses to Fudging Figures on Murray River Salinity: More Shame on CSIRO

  1. Siltstone May 24, 2006 at 11:31 pm #

    If the salinity trend at Morgan was a temperature trend we’d be sunk by claims of “global cooling”. That the CSIRO is so shoddy in it’s presentation of data is of grave concern. I look forward to your full critique.
    P.S. the Murray River Basin rainfall moving average seems to have a slight upward inclination – and check out the 1950’s wool boom years!

  2. SimonC May 25, 2006 at 11:50 am #

    “It doesn’t rain so much in the Murray-Darling Basin anymore”. I thought it was that because of water harvesting etc that not as much of the rain water finds/stays in the Murray-Darling system anymore.

    Also with your comment about the downward trend in salinity at Morgan over the last 20 years – you should also state that between 1980 and 1999 there have been 9 salt interception schemes constructed on the Murray. The amount of salt removed using these schemes is equivalent to ~200 ECs.

    So the good news that salinity is falling is due to the work of MDBC and am I sure you will acknowledge it!

  3. Boxer May 25, 2006 at 1:20 pm #


    I remember that last time we got into a squabble about salinity you terminated the discussion by saying I don’t know much about salinity. Given that’s the case, perhaps you can help me understand these graphs.

    My observations are that:

    The graphs are not inconsistent with one another but the different timescales and various degrees of smoothing confuse the issue. The Fig 18 (CSIRO) presents a simplification of the medium term trends in the past. A trend line placed through the first “salinity at Morgan” graph would approximate the gradual rise seen in Fig 18 until about 1980, followed by a dip, which I assume is a reflection of the impact of the salt interception schemes mentioned by SimonC. Given that one is a simplification of perhaps the whole river and the other is typically noisy data from a specific point (Morgan), there doesn’t seem to be any real difference between them.

    The second salinity at Morgan is every reading and the noise is so great that it doesn’t provide more information than the first Morgan graph.

    Now given that the data were sourced from the MDBC, the data are of themselves, the acknowledgement that you seek. Aren’t they? Bearing in mind that what a politician or a journalist may or may not say or acknowledge could have little to do with reality, I would rather discuss the publicly available data. Those data do acknowledge the recent decline in salinity at Morgan.

    This suggests to me that whoever designed the salt interception schemes got it about right, so causing the observed decline in salinity. Is that a reasonable possiblility?

    Now for the future as portrayed in Fig 18. I wonder if the CSIRO et al are observing water table responses in the Murray Darling Basin which are not yet impacting upon the river itself. They have applied the salt interception schemes to those water tables that have already intersected the river bed, but elsewhere in that huge complex sedimentary basin there may be other water tables that are not being tackled and are expected to intersect the river bed in the future. If this is the case, then the trends presented in all these graphs for the last 25 years may not represent the anticipated trends for the future. As you may observe, I don’t know the answer to these questions, but they would seem to be worth investigating.

    One of the few things I do know about salinity is that the response times for water tables (i.e. the time taken for a water table to intersect the surface or a river bed) may vary between a few years and several centuries, depending upon the specific local circumstances. Even at one point, there may be a short term response from shallow water table changes, and another anticipated response from deeper water tables. So perhaps a short term local success at Morgan is not a reliable predictor of the future for a very large river system.

    Have you asked the CSIRO about the basis for the modelled results? Their graph agrees with your consistent argument that recent trends for salinity have been down. If you are arguing about the veracity of their modelling, have you dissected the model(s) or discussed with them the reasons why they have such a gloomy long term outlook?

    Your fundamental concern that scientists can use their knowledge to create the “more research is necessary” argument, and so project their own research programmes into the future, is not without merit. It is normal marketing strategy, essential in the modern free market economy. The days of science for the sake of discovery alone are gone. Agreed, you should not take the word of scientists at face value, but this does not make all hydrogeology research wrong. If the modelling of the future is faulty, to demonstrate this you may need a detailed review of the models and the data upon which they are based. A discussion about the data at Morgan for the recent past may be irrelevant.

  4. Siltstone May 25, 2006 at 8:06 pm #

    The CSIRO chose to take an 18 year old graph (and by definition ignore the past 18 years of measured data) and present it to illustrate salinity at Morgan. Jennifer has taken the trouble to look at data up to 2002, not just 1988. No wonder Figure 18 bears little relation to the real world. The 1988 predictions for the future are way off beam and give no confidemce that the “modellers” were blessed with much skill. For the CSIRO to present Figure 18 in a contemporay report shows either incompetence or a deliberate intent to mislead. Maybe the CSIRO should stick to diet books or do something else.

