According to the UK Met Office Hadley Centre data set HadCRUT, the global anomaly for April 2018 was 0.629°C above their 1961-1990 base period. This is slightly higher than the previous month, and makes April the warmest, by anomaly, month of the year so far. It is the 7th warmest April in the HadCRUT record starting in 1850.
Two years ago there was a strong El Niño, which combined with ongoing global warming, resulted in 2016 being the warmest year on record, in all data sets. February 2016 in particular had the warmest anomaly for any month with GISTEMP showing an anomaly of 1.34°C, two tenths warmer than the previous month which was also a record. In fact using GISTEMP thee 7 months from October 2015 to April 2016 were all more than a degree Celsius above the 1951-1980 base period, the 7 warmest months on record up to that point. Since then monthly temperatures have been up and down, with more months over the 1°C mark in 2017, and never lower than 0.69°C.
The one point this really enforces is that global warming never went away, despite the many cherry picked pauses
.
But inevitably a record breaking El Niño can be the start of a new pause, or even a cooling trend, when as is inevitable temperatures fall below the trend.
But that's too far away for some people who need to distract from how warm the globe has become.
Welcome to the two year hiatus, or rather the greatest cooling event on record.
This all started with an article on a website called RealClearMarkets, by someone called Aaron Brown. He appears to be a financial author, and a self confessed libertarian. He asks the question Did You Know the Greatest Two-Year Global Cooling Event Just Took Place? Now he insists in this article that nothing he says should argue against global warming, that he's simply asking why there is no media attention about this event. I'll give him the benefit of the doubt, but in any event his arguments are incredibly silly.
He starts,
Would it surprise you to learn the greatest global two-year cooling event of the last century just occurred? From February 2016 to February 2018 (the latest month available) global average temperatures dropped 0.56°C. You have to go back to 1982-84 for the next biggest two-year drop, 0.47°C—also during the global warming era.
He's already had to add a qualification, it's not the greates event ever, just in the last 100 years.
Then he starts describing the big chill in terms of two little chills
and claims that if this big drop were to continue we would be back to 1980s levels in a few months.
I'm not sure if this is correct, but it probably isn't going to happen, so why mention it now?
The 2016-18 Big Chill was composed of two Little Chills, the biggest five month drop ever (February to June 2016) and the fourth biggest (February to June 2017). A similar event from February to June 2018 would bring global average temperatures below the 1980s average. February 2018 was colder than February 1998.
The dog that didn't bark in this article or any of thw many posts copying it is any attempt to put this in context.
In particular there is not a single graph showing what this cooling event looks like.
Nor does he mention that the anomaly in February 2018 was still 0.78°C warmer than the 1951-1980 average, that this was the 7th warmest February on record,
or that it is still warmer than 4 years ago.
He makes the point that it is colder than February 1998, but doesn't say by only about 0.1°C, or point out that this means that even after the greatest two year cooling event temperatures are only slightly lower than the record breaking temperature of 20 years ago.
It would be easy to read this talk of a cooling
event as implying temperatures are now cold, when they are still some of the warmest on record.
Here's the graph of GISTEMP data with the start and end points of the greatest teo year cooling event marked in red.
Of course the really obvious point is that this is only the greatest cooling because it started on such a warm month. To put this in even more context, using Brown's technique you should also point out that the two years leading up to this great cooling saw the greatest two year warming event, not just in the past 100 years, but since records began. That this warming event was over 0.8°C, far bigger than his cooling event. More than 0.25°C bigger than the comparative two year warming event of the great El Niño event of 1998.
Then there's this puzzling description of two little chills. Here's a graph of the last few years, with the start and end months marked.
There's a quick drop from the peak month in February (the first little chill), but this actually gives us all of the cooling, temperatures in November 2016 were lower as an anomaly than they currently are. His second little chill ignores the fact there was a rise in temperatures leading up to the second drop. But that just echos what happened in the years leading up to the big chill - it was preceded by a much bigger warming.
At this point it should be realized that all Brown's descriptions of chills are based on simply comparing the start month to the final month, and ignoring everything that happens in between. This is a very bad way of assessing what temperatures are actually doing. Monthly data is very variable and any difference over two years are likely to be down to chance rather than an actual trend. In this case the problem is in Brown claiming that the last two years have set a record. This is true if you ignore a bigger drop 101 years ago, but it depends entirely on this arbitrary two year period. This graph shows both the 1998 and 2016 strong El Niños, with the peak February temperature and the subsequent February tewo years later.
The obvious point here is that February 2000 was somewhat warmer than January 2000. If brown based wanted he could have taken the cooling from February 1998 to January 2000 and found it was greater than that from February 2016 to February 2018. It's arbitrary, and self serving, to call the last two years as the greatest cooling event this century.
