Chris Knappenberger has highlighted some of the problems both warmers and sceptics face with available climate data.
But it gets worse than this because (and ignoring the satellite data as these don’t suffer from the problem) all the surface temperature data sets are simply aggregations of the intensive variable temperature. Aggregations of an intensive variable is about as meaningful as calculating the median value of a set of telephone numbers.
So what you might ask, what has that to do with the global warming issue? Quite simple really.
When a surface temperature is recorded, a measurement is made of the thermal state of two objects in thermal equilibrium – the calibrated thermometer and the object whose temperature we wish to know. A simple spreadsheet example will demonstrate this.
|Mean Temp only||12.5|
Two statistics can be derived from the data set –the simple mean of the temperatures, 12.5 degrees Celsius which is the mean temperature of the thermometers themselves, (assuming identical thermometers) while the mean temperature of the various volumes of matter, is 1.08 degrees Celsius. Which is correct? Both – depends on which group of objects you want to compute the thermal state of – thermometers or volumes of matter.
Climate science restricts itself to calculating conclusions from the mean temperature only – an aggregated intensive variable that has, in reality, no physical meaning. Both the warmers and sceptics blithely use this, albeit mathematically precise, statistic to argue their positions.
Cherry-picking the data usually happens when the data is random and auto-correlated. To make matters worse, climate science seems to derive further statistics from statistics that were in the first step, improperly calculated. This usually happens when students from the social sciences start sticking their noses into the physical sciences without understanding the basics. And mathematicians are even more clueless in this regard.
Climate Science in it’s present state is spending a lot of taxpayer money on deriving conclusions from the thermal state of a set of thermometers. They are certainly not deriving statistics of the thermal state of the earth’s atmosphere, because this has not been calculated at the first step.
If they want to get close to deriving a global mean temperature from the data, then each and every temperature reading has to be initially weighted with the surface area assigned to a particular measured point – otherwise known as the “area of influence” weighting routinely used on mineral ore reserve estimations. It’s the reason mining and exploration geologists place no credence to the AGW theory.