https://quadrant.org.au/opinion/doomed-planet/2019/09/relax-weve-
“God forbid we have a real global crisis to deal with. Either we won’t believe it after suffering this present hoax or we’ll think that the equivalent of building some windmills will solve it. What global warming / climate change/ climate emergency (choose your favourite) has shown is that large sections of humankind are susceptible to hysterias promoted by elitists; and that governments are singularly incapable of acting resolutely to combat crises to which they say they subscribe. ”
A friend of mine, Geoff Hogbin (Free to Shop, CIS, 1983), made an incisive throwaway comment some time back which has stuck with me. How would the climate models do he said, sceptically, if they were required to perform across data from the end of the Industrial Revolution; from say the mid-nineteenth century. Of course, there is a data problem. Or I imagine that there is.
The official atmospheric concentration of CO2 measured at Mount Mauna Loa in Hawaii goes back only to 1958. HADCRUT temperature data on a global scale goes back to 1850 but you have to ask how consistent are the temperatures recorded in the 1850s with those in the last ten, twenty and fifty years? Satellite temperature data which is probably more reliable than land and ocean-based measures only goes back to the end of 1978.
If it were possible, I would like to see the climate models applied to separate sub-periods, at least back to 1850, to see how they perform. I will guess. They wouldn’t do too well. But my prejudice is showing.
When, four decades ago, I dabbled in econometrics, an important test of a model calculated over a whole data series was to see how it performed using only the first half of the data set and then separately the second half. Was it good for all seasons? Usually it wasn’t. Conclusion: throw it away and begin again.
Almost all of the climate models have over-predicted global temperatures. Or, again, I think that they have. It is easy to find commentaries which say that they have a pretty good record overall, as it is to find the opposite. Hard for us ordinary punters to get a handle on it.