It's past the end of 2015 so I thought this deserved an update. The original point of this thread was to discuss that 2014 was (then) on track to be the warmest year on record, which in fact it turned out to be (if only barely). Now with 2015 in the books, last year was much warmer again, making it again the warmest on record. Of course "on record" only means the last ~ 150 years give or take.
Anyway, I'm revisiting this thread to make further light of my naive predictions from last year, but also to claim that I'm not an idiot yet, we need another year (or more) to see if time (and climate) proves me for the idiot some people think I am...
So last year, I made some predictions, which I quoted in the last post. They were basedon my own naive analysis of the trends in global anomaly data, without making any effort (and in fact deliberately resisting any inclination) to consider causative relationships, but rather just let the data speak for themselves and see what they might tell us.
I started by plotting annual and decadally-smoothed anomaly data, and finding a fairly simple (second order) best fit:
Then I flattened the data by transfroming the second order best fit to a horizontal straight line:
Next I drew a crude sine curve to capture the broad behaviour of the decadally smoothed data:
Then I drew some parallel error bounds, upper and lower, that capture the 95th and 5th percentile of annual data:
Then I took the best fit sine curve and upper and lower bounds, and transformed them back to the original second order best fit to generate a "naive prediction," shown here with the best fit in reen, and upper and lower bounds in red/orange, and the annual data as orange dots. You can see this inferred trend captures nearly all the annual data:
... to be continued