Canberra water utility praises ineffective “Seasonal Streamflow Forecasts”

I noticed this article reporting on a BoM – CSIRO initiative to forecast river flows a few months in advance. I just hope they use better models than they employ for the hopeless BoM rainfall and temperature Outlooks.

What caught my eye was the gushing praise from ACTEW our Canberra water utility – whose business model seems to be to charge households increasingly rapacious water bills, and pay increasing money to Govt, while keeping water restrictions permanent as dams overflow and during the dryer times tipping large amounts of water away as “environmental flows”.

The catch for ACTEW is that their praise is exposed as unfounded puff because the BoM – CSIRO “Seasonal Streamflow Forecasts” web pages publish a check on the results of their forecasts in the form of a Table “Summary of skill scores”. Unfortunately for ACTEW forecasts for the two catchments relevant to ACT water supply, Cotter and Queanbeyan both rate pretty poorly. ACTEW comments at the end.

Forecast service predicts river, dam levels
Updated Tue Dec 14, 2010 2:28pm AEDT

Australian Government scientists have developed a new tool to predict how much water might flow into rivers and catchments up to three months in advance.

The Bureau of Meteorology (BOM) and the CSIRO have launched the Seasonal Streamflow Forecast Service online, which they say will provide water managers, farmers and the public with information on likely water flows into major rivers and dams.

The forecasts have been developed for 13 river sites and eight storages in the south-east Murray-Darling Basin, including the Murrumbidgee and upper Murray catchments.

It will be expanded to include other locations across Australia over the next one to two years.

BOM deputy director Dr Rob Vertessy says it is cutting edge science.

“It’s the first operational seasonal streamflow forecasting in the world, that we’re aware of, that’s accessible to the open public,” he said.

He says the forecasts will assist decision-making on water allocation, environmental flow management and storage operations.

“Water managers and users now have access to skilful, reliable and robust forecasts of seasonal streamflows,” he said.

Dr Vertessy says this year’s extreme climate variability and the recent drought-breaking floods have highlighted the need for sophisticated tools to predict stream flow.

“Although there is a lot of water in some areas of Australia at the moment, water availability will continue to vary in the future as demand continues to grow and climate inevitably varies,” he said.

“This poses major challenges for water resource management.”

Dr Vertessy says he is pleased with the accuracy of the predictions.

“It’s really tough to predict the future but basically we do three things – we look at climate models and ask them what is the rainfall prognosis for the three months ahead. We look at the recent historical rainfall over the last month or two and then we look at the current wetness conditions of the catchments,” he said.

‘Invaluable’ resource

Canberra’s water supplier ACTEW has been using the tool.

Managing director Mark Sullivan says it allows for better water security planning.

“We have understood a whole lot more about the development of flows in recent months than we ever would have without the tool,” he said.

“We have been guided and we’ve understood what’s happening in our catchments. It’s helped in decisions about water restrictions and it’s already been proved invaluable.”

14 thoughts on “Canberra water utility praises ineffective “Seasonal Streamflow Forecasts””

  1. While publication of the uncertainty is to be encouraged, I can’t follow their RMSEP at all, and can’t find a definition of it. The definition on the BOM site doesn’t help as it has no equation, and indicates like most RMSE that better predictions should have lower values (less residual between observed and expected). The summary on the BoM site has higher values for higher skill????

  2. Click on ‘Glossary’ in the box at top right of the summary page,click ‘ Bayesian Joint Probability’ then the embedded ‘research papers’ and there are four papers are listed.

    The way I read the chart,the Cotter and Queanbeyan predictions have showed moderate skill for 3/4 of the year,and lower skill in summer. I suppose the summer thunderstorm season introduces a lot of potential local variability to rainfall ,thus streamflow ditto. Both those catchments also have rainshadow effects at play,Cotter having the Brindabella Range as a barrier from the west,and the Queenbeyan being distant both from the western slopes rain harvest and the eastern escarpment’s similar role for weather from the Tasman,so I’d expect a lot of variability in that area,compared with say the Murray at Biggara,or the Goobarragandra.

  3. From the Parliamentary Cttee on Forecasting:

    Professor Nicholls explained further that there is a major limitation with forecasts based on the ENSO in that:
    … prediction across March/April (e.g. of early winter rainfall) is very difficult, because this is the time that El Niño events are starting to develop but may not yet be sufficiently strong to be observed. This is known as the “autumn predictability barrier”, and its causes are still not understood.

