Sunday, 4 October 2009

How reliable are the input date to climate change models?

As is probably evident from the earlier posting on the climate debate, I have some degree of scepticism regarding totally anthropogenic climate change. The robustness of the models has always been a concern, and whether all the possible variables, such as natural earth and sun cycles, have been taken into account, as well as the interactions between dependent variables. Questions I hope will be answered at Climate Change '11.

A few days ago Amanda and I attended a lecture in Falmouth by Dr. Stephan Harrison, a quaternary scientist at the University of Exeter, and an advocate of human-induced global warming. I have to say that his lecture, crammed with data and graphs, was very impressive and presented a most convincing case for anthropogenic warming. It was evident that the mathematical models are now very sophisticated and take into account all known variables, including cycles of the sun, oceans, and vulcanism, but I have to admit that I left with the same degree of scepticism.

No matter how big and powerful are the computers, or the sophistication of the models, the latter are only as good as the input data. The two most important inputs, and the subject of the debate, are atmospheric CO2 content and mean atmospheric temperature and I am not at all clear as to how the levels of these variables can be effectively compared over long time intervals.

In recent years satellite technology has been employed to measure temperature, but prior to that temperatures were measured by thermometers in stations around the world. During the 20th century the number of stations providing data must have fluctuated, and those situated in or near cities must have been influenced by the 'urban heating effect'.

So, in my naivety I ask if anyone can provide me with a convincing argument that average annual temperature readings over a period of, say, one hundred years can be reliably compared. The same question applies to CO2 levels- were they measured as accurately a century ago as they are now?


  1. By email from Stephan Harrison, University of Exeter:

    You said: "During the 20th century the number of stations providing data must have fluctuated, and those situated in or near cities must have been influenced by the 'urban heating effect'".

    This is a good point. However, the inhomogeneities in the data sets are well known and accounted for. Climate data in the UK goes back to the 17th century (CET record) and we also use an enormous amount of proxy data to calibrate the recent instrumental record and to reconstruct past climate (eg dendroclimatology, speleothems, borehole records, glacier records, oxygen isotope records, ice cores etc). All show that our reconstructions are accurate and that the last few decades are the warmest over the past 2000 years or so.

    As you can see, this gets rid of the 'urban heat island' as many of these records are from remote regions unaffected by UHI. In addition, the UHI is well known and can be accounted for in climate records.

    We can measure C02 levels from proxy records too (including ice cores, plant stomata etc) and they also show a consistent rise with GHG emissions.

  2. From Jerry Perkins VP Development & Operations at Exeter Resources, Canada (via LinkedIn):

    Evidence is mounting that the models and ancillary data used by the IPCC are quite unreliable and in many cases manipulated and peer-review access withheld. As the evidence grows that climate change is non-anthropogenic it appears that the obvious trend to discount debate and pronounce the need for it to be over, increases exponentially. There could be a useful forum for evidence on both sides here as long as the debate is outward looking and seeking to influence a broader audience, whichever way it goes.

  3. From Terry Jackson Consultant at Maiden Creek Consulting, LLC, USA (via LinkedIn):

    I am also very skeptical of the anthropogenic climate change theory or our ability to control the climate. Anyone who wants to learn more about this subject should read "Unstoppable Global Warming... Every 1,500 Years" by S. Fred Singer. In addition to pointing out that the zeal behind this movement has developed a definite bias, there is considerable information that shows that Artic temperatures correlate closely with solar activity and much less so with atmospheric CO2.

    No doubt CO2 levels in the atmosphere have risen greatly in recent decades. However surface temperatures deduced from ice core and tree growth analysis indicate that temperatures have cycled up and down over the centuries. A closer analysis of recent surface temperatures indicates that the 1930s were hotter than the 1990s and that temperatures in the first decade of the current century has declined slightly.



If you have difficulty posting a comment, please email the comment to and I will submit on your behalf