The Royal Society of New South Wales held a meeting on Saturday in Mittagong on 'Global Warming and The Cosmos'. Speakers included the director of the Danish National Space Centre, Eigil Friis-Christensen, and Graeme Pearman, former head of the CSIRO Atmospheric Division and now a consultant with GP Consulting Pty Ltd and an advisor to Al Gore and Ross Garnaut.
Graeme Pearman and Eigil Friis-Christensen, Mittagong, April 5, 2008
Dr Pearman spoke first and focused on global warming from carbon dioxide as a “policy driver”. I was offended by the presentation.
Dr Pearman suggested that much of the 0.7 degree Celsius increase in the earth’s temperature over the last 100 years has occurred in the last 10 years. Yet the last really hot year was in 1998 and global temperatures have since plateaued.
Graph and fitted spline curve from 1979 through to February 2008, from Professors John Christy and Roy Spencer, University of Alabama, Huntsville
Dr Pearman referred to 95 and 99 percentiles as measures of the “proof of an hypothesis” in the same breathe claiming that that there was more than 90 percent proof that global warming is a consequence of greenhouse gas emission. Yet this 90 percent figure, sometimes used by the United Nation's Intergovernmental Panel on Climate Change, is not from the testing of a falsifiable hypothesis but rather a political expression of the strength of opinion.
Dr Pearman began his presentation by suggesting that the break-up of the Wilkins Ice Shelf was a consequence of global warming. When I questioned him on this issue, he told the audience that Antarctica is warming.
Yet it is generally accepted and uncontroversial that 95 percent of the landmass of Antarctica has cooled over the last 20 years.
Image depicting the heating and cooling trends over and around Antarctica (1982-2003). Blue indicates cooling trends and red indicates warming trends.From NASA/Goddard Space Flight Center Scientific Visualization Studio.Data provided by Larry Stock.
There has been warming at the edge of the continent including where Wilkins Ice Sheet recently collapsed. The collapse could be due to global warming , oceanic volcanoes, and/or from internal stresses associated with the accumulation of ice in the bay.
-----------------------------------
I am grateful to John McLean for information on temperatures in the vicinity of the Wilkins Ice Shelf and Joe D'Aleo for other temperature data, and to Bill Kininmonth and Garth Paltridge for information on the Wilkins Ice Shelf collapse.
I shall elaborate on the presentation by Dr Christensen in Part 2, to be posted in the next day or two.
[...] 2. Marohasy, J. Graeme Pearman Claims Antarctica is Warming. April 7, 2008. https://climatelab.com.au/blog/2008/04/graeme-pearman-claims-antarctica-is-warming-global-warming-... [...]
As part of the work undertaken for the Fourth Assessment Report of the IPCC13, about 20 different climate models were run with historical changes to natural and anthropogenic forcing factors to simulate the climate of the 20th century. The simulated changes in Antarctic surface temperatures over the second half of the 20th century vary greatly from model to model with no single model reproducing exactly the observed pattern of change. However, when results from all models are averaged, the resulting pattern of change bears some resemblance to that observed, with greatest warming in the Peninsula region and little change elsewhere. This result suggests that some of the observed change may have an anthropogenic origin, but the lack of a clear and consistent response to changed forcing between models also suggests that much of the observed change in temperatures may be due to natural variability. The IPCC model experiments fail to reproduce some of the observed features, notably the rapid warming of the lower atmosphere. These differences between modelled and observed changes could be used to argue against attributing change to anthropogenic forcing but some caution is called for as the models used may not adequately represent all of the complex processes that determine temperatures in the polar regions.
http://www.antarctica.ac.uk/bas_research/our_views/climate_change.php
Looking back a couple of threads we could guess this spline curve thingy would turn up again and again. How easily Jennifer tosses it in as a contrast with the Pearman view.
As two other aspects have not been challenged properly, let’s open it up. Lost in space are both steps in temperature trends over the last half a century or so. Also is it fair math to employ anything other than Fourier type analysis with so much rhythmic forcing around?
IMO eliminating any steps is foul play by a would be pure statistician.
Satellite work is not for the naïve and neither is high speed signal analysis. Working back to fundamentals from complex events is still an art not a science.
Sure smoothing the cake with icing provides a fake finish and the candles provide a beacon for its rites of passage. Despite all the decoration I for one am not convinced by the extra layers of icing promoted particularly on these blogs. Icing is for kids. Standing by your nearest monitoring station seems quite beyond our average trooper.
Any one who has sat with a CRO looking for distinguishing features in say a human voice will know where I’m coming from. Yes, vibrations are present as in all music but ask yourself this, what’s your voice look like after its passage over a mobile phone system? Fudge won’t do. Good sampling and filtering techniques are essential in our modern communications.
Let’s add though, patchy data is most bewitching for our come lately non engineering science writers.
Arguments over sparse signals and poor coverage are best paralleled here in the CDMA v TDMA choice some of us had with mobile phones. This war is not over.
http://www.arcx.com/sites/CDMAvsTDMA.htm.
The study of energy distribution goes on and the case for extra wide bandwidth, finding signal energy on the cell fringe etc comes down to our practical experience over time.
http://en.wikipedia.org/wiki/Spread_spectrum
BTW That NASA image can be used for either scenario, warming & cooling
There was that saying that "a staticician (protesting if good) could fit a curve through three points, an engineer through two and an eco-fadist through one"
Dave: "Who knows how the satellite data was "adjusted" to make this graph."
Christy and team area open about how they do things, unlike some others I could mention who change things without telling.
If you wnat to track them you can see the 'readme' in the data directory for example:
http://vortex.nsstc.uah.edu/data/msu/t2lt/readme.03Jan2008
Jennifer,
The thing to keep in mind from Mr. T's link
"Instead, the team checked the satellite records against ground-based weather station data to inter-calibrate them and make the 26-year satellite record. The scientists estimate the level of uncertainty in the measurements is between 2-3 degrees Celsius"
Who knows how the satellite data was "adjusted" to make this graph.
The data is based on the same data as your picture .
Mr T, "Jan you saved me from Louis!"
I'm sure you realise that was not my intent. It's just on this occasion he was IMO less right than you.
Even sawtooth triangular and even squarewaves can according to Fourier theory be built out of adding a series of sinusoids together ( int the cases mentioned they are all harmonics of the fundamental of devreasing amplitude take to infinty). Given that the main drivers of our climate are all rotational or orbital factors it stands to reason that there are many sinusoids making up historical climate data. Removing the noise (cloud movement etc effects) and the short term (ie diurnal, annual, ENSO etc) one by one should reveal eventually a sinusoidal signal relating to the longest term orbital factor. I don't think that has been reached here but I do think one of the shorter term ones has surfaced - it will take a little/lot more time to check that to my satisfaction at least.
Jan you saved me from Louis! Thanks muchly
And yes, I took a guess about the three point thing, it just looks like the spline hasn't used many data points.
Louis: "All plots are 2D representations, so only one solution is possible"
I'm not sure what you are on about. You said "A fit between 3 data points is a linear one, not a spline."
This is not correct mathematically or physically a line is defined by two points if a third point lies on the line good and well. If not then you need either a curve of some sort (includes splines) or you may do a linear regression. Since the time temperature series have orbital factors driving them which gives rise to sinusoids I don't think a straight line is going to accurately represent them any time soon.
Jan
All plots are 2D representations, so only one solution is possible.