    From BOM, on rainfall predictions for autumn:
    However, the skill of the forecast model is low in southeastern Australia for this time of the year, so this outlook should be viewed with caution.

  4. Canberra water utility – whose business model seems to be to charge households increasingly rapacious water bills, and pay increasing money to Govt, while keeping water restrictions permanent as dams overflow and during the dryer times tipping large amounts of water away as “environmental flows”.

    Don’t you Aussies have some legal recourse here? In the US it would be possible to launch a law suit against them for the areas effected by this nonsense. We just need a group of people willing to fund the legal action.

  5. OPEC does it with oil. De Beers did it with diamonds.

    I suppose that by a combination of monopoly, waste, prohibition and underdevelopment, it is possible that water could go the same way. Pretend specialists, such as now abound under the “climate” banner, will be ideal for the job. Any objections can be dealt with by the ingenious process of peer-reviewing one another’s idiocies. (Handy tip: If you see the word “model” used outside the context of Leggo or Hugh Grant’s love life, you are likely dealing with a born idiot.)

    More likely, Australians will soon wake up and remember what they are. You only have to look at the lightning emergence of our wool industry: from nought to a million pounds in a few years, under all but impossible conditions. When you read of the achievements of Elizabeth Macarthur, left alone with seven (surviving) children, you are left in a state of near-disbelief. This country was shaped so quickly, and the people who achieved it were fairly common types – along with plenty of drunks and crims – and there was barely an expert to be found. Explanation?

    As Peggy Lee used to sing: You gotta have heart.

  6. Gidday woodNfish – you might have trouble realizing the dominance of the GreenLeft in the ACT Govt. Results of the last election in 2008 show the numbers. Technicalities in ACT electoral law acted against the splintered right-of-centre groups that dominated “Others” – so they got ZERO seats while the Greens with a similar vote got 4 seats in the 17 seat assembly. We have a GreenLabor coalition Govt – and our Liberals are a weak imitation of conservatives. At our next election I would suggest that all the various right-of-centre splinter groups do not stand but tell their supporters to vote Liberal. I see no other hope to tip out GreenLabor.
    Re your suggestion of a lawsuit – I am not aware of any grounds for a challenge but will pass your idea on.

  7. Hi Warwick,

    I could be totally off base about a lawsuit, but I understand your frustration at dealing with this insanity. It just seems that if you don’t have the political representation and you don’t have the media to bring it to the publics attention, then the only recourse left is the courts. Not the preferred path, but if its all you have then you must use it or sit back and hope the next election turns it around.

  8. here’s the overview document from NIWA’s site; doesn’t seem consistent with the NZCSC’s press release

  9. Val, I have not yet found the BoM NIWA review – only this covering letter. Which is from the pdf “Report-on-the-Review…..” the last URL below.
    I started at this page and worked through their links. Where is the BoM Review of the NIWA 7 Station Series ?

    This “overview document” discusses the revision of the ‘seven-station’ series in more detail.

    The adjusted data

    “This spreadsheet” contains the adjusted data used to compile the ‘seven-station’ series
    NIWA has “documented its analysis for each of the 7 locations” – “Documentation of adjustments at each location” links to info re each site

    Scientific references

    A “PDF is available” here which includes the overview and the reports for each of the seven stations.

  10. found it, it’s on page 13 of the PDF
    The review does not constitute a reanalysis of the New Zealand ‘seven station’ temperature
    record. Such a reanalysis would be required to independently determine the sensitivity of, for
    example, New Zealand temperature trends to the choice of the underlying network, or the
    analysis methodology. Such a task would require full access to the raw and modified
    temperature data and metadata, and would be a major scientific undertaking. As such, the
    review will constrain itself to comment on the appropriateness of the methods used to
    undertake the ‘seven station’ temperature analysis, in accordance with the level of the
    information supplied.
    In general, the evidence provided by NIWA supports the homogeneity corrections that have
    been applied to the temperature record to create the ‘seven station’ series. The scientific
    papers clearly report on major issues which have been identified in the metadata and past
    scientific literature. It is also clear that a number of significant adjustments (as identified by
    NIWA in the reports) are clearly required for the raw/composite station series owing to
    inhomogeneities which would otherwise artificially bias results.


Leave a Reply

Your email address will not be published. Required fields are marked *