Temperature Update - September 2018 - UAH

Update - 5th October 2018

Added graphs and a correction.

Data

According to the UAH satellite data, global temperatures were 0.14°C above the 1981-2010 average. This makes September the coldest month (by anomaly) of 2018 so far, with the month being the equal 16th warmest September in the 39 year history of satellite data.

This is a big drop from last September, though the underlying trend for Septembers remains somewhat higher than the general UAH trend, at 1.49°C / century. The underlying trend for all months is still 1.28°C / century.

This month is somewhat below the trend and the average of the previous 12 months is exactly on the trend.

Forecast

The average for the first 9 months of 2018 is 0.218°C. The statistical prediction is 0.234 ± 0.059°C, compared to last month's prediction of 0.252 ± 0.066°C.

This is one of the bigger drops in the forecast, though it is still pretty close to the forecast from February.

The forecast has been surprisingly stable so far but I expect it will drop in the following months, baring a big last minute surge. To reach the current mid-point prediction will require the next three months to average over 0.35°C. This is more than any month so far this month. My calculations were wrong. The next three months have to average a little more than 0.28°C, which seems more likely, but still a tall order. Only one month so far this year (July) has been warmer than this at 0.32°C. Given that, take the following probabilities with a grain of salt.

My statistical method suggests that 6th place is still the most likely with around a 62% chance of 2018 finishing between 2002 and 2015. There's a 10% chance of finishing 5th, and a good chance (over 25%) of finishing below 6th - possibly as low as 10th.

UAH 6.0

Rank	Year	Anomaly	Probability	Cumulative
1	2016	0.52	0.00%	0.00%
2	1998	0.48	0.00%	0.00%
3	2017	0.38	0.00%	0.00%
4	2010	0.34	0.04%	0.04%
5	2015	0.27	10.68%	10.71%
6	2002	0.22	62.03%	72.74%
7	2005	0.2	16.21%	88.95%
8	2003	0.19	5.66%	94.61%
9	2014	0.18	1.55%	96.16%
10	2007	0.16	3.17%	99.33%
11	2013	0.14	0.54%	99.87%
12	2001	0.12	0.12%	99.99%

But note that 6th place year 2002 has an anomaly of 0.22°C, slightly above the current average for 2018. I suspect it is more likely than not that 2018 will be cooler than 2002, whatever my statistical analysis says.