A better way of determining cooling trends would be to look at the actual trend over a two year period, using linear regression. This is a standard technique, though it still doesn't make much sense to look at a two year period it does at least take into account all the data over that period. The problem is, if you do that Brown's argument that there was anything unusual about the last two years goes out the window. The fastest recent cooling two year trend was around 17°C / century. This sounds dramatic, but between 1998 and 2000 it was over 23°C/ century. More importantly there have been 7 periods over the last century with two year trends faster than the post 2016 big chill, and many more almost as steep.
This graph shows every 2 year trend over the history of GISTEMP.
Strong two year chill rates happen all the time, just as do strong warming events.
Brown's main point is that this is not about the reality of global warming, but about bias in the media.
My point is that statistical cooling outliers garner no media attention. The global average temperature numbers come out monthly. If they show a new hottest year on record, that's a big story. If they show a big increase over the previous month, or the same month in the previous year, that's a story. If they represent a sequence of warming months or years, that's a story. When they show cooling of any sort—and there have been more cooling months than warming months since anthropogenic warming began—there's no story.
Now I wouldn't disagree that the media has biases, and that in reporting on climate change they are often wrong or simplistic. I'd also agree that it is a problem that reporting on climate change is often limited to mentioning when record breaking years, and focusing on statistical outliers, rather than the more important concerns of ongoing warming.
But, constant complaints of bias often show more about the accusers own bias. They want the things that support there beliefs to be reported.
Brown tries to argue that the recent two year cooling event should be as newsworthy as 2016 being the warmest year on record based entirely on the statistical likelihoods of the two events. He says that the cooling event was a three sigma event, something that only happens rarely, whereas warmest years happen every 8 years or so.
But apart from the fact that the two year cooling event isn't really a 3 sigma event, one you understand the large amount of cherry picking associated with it., the problem is on wanting to decide what stories are covered based only on how likely the event is. He thinks events that are rare are more newsworthy than events that are important. Record breaking hot years might happen 12% of the time, but that's still an important news story given both what it tells us about global warming and the effects on the planet. By contrast his great cooling means nothing except the planet is back on it's regular warming track.
Aaron Brown's piece has had an odd history since it appeared on the 24th April. Almost immediately James Delingpole was ripping it off for Breitbart, in a piece called Earth in ‘Greatest Two-Year Cooling Event in a Century’ Shock, 26 April. This essentially repeats what Brown says, but more forcefully, and adds even more confusion. Delingpole tries to put the shocking event in context:
To put this temperature drop in context, consider that this is enough to offset by more than half the entirety of the global warming the planet has experienced since the end of the 19th century.
This is a common trick, point to a drop lasting a few months, claim it wipes out a certain percentage of the last century of warming, fail to distinguish between long term warming averaged over decades and a single month, and ignore the fact that the actual drop was from a much higher starting point, so that even after this drop nothing has been offset. In fact February 2018 was over a degree warmer than the late 19th century average - far from offsetting any warming it has added 25% to Delingpole's claimed warming.
He finishes by asserting that
As we know from long experience, if it had been the other way round – if the planet had warmed by 0.56 degrees C rather than cooled, the media would have been all over it.
Except we don't have to imagine that, the world did warm by far more than that, between February 2014 and February 2016 the world warmed by over 0.8°C, over 50% more than the shocking cooling, and adding as much warming as all the warming of the past century. Yet no-one seems to have mentioned that fact, as they were too busy talking about how February had been the warmest month on record.
Three weeks after the Brown piece, something called Investor's Business Daily had an editorial called Don’t Tell Anyone, But We Just Had Two Years Of Record-Breaking Global Cooling again referencing the Brown piece and complaining that no other news outlet had picked up on it in three weeks. GWPF reproduced the editorial, WUWT copied the GWPF copy of the original without acknowledging either.
The WUWT article prompted me to add some graphs putting the record breaking global cooling
in context in their comments.
Had an usually large set of angry responses, suggesting I might have hit a nerve.
The main complain was that I was using GISTEMP, which I was repeatedly told was garbage, and not caring that that was the data set Brown uses.
In fact different data sets won't show the record.
Then we had Tony Heller, suddenly pointing out that there had been the greatest two year drop in temperature in a century - two months after it had happened and without acknowledging Brown's original, or any of the other articles that had already mentioned it. In his post Earth Cooling Fastest In A Century – Since Trump Took Office, he makes a lame joke about how all the cooling happened under Trump.
Last year, scientists said Earth was record hot and it was Donald Trump’s fault. But they had it backwards. The prior warming was Obama’s fault, as Trump wasn’t in office yet. The cooling since 2016 is under Trump’s watch.
Not that it matters but this isn't really true. Most of the cooling happened before Trump was elected. Temperatures went up after his election.
