Some time ago I was sent a link to a paper by Myanna Lahsen, an anthropologist who spent seven years studying climate modelers at the National Centre for Atmospheric Research.
Her research findings seem to focus on what the modelers said, without scrutinizing the extent to which what they said about the models, accorded with reality. In particular Myanna found that the modelers were very attached to their models and sometimes confused model output with reality. But it is unclear from her work how accurate the models were – the extent to which they did accord with reality?
Last time I looked I was impressed with the extent to which well known global warming scientist, James Hansen, was still on track with his 1988 prediction (Scenario B) about global temperature increase.
Then again between his Scenario A and C, he was covering a range of possibilities?
But hey, all predictions, made back in 1988, have been consistent with what has been a warming trend over the last 18 years.
In contrast, government scientists who made predictions about salinity along the Murray, in particular the NSW Riverina, got it really wrong.
The following graph shows what the models predicted would be the extent of the problem in the Riverina with, and without, a commitment to catchment and farm drainage plans.
The problem of rising groundwater in the NSW Riverina once seemed intractable. In 1990 123,300 hectares was considered at high risk of salinity because the water table was within two meters of the surface. At that time it was predicted that if the irrigators did nothing, by 2006 228,700 hectares would be lost to salt. If the irrigators committed to a $473 million program with $150 million from the state and federal governments, it was predicted that only 182,620 million hectares would be lost.
The irrigators committed to the program in the early 1990s including the implementation of drainage works often including water recycling systems to reduce recharge to the groundwater and improve water use efficiency.
The actual area now affected by shallow water tables is just 3,758 hectares – this is just two percent of the area that the NSW government thought would be affected under the most optimist scenario.
While I am pleased salt levels have been falling in the Murray River, and that the area at risk of salinity in the NSW Riverina has reduced to 2 percent of what was predicted, I am always amazed at how many people ignore this great news story.
Earlier this week the Australian Parliament’s Senate Environment Committee released a report about salinity. The report reads as though hardly anything has been achieved in address salinity in the Australian landscape.
The report recommends an extension of funding for the National Action Plan for Salinity and Water Quality – a project that began in 2001 with a budget of $1.4 billion. It is unclear from the senate report how this $1.4 billion has been spent.
Senator Andrew Bartlett chaired the committee and has a blog piece on the report here.
The Senate report repeats the finding from the National Land and Water Audit’s Australian Dryland Salinity Assessment 2000 that 17 million hectares of Australian farmland could be lost to dryland salinity by 2050.
Yet various recent reports have shown the 17 million hectare figure to be a gross exaggeration.
As Mick Keogh from the Australian Farm Institute recently explained:
“Increasingly, researchers are concluding that many of the assumptions and much of the data used in generating this estimate were wrong, or should not have been used. There are suggestions, for example, that some State salinity assessments used to calculate the national estimate overstate the current extent of salinity by factors of between three and seven times, let alone the projected future extent. Several of the state reports had no reliable data to base estimates on, and many made assumptions about future groundwater levels – a critical element in salinity assessments – that defy the laws of gravity and science, and are not supported by available data.”
At some point in time, the Australian community and the Australian Senate should accept that farmers have learnt how to manage salt. It hasn’t gone away, but the area affected by salinity is contracting and this is a great news story everyone should be shouting about.
But they are not.
And I am reminded of a comment posted at this blog about the same day the Senate released the report. While Geoff Sherrington made the comment in the context of global warming, it seems much more applicable to salinity modeling:
“Modelers would be prudent to keep their frameworks up to date, with periodic testing and private comparison with others, until consensus is reached that the methodology is scientifically good and all plausible effects are quantified.
Economists should not make predictions until that consensus is reached, unless they like eating humble pie.
If, in my earth sciences past, our company had announced a new ore deposit and given figures for its value to the Stock Exchange that was premature, we would well have ended up incarcerated.
If we got our maths wrong and mined a body that turned out a dud, we could go out of business and on the street. These outcomes instill a certain caution and accountability. Greenhouse modelers who produce premature estimates don’t have the same sword hanging over their heads. Their reward is more likely idolatry from supplicants.”
Certainly the doomsay salinity modelers have received nothing but praise, and the science managers who repeated their predictions promoted including to the National Water Commission, while I am often called all sort of nasty names at the popular blogs including at John Quiggin’s and Andrew Bartlett’s for daring to suggest they might have got their predictions very wrong.
Jennifer Marohasy BSc PhD has worked in industry and government. She is currently researching a novel technique for long-range weather forecasting funded by the B. Macfie Family Foundation.