Baloney Won't Stop Global Warming

Most recent answer: 09/18/2017

Q:
Two questions: Firstly, among other claims, Kelvin Kemm claims that the reduction of arctic ice is unlikely to be related to climate change. Is Kemm correct? Secondly Kemm claims that Al Gore's comments about the Bell curve are dumb. Who is right - Kemm or Gore?
- Ngoato (age 35)
Mamone, Limpopo South Africa
A:

The comments denying the existence of global warming due to man-made changes in the atmosphere are just plain wrong. The basic physics behind the effect have been understood for over 100 years. The results to date have been quite close to what Arrhenius predicted in 1906 based on those physics. 

The numerical arguments in that article are baloney. Of course the ice at the coldest polar spots isn't going to melt soon. In fact, at the South Pole increased precipitation caused by warmer oceans will cause the ice there to build up. The melting is occurring near the edges of the ice caps, which are near the melting point of ice in sea water.

The article doesn't get off to a good start when it starts to discuss actual science: Kemm writes "Now, for some basis physics: ice melts at 0 ºC. Okay, got that? " In sea water, ice actually melts at about -2°C, not 0°C.

Then it gets worse.

I hadn't followed Gore's discussion of the distribution of daily temperatures, using a bell curve, but just looked it up to see if Kemm's criticism makes any sense. It doesn't. Gore just pointed out that even allowing for the usual statistical variation in temperatures, there's been a very distinct shift toward increased temperature. Kemm seems to be attacking some imaginary argument that Gore explicitly didn't make, that a single warm day is a big deal. Although the real scientific argument about statistical significance involves time-dependent correlations, not mentioned by Gore, Gore's pictorial representation is much closer to reality than Kemm's incoherent rant.

I have paid a lot of attention to the Hockey Stick question, concerning the history of global temperatures over the last 1200 years or so. The arguments discussed with such intensity in that article concern primarily how big the error bars should be on the estimated temperature about 1000 years ago. Not only is Kemm much further off on that than Mann was initially, but the whole argument is irrelevant to what we need to know from those reconstructions. What the hockey stick shows, correctly, is that there's been a sharp warming in the last 100 years, unlike anything else in the record. It's that rise that's used to train the models, fine-tuning them to the data. The information we need from the hockey stick is whether rises like that sometimes happen for unknown reasons, which would mess up the model calibration. They don't. Even if the harshest critiques of Mann's initial work had held up (they haven't), that would mean that slow changes in temperature had bigger error bars than he thought. Since those slow changes aren't very important to train the models, that would have little effect on the uncertainty already present in the standard predictions.

If you'd like to follow up further, here's a serious recent paper on the topic:

Tingley, Martin P. , Peter F. Craigmile, Murali Haran, Bo Li, Elizabeth Mannshardt, and Bala Rajaratnam. On Discriminating between GCM Forcing Configurations Using Bayesian Reconstructions of Late-Holocene Temperatures. Journal of Climate. 28: 8264–8281, 2015

Mike W.


(published on 09/18/2017)