Communicating risk – (Pt 1) just the facts


The Newcastle 2007 storm

The truth is incontrovertible

According to Drew Warne Smith and James Madden writing in the Nov 7th edition of the Australian:

 “The science is in on sea-level rise: 1.7 mm” , …we don’t need to worry about sea level rises in Australia as a ‘scientific’ 1.7 mm rise is a third less than the government’s overheated predictions…

How Smith and Madden set out to construct a case that government predicted sea level rises are exaggerated provides an excellent example of how fallacious arguments can be used to misinform the unwary, and in this case skew the reader’s perception of risk.

This first post will look at the facts upon which their argument is based, in the next post I’ll take a look at broader issues of the psychology and logic of their argument, while in the final post in this series I’ll look at the role of the media in mediating perception of risk.

Instead of working on your opponent’s intellect by argument work on his will by motive; and he, and also the audience if they have similar interests, will at once be won over to your opinion, even though you got it out of a lunatic asylum.


What Smith & Madden are arguing

Wading through the journalistic effort of Madden & Smith they base their argument upon the following premises:

1.  the average global sea level increases for the 20th century was 1.7 mm per year,

2.  the numbers the government (e.g IPCC & CSIRO derived) uses are a global average not local values and so are not representative,

3.  at Port Kembla (a local gauge for NSW) there has only been a 1.9 mm rise per annum since 1991,

4.  in the last twelve months Pt Kembla has only shown a 0.1 mm rise,

5.  only a thin layer of the earth’s ocean is warming,

6.  there is little compelling evidence that the polar caps were melting, and

7.  local resident’s haven’t noticed a change.

So, because there is no evidence that sea level rise is accelerating (1),(2),(3),(7) or that anything will cause it to accelerate (5),(6), the predicted 6.6 mm per year rise (based on IPCC and CSIRO modelling) is exaggerated and because it’s an average anyway you shouldn’t use it (4) and if you use local historical gauge data (3),(4) the rise is much less than your predicting.

Is the sea rising?

So let’s have a look at the first of the premises of the argument e.g. that ‘the average global sea level increases for the 20th century was 1.7 mm per year’. Douglas (1997) estimated that sea level rise has occurred at a mean rate of 1.8 mm per year for the past century while Church and White (2006) found 1.7+/-0.3 mm rise per year, so far so good. However Church and White also found that the rate was increasing at 0.0013 +/- 0.006 mm per year per year (Church 2006) giving a 21st century rises consistent with the IPCC TAR estimates of 280 to 330 mm rise in seal level from 1990 to 2100. Over the period 1993 to 2003 satellite radar altimetry also indicated sea level was increasing at 3.1 ± 0.4 mm per year (Leuliette et al. 2004), although it’s not entirely clear (at the moment) why this is greater than tidal gauge figures.

Truth is incontrovertible, ignorance can deride it, panic may resent it, malice may destroy it, but there it is.

Winston Churchill

So what the authors are doing in this instance is basing their proof on a carefully selected  figure that does not represent the current rate while conveniently omitting that the rate of sea level rise is increasing.

Using local data is better

The statement that local data is better than global averages is on the face of it a reasonable assumption. Of course one should also then recognise that we don’t have ‘local data’ for future events, thus the argument is constructed so that historical data (Port Kembla) is introduced as a predictor of future sea levels. The CSIRO has actually constructed models for local sea levels based on best to worst case scenarios (CSIRO/BOM 2007) so the presentation of this ‘assertion’ is in fact simply a ‘straw man’ that allows the introduction of a carefully selected historical data set as a predictor of future sea level.

Cherry picking

The author’s also state that at  ‘Port Kembla there has only been a 1.9 mm rise per annum since 1991, and in the last twelve months has only shown a 0.1 mm rise’.  Again the author’s fail to state that the average for Pt Kembla (across >25 years of data up-to 2003) was 0.65 mm/year (NTC 2003) indicating that the average from 1991 to the present is substantially greater, so again the data quoted is selective and misleading. The figures also don’t account for barometric corrections which if included will increase the sea level.

Ignoring known climate effects

Finally the smaller 0.1 mm increase in the last twelve months would be expected given that we are currently experiencing an El-Niño climate event that results in lower sea levels than normal in the Australian region.

Only the top bit of the ocean is warming

Therefore, although this is an unstated implication, sea level rise will be limited. Now certainly the uncertainty over the mechanisms driving sea level rise, for example the involvement of deep oceans in taking up heat, is recognised as an area of weakness in climate models. Even after AR4 (the fourth IPCC report) this uncertainty has not diminished appreciably. But, and it’s a big but, when we compare actual data against current model predictions we find that the actual data is still trending to the upper predicted values of the models. So the problem remains that actual data is trending at the extreme end of what current models predict or to put it another way we are under predicting not over predicting. So fundamentally this assertion is what’s called a ‘red herring’, it may be true, partially true or not true at all but it doesn’t address the known under prediction of current models regardless of what ever set of assumptions they are based on.

The ice caps aren’t melting

This statement implies that the ice cap melt is the predominant driver for sea level rise. In fact there are various sources, one of the most significant being the thermal expansion of the ocean. Interestingly the range  of ice cap contribution (Antarctic and Greenland) ranged from a net negative impact on sea level rise to net positive in the third IPCC report which still predicted significant overall sea level rises. Again a red herring.

But the local’s haven’t noticed it

Human’s are notoriously bad at noting small incremental changes, so this is yet another non sequitur that proves nothing one way or the other, sigh…

In the Next Post

Communicating Climate Risk – (Pt 2) Fallacious Arguments.

Further Reading

Douglas, B.C., Global Sea Rise: A Redetermination, Surveys in Geophysics 18: 279–292. doi:10.1023/A:1006544227856, 1997.

Church, J. A., and White, N. J., A 20th century acceleration in global sea-level rise, Geophys. Res. Lett., 33, L01602, doi:10.1029/2005GL024826, 2006.

Commonwealth Scientific and Industrial Research Organisation (CSIRO) and Bureau of Meteorology (BOM), Climate Change in Australia – technical report 2007, 2007, CSIRO.

Leuliette, E.W., Nerem, R.S., and Mitchum, G.T., Calibration of TOPEX/Poseidon and Jason Altimeter Data to Construct a Continuous Record of Mean Sea Level Change. Marine Geodesy, 27(1-2), 79- 11 94. 12, 2004.

National Tidal Centre, Australian Bureau of Meteorology, Australian Mean Sea Level Survey 2003.