  5. Boxer May 25, 2006 at 11:58 pm #

    I thought the 18 year old graph agreed with the results that have subsequently come to light since the 1988 prediction. So perhaps they could be accused of not producing a new graph, but if the old graph correctly predicted the past 18 years, is it it such a sin to use the old graph? The new one would presumably look much the same as the old one. By using the old graph they indirectly illustrate that their 18 year old modelling result was pretty good. Of course they used the historical data to tune their model; normal practice for any model, be it a model of catchment hydrology or turbulence in a cup of tea.

    I accept that at first glance the two graphs are different in appearance, but if you plotted the actual data (1988-2006) on the old graph’s axes, the scatter of points would land around the predicted, heavily smoothed, line modelled in the 1980s. Slow rise, then a dip starting from about 1980. It is a debatable point about where the circa 1980 down-turn starts to swing back up. But if you are arguing that the model implies the upswing should start on Tuesday 2nd Jan 2003 at 1432 hrs, then you may be asking a lot of a model. The model says that there will be an up-swing. If it’s a decade later than originally predicted, then so what? In the time scales of catchment hydrology, and particularly in a catchment as large and complex as the MDB, a decade or two is hardly here or there.

    Perhaps you are right, the trend since about the mid 80s may continue indefinitely. That will be good for all of us. But it might also be worth asking the modellers why they predict an upswing after a temporary decline. The duration of the predicted dip-and-rise may be a decade, or three decades, but that is a pretty minor point compared to the predicted rise. Will it occur or not? Why?

    Anecdote for you Siltstone. In the WA wheatbelt the boreholes in the valley floors are, in about half the cases, showing a small decline in the last few years of lower-than-average rainfall. Reason for some relief. Climate change is helping us out. However, given that the valley floor water tables are frequently at or near the soil surface, and it is not unusual to see valley water tables above ground level (that is, the deeper water tables have built up so much pressure under confining layers that they will flow continuously if not capped off), the dry period in recent times has allowed the evaporation from the valley water tables directly to the atmosphere to catch up with rainfall. Will climate change, if it is more than just a short term fluctuation, solve salinity for us?

    If you look at the bores up on the slopes, where the water tables is still several metres down from the surface, there is no evaporation directly from the water table into the atmosphere. Here, nearly all bores are still rising, a few are static.

    So if you take the data from the valley floors where the visual appearance of salinity is most apparent, you can argue that the trends are looking good. But if you think about the water tables under the slopes, where will they end up? Is the current dry period a reliable indication of rainfall for the next few centuries? Is it a little premature to celebrate? These systems are very complex and respond in different ways as you move across the landscape and down into the soil profile. I don’t understand it all, as Jen will testify, but it is likely that the hydrogeologists know something I don’t. Might be worth talking to them rather than assuming they are incompetent and implying their motives are corrupt.

  6. Jennifer May 26, 2006 at 8:53 am #

    Boxer, I think you are ‘clutching at straws’… and trying to impose Western Australian issues on a Murray Darling Basin success story and please don’t assume I don’t talk to anyone.

  7. SimonC May 26, 2006 at 9:17 am #

    As far I can tell the report says that a review conducted in 99/00 of the 1988 model confirmed the model’s validity. The model predicted drop in salinity associated with the salt interception schemes (which I’m sure that Jennifer will acknowledge). The report also says that if things continue as they salintiy levels will begin to rise again.

    Jennifier are you going to acknowledge that the drop in salinity is due to the salt interception schemes?

  8. Jennifer May 26, 2006 at 9:29 am #

    Simon C,
    I have been acknowledging that the drop in salinity is due, probably in large part to the salt interception schemes for 3 years … including in ‘Myth and the Murray’ published in December 2003.
    All you need to do Simon C, is read one or two of the links I provide at the end of the above post.
    And I suggest there is NO reason to expect that salt levels are going to deteriorate in the short, medium or longer term … as detailed in ‘Myth and the Murray’.

  9. Boxer May 26, 2006 at 10:12 am #


    My previous closing remark was more aimed at Siltstone’s closing remark. However I do wonder how open the communication channels between yourself and the CSIRO’s hydrogeologists really are. Your comments in this subject area always pivot around salinity at Morgan, and they are also often responses to policy documents. Policy documents are an ugly combination of of science and politics. The introduction of inevitable “spin” often detracts from the quality of the underlying science. (Just as science in the absence of policy can become a detailed discussion about how many angels DO fit on the head of a pin.)

    Federal policy documents also tend to regard salinity in terms of a single, albeit very large, river. By discussing such documents, we are to some degree reducing this very complex issue to a discussion about salinity at one point in a single river. It is not surprising that those of us who don’t live on that individual river see the issue as being more complex. Knowing that federal pollies read this blog, I like to point out that salinity and the Murray River are not the same subject. The two topics happen to overlap, but only a bit. The recent fed budget allocated 0.5 billion to saving the Murray. “There,” say the PM and the Treasurer, “we’ve fixed salinity”. No, you chucked a lot of money at one of very many rivers because the electorate thinks fixing one river fixes salinity in total.