Yet it's strange that whilst every one is complaining about the press not noticing or caring about the greatest two year cooling event
, even the sceptical websites failed to report on this momentous event until someone else pointed it out to them.
Quick update, as I'm pressed for time and have a few more interesting post in the works.
The Goddard Institute for Space Studies have released the update for their GISTEMP data set. April 2018 was 0.86°C above the 1951-1980 base period. Only slightly down on last month's 0.88°C. This makes 2018 the 3rd warmest April in the data set, with the last 3 years being the three warmest Aprils on record.
The RSS satellite global temperature for April 2018 shows a slight drop on March. Compared to UAH satellite data, there has been slightly more movement in RSS data, with a range of around 0.1°C compared with UAH's 0.05°C, but overall the picture is very similar so far, with temperatures being slightly below the long term trend, but no sign of the big reversion seen after the 1998 El Niño.
RSS has April 2018 at 0.45°C above the 1979-1998 base period, compared to 0.54°C for March. This makes 2018 the 6th warmest April in the RSS data set.
The trend since 1979 is 1.93°C / century.
The prediction for 2018 shows a slight drop from last month. My simple statistical analysis suggests 2018 will average 0.527 ± 0.118°C, compared with last month's prediction of 0.541 ± 0.136°C.
The average for the first 4 months is 0.506°C, so this will require warmer temperatures to develop later in the year. As with other data sets this prediction is based mainly on the assumption that temperatures will tend towards the trend, not on any specific predictions of weather conditions.
On this basis, there is a good chance (~60%) that 2018 will be the 6th warmest on record. There's a small chance (~7%) that it will finish below 6th, though if it does it could fall someway down the rankings. There's a reasonable chance (~30%) that it could be 5th or better, though almost certainly not 1st.
| Rank | Year | Anomaly | Probability | Cumulative |
|---|---|---|---|---|
| 1 | 2016 | 0.75 | 0.02% | 0.02% |
| 2 | 2017 | 0.64 | 3.11% | 3.13% |
| 3 | 2010 | 0.58 | 14.88% | 18.01% |
| 4 | 1998 | 0.58 | 0.21% | 18.23% |
| 5 | 2015 | 0.55 | 13.61% | 31.83% |
| 6 | 2005 | 0.43 | 61.93% | 93.77% |
| 7 | 2014 | 0.43 | 1.02% | 94.79% |
| 8 | 2003 | 0.39 | 3.95% | 98.74% |
| Below 8th | n/a | n/a | 1.26% | 100.00% |
Both satellite data sets are predicting 2018 will probably be around 6th warmest, whereas surface data sets are tending towards 4th warmest. The difference between the two is mostly to do with the satellite data showing 1998 and 2010 as being much warmer than in the surface sets. In terms of recent history, all data sets are suggesting similar things, that the past 4 years have all been pretty hot, and this is very different to the previous El Niño peaks, that only lasted for a single year.
All this suggests we are not likely to see the return of the old Great Pause
any time soon, and probably ever.
If warming did stop in 1997 it's certainly going again, and had no effect on the long term trend.
It's easy to predict what happens next, and this can be seen in many discussions I've had on WUWT.
1. The claim of a new pause, starting just before the latest El Niño. Last time we had to wait seven years before claims that there had been no warming since 1997, or even that there had been rapid cooling, but now I'm already seeing claims based on the last two years.
2. Some, such as Lord Monckton, will deny they ever thought the pause was important and will insist that the real point is that temperatures may not warm as much as expected, and that warming is a good thing.
Roy Spencer announced the April update for UAH Global Temperature, and it continues the theme of being not too interesting.
According to the UAH satellite data, April was 0.21°C above the 1981-2010 base period. This compares with 0.24°C last month. This makes 2018 the 7th warmest April since the start date of 1979.
The trend over the entire data set for the months of April is slightly less than the trend for all months at 1.17°C / century. Or if you prefer, Tony Heller would argue there has been no trend in April temperatures in 20 years.
Looking at all months the trend continues at 1.28°C / century, substantially slower than all other data sets, with this month again being just below the trend.
The mid-point prediction remains unchanged from last month, but with slightly better confidence. My prediction, based purely on simple statistical analysis is that 2018 will be 0.243 ± 0.117°C. This compares with last month's prediction of 0.243 ± 0.136°C.
The current average for the first four months is 0.228°C, so the prediction requires temperatures to rise slightly over the rest of the year. This is simply because so far 2018 has been below the trend line and there is an expectation that it is more likely temperatures will move towards the trend.
