Tracking 2016 - The Begining of the End

A bit late due to the holiday season, here's the update for November and the final satellite data for 2016

For November the various surface data sets were mostly up or the same, with HadCRUT dropping slightly. For December all the satellite data show a big drop, but not enough to avoid beating 1998.

The margin graph, showing the expected value for each data set compared to it's previous record, shows a bit of a drop for HadCRUT and GISS in November. For the satellite data, UAH 6.0 (no longer in beta) and RSS 3.3 both beat their previous records by a small margin, around 0.02°C. In contrast the old UAH 5.6 and the new RSS 4.0 beat the old records by a very substantial 0.17°C.

What's really curious about this is how similar the final margins were for UAH and RSS. The two cooler sets (the new UAH and the old RSS) both beat 1998 by the same margin, and the two warmer sets (old UAH and new RSS) both beat 1998 by very similar margins. It's just coincidence, but it does demonstrate that 2016 probably was warmer than 1998.

The probability graph shows the probability (using a naive linear trend) of each data set setting a record. No real surprise here - most data sets are looking very likely to set a record, but there's still some doubt about HadCRUT, with a 96% chance of setting a new record.