Temperature Update - July 208, GISTEMP

According to the NASA GISTEMP data set, July 2018 was 0.78°C above the 1951-1980 base period. This is almost identical to the anomaly for last month and continues something of a downturn compared with earlier in the year when the anomaly was over a tenth of a degree warmer. In a way this seems surprising when you consider how hot it has been over much of globe, and is a good illustration of why it's important to look at the whole picture and not just the part of the world you are living in.

It also demonstrates a curious discrepancy between the satellite and surface data sets. All surface data sets have shown, at least using 12 month smoothing, a fairly consistent cooling since the height of the 2016 El Niño, whereas satellite data shows most of the cooling stopping in 2017, with temperature hardly changing since then.

All this is relative though. GISTEMP might be a lot cooler than it was in 2016, but 2018 was still the 3rd warmest July on record, with the last 3 July's being the three warmest on record, with 2015 being the equal 4th warmest July.

As the graph shows the last three years have all had relatively warm Julies. Yet despite this, July 2018 was slightly below the trend counting all months and starting in 1970. This trend is 1.80°C / century.

The average anomaly for 2018 so far is 0.817°C, continuing to decline from a peak of 0.840°C in April. Despite this my prediction for 2018 rose ever so slightly. It now stands at 0.823 ± 0.079°C, compared with the prediction based on data up to June of 0.820 ± 0.087°C.

NB, I don't always spell this out but when I quote the previous months prediction I'm recalculating it with current data. This is not necessarily going to be the same as the actual prediction I made the previous month. Differences can arise because the earliest figure published around two weeks after the end of the month will be modified over the coming weeks. I prefer to use the most recent data, and it's easier to just read the corrected values from my R script than go back over previous postings. For what it's worth my initial prediction made from June 2018 was 0.825°C, so the prediction from July is slightly down not up, but it really makes almost no difference.

That disclaimer over, the prediction is now increasingly that 2018 will be the 4th warmest year on record, placing it between 2010 and 2014. There's around an 85% chance of it finishing exactly in 4th place, but only just over 1% that it will finish below 2010. There's around 14% chance that it will finish above 2014 for 3rd or higher, including a small but non-negligible chance of finishing in 2nd place.

Rank	Year	Anomaly	Probability	Cumulative
1	2016	0.99	0.01%	0.01%
2	2017	0.9	3.55%	3.56%
3	2015	0.86	10.67%	14.23%
4	2014	0.73	84.52%	98.74%
5	2010	0.7	1.13%	99.87%
6	2005	0.67	0.11%	99.98%
7	2013	0.64	0.02%	100.00%