    The WA anedote was to make the point that within a single paddock you can have rising and falling water tables at the same time after a series of dry years. Within something as vast as the Murray Darling Basin the situation is not likely to be any less complex than a single paddock. I wish a CSIRO hydrogeologist would come in here and describe why they think trends at Morgan are just a tiny tiny component of the Murray Darling Basin.

  10. SimonC May 26, 2006 at 11:16 am #

    “It’s good news that salt levels are falling. But no-one will acknowledge it!”

    From the first line of the section of the report you that the graph is from:

    “The 1988 Murray-Darling Basin Salinity and Drainage Strategy6 has led to a reduction of salinity in the River Murray (Figure 18).”

  11. Gummo Trotsky May 26, 2006 at 1:14 pm #

    This discussion has gone a long way on the assumption that the CSIRO graph was actually plotted 18 years ago. But was it? At the left (over the continuous line) we see the caption “recorded salinity” and at the right, after the line forks, marking (I presume) the upper, mean and lower salinity values derived from modelling, the caption “uncertainty associated with predictions”.

    So possibly we’re dealing with two data sets here – historical data before the year 2000 (approximately) and model predictions beyond that. Was that possibility checked with the report’s authors?

    It might have been wise to do so – there’s no telling what might happen if you start confecting conspiracy theories (“Perhaps so much funding is dependent on the perception that salinity is a continuing problem, and so many reputations have been made on the myth, that there is almost an obligation to repeat the falsehood?”) on the basis of a single graph in a report of 77 pages in total.

    For an example of a writer confecting a conspiracy theory on the basis of inadequate information, see http://larvatusprodeo.net/2006/05/25/fudging-facts-on-murray-river-salinity-reports/

  12. Jennifer May 26, 2006 at 1:51 pm #

    I posted a response at larvatus prodeo, which has not been published yet? Perhaps Gummo, you could advise whether they usually hold comments for very long?

    And if, Gummo, you want to understand the origin of the graph read the 1988 MDBC salinity strategy and also my ‘Myth and the Murray’, link at end of above post.
    Also I did try and speak to the authors of the reports but they said I had to clear it all with CSIRO media people first…
    and the saga continues here … http://www.jennifermarohasy.com/blog/archives/001389.html .

  13. Gummo Trotsky May 26, 2006 at 2:43 pm #


    Comments at Larvatus Prodeo have to get past the spam filter, which might refuse posts with too many links (more than 3). Beyond that it’s almost, but not quite open slather as, unfortunately, it’s been necessary to ban a few commenters for excessively abusive comments and “trolling”. Your comment has appeared at Larvatus Prodeo since I was last on-line. In the absence of any other information, I’d chalk the delay up to server response times, or something equally arcane and technical.

  14. steve munn May 26, 2006 at 5:51 pm #


    Environmental flows for the Murry River are required for a number of reasons, as you well know. This includes reducing the carp infestation; increasing native fish numbers; increasing platypus numbers and possibly allowing them to re-esablish in South Australia, where they are extinct; improving the health of the Coornong- which is badly salted to such an extent that fish have become extinct and the mix of waterbirds has changed; and restoring to health badly degraded flood plain ecological systems.

    Your posts so often focus on individual aspects where you can score an easy point or two but dodge the wider issues.

    When will we see your long term plan for the Murray River?

  15. Jennifer May 26, 2006 at 6:16 pm #

    I reckon that the Murray River has had too much water in it over recent years. I reckon stretches of it should have been allowed to dry up over the recent drought … as they would have done naturally in the past before all the dams and levies.
    I cover all of these issues and more in ‘Myth and the Murray”. I’ve directed you to the link before, but from memory, you said you prefer reading The Age and blogs … but go on, have a read, its a bit long, but that’s a consequence of me covering all the issues.

  16. Jennifer May 26, 2006 at 6:18 pm #

    PS. Letting stretches of the river dry out would probably have a really big impact on carp numbers, native fish are much better able to deal with drought and salt.

  17. rog May 26, 2006 at 7:47 pm #

    Hey Steve, why not set an example by not ‘dodging the wider issues’; just who is your employer + just how much are you paid?

    Not that I particularly care, nothing personal mind you but your fair weather comrades think it is of significance and are busily beating it up as an issue – in the past they made it a matter of life and death. Down with the bourgeois! Off to the Gulags! Off with their heads! Solidarity comrades!

    As for me, if you are successful then good on you, you deserve it. If you are not at this moment successful-well keep trying, your turn is just around the corner. Just dont waste to much energy on being unsuccessful.

