Tristanc you don't think it odd that there are so many, or to be more accurate, so many and so many higher-reading calibration errors are on the red side, compared to the blue side, of that graph that Greenhart and others have posted here on this thread and on HC so often?
I wonder why the scientists/meteorologists chose to "homogenise" so many of the too-warm-too-early last century anomalies [eg. 1913-1973] as outliers rather than going the other way instead [now that really would give an entirely different trend line] and simply including all real data initially and then if deemed necessary removing a defined number or a like percentage of extreme examples from both sides rather than selectively removing double the number of 'too-warm-too-early' red side "errors" and in amongst that doubled selection also picking/choosing/including such a large percentage of "errors" with a quantifiably greater "error"?
eg. Tristanc there are a massive20 sites, deemed in need of homogenisation, with a difference in trend in access of 0.8deg. on the right, red side, compared to a mere 3 sites chosen in excess of 0.8deg. on the left, blue side, of the graph.
I wonder how choosing to do that has impacted the end result Trend Line?
I'd love to be the arbiter doing the selecting of which sites may or may not have records which need data homogenisation apparently based on some dreamt up long forgotten, unprovable anomaly.
Like for example I'd not be picking Rutherglen's data which didn't actually ever shift its site. Yes that would be a sound basic point.
I reckon I could come up with a global cooling Trend line over 100 years if I did the homogenisation site picking but then I'm not a meteorologist-scientist and I don't have an agenda of Global Warming to prosecute thereby to protect my wage.
I believe a 5year old on average would be able to estimate that the area in blue, when shaded, is not a third of the red shaded area.
Even a non-scientist like me then sits up and starts looking at the magnitude of those red 20 : 3 blue, large 'differences in Trend' and realises the connivance.
Odd that these homogenisations always favour the Alarmists case don't you think Tristanc and Greenhart?