As far as rankings are concerned, 2018 is most likely (~60%) to finish either 5th or 6th warmest, but more likely to finish below 6th, than above 5th. There's a small chance (~5%) that 2018 will finish 4th or 3rd, and virtually no chance of finishing 2nd. There's about a 1 in 3 chance it could finish below 6th place, in which case it could easily finish anywhere done to 12th place and possibly lower.
Here are the probabilities for each ranking down to 12th.
| Rank | Year | Anomaly | Probability | Cumulative |
|---|---|---|---|---|
| 1 | 2016 | 0.51 | 0.00% | 0.00% |
| 2 | 1998 | 0.48 | 0.01% | 0.01% |
| 3 | 2017 | 0.38 | 1.43% | 1.44% |
| 4 | 2010 | 0.34 | 4.48% | 5.92% |
| 5 | 2015 | 0.27 | 29.11% | 35.03% |
| 6 | 2002 | 0.22 | 32.69% | 67.72% |
| 7 | 2005 | 0.2 | 9.87% | 77.59% |
| 8 | 2003 | 0.19 | 5.81% | 83.40% |
| 9 | 2014 | 0.18 | 3.25% | 86.65% |
| 10 | 2007 | 0.16 | 5.69% | 92.35% |
| 11 | 2013 | 0.13 | 4.33% | 96.68% |
| 12 | 2001 | 0.12 | 1.65% | 98.32% |
| Below 12 | n/a | n/a | 1.68% | 100.00% |
A brief update for GISTEMP, whilst I try to get some more interesting posts completed. I'm only updating a few data sets a month for now, as they are mostly pretty similar.
GISS Surface Temperature Analysis (GISTEMP) has March 2018, at 0.89°C above the base period of 1951 - 1980. This is 0.1°C warmer than the anomaly for February. It makes 2018 the 6th warmest March, with the last 4 Marches being 4 of the 6 warmest on record.
As is something of a theme so far for 2018 across all the data sets, March was close to the trend since 1970.
The average anomaly for the first 3 months of 2018 is 0.817°C. My forecast for 2018 is 0.802 ±0.166°C, which is almost identical to the forecast from February.
This would put 2018 between the big gap between 2014 and 2015. There is around a 74% chance of 2018 being the 4th warmest, and a greater than 99% likelihood of being between 7th and 2nd warmest.
| Rank | Year | Anomaly | Probability | Cumulative |
|---|---|---|---|---|
| 1 | 2016 | 0.99 | 0.13% | 0.13% |
| 2 | 2017 | 0.89 | 5.95% | 6.08% |
| 3 | 2015 | 0.86 | 8.70% | 14.79% |
| 4 | 2014 | 0.73 | 73.83% | 88.61% |
| 5 | 2010 | 0.7 | 7.10% | 95.71% |
| 6 | 2005 | 0.68 | 2.30% | 98.01% |
| 7 | 2013 | 0.64 | 1.56% | 99.58% |
Apologies for not updating this blog much recently. I've been having more fun commenting on Watts Up With That, which I expect is a better use of my time as people actually read that blog. I want to write some posts about some of the discussions there, but for the time being lets get back to charting global temperatures. Note, I've skipped February's updates - none where particularly interesting being mostly slightly cooler than January, and this early in the year making little difference to the annual predictions.
The global temperature anomaly according to The University of Alabama in Huntsville was 0.24°C, compared with the 1981-2010 base period. This is the equal 6th warmest March in the data set.
This is a slightly higher anomaly than February, which was 0.2°C. The start of 2018 continues to be slightly below the trend since 1979. There's still little sign of a big La Niña drop, as happened after the 1998 El Niño.
My statistical prediction based on the 3 months so far, and the trend is that 2018 will have an average anomaly of 0.243 ± 0.136°C. This is up 0.01°C since last month.
An anomaly of 0.243°C would make this the 6th warmest year on record. The probability spread suggests there is about a 28% chance of finishing in 6th place (ahead of 2002), and a similar chance of finishing 5th, ahead of 2015. However, there is only a slim chance (less than 10%) of finishing 3rd or 4th, and virtually no chance of 2018 being the 2nd or 1st warmest. There's about a 35% chance of finishing cooler than 2002, in which case it could finish just about anywhere as there are so many years with similar temperatures.
How much credence do I attach to these predictions? If I'd posted last month I'd have said not much. The predictions at this stage are heavily dominated by the trend and it seems likely that we are overdue a below the trend year, especially given the possibility of a La Niña. However the small rise this month and the fact during the first quarter of the year anomalies have remained close to the trend, makes me slightly more confident that this prediction method might not be too far of this year. I still expect it's more likely that we will finish below the mid-point prediction, but maybe not so far off.
All this will be bad news to those yearning for a return of the so-called pause. At present even the lowest trend using Christopher Monckton's cherry picking exercise is for a trend of 0.72°C / century starting in July 1997.