  18. steve munn May 26, 2006 at 8:43 pm #


    You have me confused with someone else. I did peruse you Myth and Murray document although I must admit I didn’t read it in full.

    Re carp, I was under the impression that there was some evidence that carp numbers decrease subsequent to a flood plain inundation and an accompanying commencememnt of the life cycle of native fish.

    How well do platypuses cope with increased salinity levels and low water levels? I thought low water levels and waterhole formation would render them more suspectible to fox kills.

  19. Schiller Thurkettle May 26, 2006 at 11:24 pm #


    From the data provided by CSIRO and others, it is quite obvious that the Precautionary Principle is being applied.

    According to the Precautionary Principle, where there is scientific uncertainty about future risks, immediate steps should be taken to ameliorate them.

    It is glaringly apparent that it is uncertain whether salinity will rise to unacceptable levels, therefore we are obligated to take immediate steps to prevent the salinity crisis from occurring.

    The fact that data suggest a salinity crisis is highly unlikely merely demonstrates that scientific uncertainty about its inevitability is correspondingly high, and the urgency of the matter as well.

    This is, indeed, how the Principle would apply in this case. Indeed, it looks like a textbook application of the Principle.


  20. Siltstone May 26, 2006 at 11:24 pm #

    A lone interested individual can produce a couple of very informative graphs of salinity at Morgan over mnay decades (thanks Jennifer) and the might of the CSIRO can reproduce an 18 year old cartoon that any professional scientist worth their salt (ha ha) would be ashamed of. That is the telling point. Is the CSIRO really worth supporting any more?

  21. Jennifer May 27, 2006 at 8:44 am #

    Thanks Schiller … very insightful!

    And, thanks Siltstone.

    Steve M.,

    Are you suggesting we keep the river full of water for the playpus?

    It may in fact be a bit difficult creating perfect conditions for both Murray Cod and Platypus.

    I’d suggest a focus on Murray Cod because there are Platypus almost ‘everywhere’.

    For example, if you wander up the Yarra a bit, you’ll probably find some burrows in Heidelberg … you do live in Melbourne don’t you?

    In fact there are even playtypus in Hobart in Tasmania and then in Queensland on sugarcane farms. They aren’t that fussy in the ecological scheme of things.

  22. Jennifer May 27, 2006 at 8:54 am #

    PS Carp were recorded in similiar numbers to which species of native fish at the Torrumbarry Weir after the 2000 flood? (Steve M, you can find the answer in ‘Myth and the Murray’.)

  23. Richard Darksun May 27, 2006 at 7:01 pm #

    It seems poor science to put up concentration graphs as some sort of “proof” that things are getting better or worse, at a minimum one needs to know flow volume so one can estimate tonnes of salt and not just a concentration. One would hope that someone is measuring tonnes of salt and water into the evaporation pans as part of the salt interception schemes and doing a mass balance. The arguments on both sides of this post seem less informed that average greenhouse skeptic!!! without access to good data comentrary might as well be just junk!

    Is some one collecting the data needed and is it accessable? who is doing a salt mass balance? and tracking salinity and height in lots of bore holes?

  24. Siltstone May 27, 2006 at 10:45 pm #

    Aquatic organisms don’t respond the load (gross mass per unit time), they respond to concentration (mass per unit volume). At Morgan there has been a pronounced decline in concentration in the river since about 1980, based on the graph Jennifer presented. Salinity has halved, roughly. For any organism sensititive to salt, that’s a good news story. Unless catchment discharge at Morgan has doubled since 1980, then load has also declined. Page 19 of the report Jennifer links to says that salt loads have indeed declined. Another good news story.

  25. bad boy June 9, 2006 at 2:49 am #


    Don’t worry so much about Trotsky. Anyone who takes the moniker of a mass murderer who wants to talk salt levels accusing you of r bad things I just say ignore it.

    Munn’s another critter from the swamp that only comes calling (like Trotsky) when he thinks there’s a blood around- lots of it. Don’t worry he “graduated” from RMIT, not MIT, his alma mater has and R in front of it so nothing else needs to be said other than he will hold a chart upside down and thinks it’s right side up. He never got through remedial at that “ exulted institution” of higher learning. He disclosed his alma M on the web once.

    It’s the same old saw with hard lefties, when they can’t attack on facts they attack your integrity. They get it from Quiggin and take it round the web.

    Trotsky aren’t you the least embarrassed attacking people about their integrity when you traipse around with the moniker of a horrendous mass murderer. Seriously are you that much of a freak? Have you thought of going into the circus as a freak act or something? Anyone calling himself Trotsky wouldn’t be demonstrating concerns about the moral ethics of attacking someone at a personal level.

    Munn’s a greenie that’ll tell lies and ram people if he can get away with it.

    What a class act these two animals are.

Website by 46digital