Temperature Updates - November 2017 - HadCRUT

A Christmas present from the Met Office and Hadley Centre as they release the November HadCRUT data early.

This shows the anomaly in November to be 0.547°C, compared to the 1961 - 1990 base period. This is slightly cooler than last November, and in contrast to GISTEMP and BEST is only the 10th warmest November on record.

This is much more like the pattern for November in the NOAA data set. The differences between the different data sets possibly is due to the extend different sets cover the Arctic.

The 12 month rolling average for HadCRUT looks like this.

The average anomaly for 2017 so far is 0.685°C, down from 0.699°C up to October. My prediction for 2017 is now 0.681 ± 0.026°C, compared with last months prediction of 0.692 ± 0.041°C.

As can be seen from the graph it's difficult to see HadCRUT finishing anywhere other than 3rd warmest. In fact my statistical model makes this 100% certain.

HadCRUT4

Rank	Year	Anomaly	Probability	Change
1	2016	0.8	0.00%	-0.00%
2	2015	0.76	0.00%	-0.05%
3	2014	0.58	100.00%	+0.05%

I'm not going to disagree with that. December could be anywhere from -0.57 to +1.63°C, and 2017 would still finish 3rd.

December Round Up

With all the data sets I follow updated to November we have a pretty clear view of where 2017 will finish. GISTEMP, BEST, and RSS will probably finish 2nd warmest, with only an outside chance of dropping to 3rd. NOAA, HadCRUT, and UAH will probably finish 3rd warmest, with only UAH having a small chance of finishing 4th.

Temperature Update - November 2017 - GISTemp

NASA's released the latest GISTEMP data.

The anomaly for November was 0.87°C compared with the 1951 - 1980 base period. This is down slightly from 0.90°C in October. This is the 3rd warmest November on record. The last three Novembers have been the three warmest Novembers on record.

It is extraordinary how much warmer the last three years have been in November compared with any previous year. The equal 4th warmest Novembers were 2010 and 2013, and they were almost a tenth of a degree cooler than this current month.

The 12 year rolling average continues to cool from the el Niño peak.

The cumulative average for 2017 is virtually unchanged, down to 0.901°C from 0.904°C up to October.

Forecast

My forecast for 2017 is now 0.895 ± 0.025°C, compared with the forecast from October of 0.897 ± 0.04°C.

It is increasingly likely that this will be the 2nd warmest year on record. In any event 2017 will be very close to 2015, two very warm years sandwiching a much warmer year.

GISTEMP

Rank	Year	Anomaly	Probability	Change
1	2016	0.99	0.00%	-0.00%
2	2015	0.87	97.99%	+5.38%
3	2014	0.73	2.01%	-5.38%

For 2017 to drop to 3rd place will require December to drop to around 0.53°C. The last time a December was this cool was 2012.

Other Data Sets

I'm not giving detailed monthly updates on BEST and NOAA, but with one month to go here's a brief update on them.

BEST is very similar to GISTEMP. November was also the third warmest on record, and it is also shows 2017 on track to be second warmest, with similar (slightly better) odds to GISTEMP.

NOAA has a somewhat cooler month with November 2017 being identical to 2016 at equal 5th warmest. According to my statistical model NOAA is now 100% certain to finish 3rd warmest, and looking at the graph it is easy to see why.

As with HadCRUT the fact that NOAA is finishing in a different position to GISTEMP and BEST has more to do with how warm 2015 was than with a big drop in 2017.

Temperature Update - November 2017, RSS

As with UAH, the Remote Sensing System satellite lower troposphere data shows a big drop in the anomaly from October to November.

Details

The anomaly for November was 0.549°C compared to 0.805°C in October (anomalies relative to the years 1979 - 1998). This still makes this November the third warmest November, with the last three Novembers being the warmest three in the data - with all three being over 0.1°C warmer than the previous warmest November.

The cumulative average for the year dropped slightly, from 0.646°C to 0.637°C.

Here is the updated 12 month rolling average.

Predictions

My prediction for 2017 is unchanged from that for October, but with more confidence. The expected average for 2017 is 0.624 ± 0.028°C, compared with last months prediction of 0.624 ± 0.043°C.

Baring a major catastrophe it seems certain RSS will have its second warmest year on record, the warmest being last year.

RSS 4.0

Rank	Year	Anomaly	Probability	Change
1	2016	0.74	0.00%	-0.00%
2	1998	0.58	99.86%	+2.16%
3	2010	0.57	0.13%	-1.58%
4	2015	0.54	0.01%	-0.56%
5	2005	0.43	0.00%	-0.03%

My statistics suggest there is less than a 1 in 700 chance of RSS not finishing 2nd. But this would require December to have a negative anomaly (around -0.041°C). The last time December was this cold was 1995. There have only been 3 months during the 21st century this cold.

Temperature Update November 2017 - UAH

Dr Roy Spencer has released the November 2017 global temperatures for the UAH 6 lower troposphere satellite data set. This shows a big drop in the anomaly compared to last month's record breaker, but still pretty warm.

Details

The anomaly in November was 0.36°C compared to 0.63°C in October, and down a tenth of a degree from last November. However, this still makes 2017 the 2nd warmest November in the UAH data set. The last three years have had the three warmest Novembers in the set.

The mean of the first 11 months of 2017 is 0.372°C compared with 0.373°C for the first 10 months.

The 12 month rolling average is beginning to move down slightly from its unexpected jump upwards, but I expect it should start moving down more over the next year as La Ni&ntilda;a takes hold.

Predictions

With just one month of 2017 to go, my simple statistical analysis predicts 2017 will finish at 0.364 ± 0.028°C, compared with last months prediction of 0.359 ± 0.043°C.

This has made it far more likely that 2017 will finish as the 3rd warmest year, with less than a 1 in 50 chance of dropping below 2010 for 4th place.

Probability of different rankings for UAH 6.0
based on data to November 2017

Rank	Year	Anomaly	Probability	Change
1	2016	0.51	0.00%	-0.00%
2	1998	0.48	0.00%	-0.00%
3	2010	0.33	98.30%	+10.02%
4	2015	0.27	1.70%	-10.02%

However, I suspect this is too certain. For 2017 to finish below 2010, December would only have to drop to around 0.13°C. This would be equal to December 2013 and 2012, so with temperatures dropping from the very high values of the last few years it doesn't seem that unlikely. In any event, it's clear that 2017 will be another hot year, similar to 2010, hotter than 2015, and the hottest El Niño year that was not a major El Ninño.

How remarkable the last few years have been is much clearer if rather than focusing on one calendar year, we look at 2 yearly averages.