In the HOME benchmark the inserted inhomogeneities had a standard deviation of 0.8°C, while the detected inhomogeneities had a standard deviation of 0.94°C. Thus this difference between actual and detected inhomogeneities is important. In Peru, when only the largest inhomogeneities could be detected the difference may be even larger.

Compared to Peru this is a factor 20 (40) higher station density for temperature (precipitation).

The distance dependence of the spatial correlations are also not that high. So you do need a clear difference in station density to get an effect. The number of stations and thus sometimes references may have some influence as well.

Also important is that this by itself also reduces the correlations between the stations and thus makes homogenisation and QC harder.

I am the first author of this paper.

For some weaknesses of this study and how to improve the benchmarking, see this blogpost of mine. http://variable-variability.blogspot.de/2014/06/problems-with-home-benchmark.html

I would call these methods multiple-breakpoint methods. PRODIGE and ACMANT use joint correction, MASH does not.

Which Lindau and Venema (2013) showed could be improved.

It finds the optimal solution with respect to the explained break variance for a given number of breaks. Whether this is also the optimal solution for homogenisation is under debate; see Lindau and Venema (submitted, 2017).

At least in climatology, joint detection is more recent. HOMER in 2013 was the first implementation.

At least I would define joint detection as the detection of break on multiple series simultaneously, which can be distinguished from multiple-breakpoint methods, which detect multiple breaks in one series simultaneously.

In my view HOMER is not HOME recommended. We made these recommendations in part based on the performance in the HOMER benchmark. HOMER did not participate in the HOME benchmarking, it did not exist yet.

One could argue that when HOMER is operated like PRODIGE it is a HOME recommended method. Operated automatically with joint detection it regularly produces problems; see for example Gubler et al. (2017).

No evidence. One government would not be enough. There are climate scientists in many countries all over the world.

Also without computer models we would know it is warming and that the warming is due to human activities. For example, we see the influence of CO2 in the climatic changes in the deep past. We measured how the global temperature responds to volcanic eruptions.

One of the most bizarre conspiracy theories ever created. A conspiracy over decades on the central tenets of a natural science spanning thousands of scientists in many countries and universities would be impossible to pull off.

Other conspiracy theories at least try to provide some evidence, suggestive material. This piece just makes an empty claim.

The actual warming estimate clearly shows continued warming.

"Truth And Action" (TAA) does not report on these mentioned tweets and emails, that purportedly would provide some evidence for the claims in this article and in the TAA article.

P.S. TAA also "reported": FBI Publishes Report Stating Nobody Died At Sandy Hook.

One of the most bizarre conspiracy theories ever created. A conspiracy over decades on the central tenets of a natural science would be impossible to pull off.

I do. I do not work in America. (And naturally my American colleagues also follow the evidence.) I would not be in science if the evidence would not decide what it true.

This is wrong. This is not how decisions on science funding are made. It would be impossible for any American to do this, even if they pervert the process, because a large part of climate research is performed outside of the USA. Under left, right and whatever governments.

Trump is in power now, where are all the credible scientific papers that John Coleman myths are right?

Let him write a scientific article and explain in detail where all the other scientists went wrong for all the different independent lines of evidence that the climate is changing and all the different independent lines of evidence that CO2 emissions are the main reason.

It used to be part of the Republican party platform. Conservatives parties all over the world accept that climate change is a problem. Scientists from all countries work on it. All countries are part of the Paris Climate Agreement, only the Trump administration has pledged to leave the agreement.

John Coleman may not agree with it, but there are reasons to expect warming. In fact, nearly all scientists do.

His friends like to talk about the (little) ice age and that the climate is always changing. That is at least factually true, while no argument that this time the warming cannot be due to human activities.

I guess there were also no tropical forest on the Arctic island Svalbart.

The warming is due to human activities.

I beg to disagree.

A wide range of observations showing the climate is changing; taken from the last IPCC report.

Science is a process to understand reality better and better.

If there is a clear consensus understanding of something one can call it a "fact" and it is an acceptable term in daily live, but not really well suited to describe science where knowledge is always provisional.

It is an ironic that in the same paragraph Coleman claims consensus does not exist in science, he does claim facts exist. Which is a more wide-ranging claim.

There is no consensus on the theory of gravity? Of course that exists.

The Earth is big. Time matters for an explosion. This is comparable to 0.000,000,01 bombs per meter per second.

It would be exceedingly difficult to manipulate the global temperature estimates significantly without scientists noticing.

The start of the second world war also changed the composition of the fleet making the sea surface temperature observations and with it the measurement methods used changed quickly. Climatologists have made adjustments to account for this problem, but it is well possible that this was incomplete. Uncertainties in this period are larger than in the decades before and after.

Thus it could well be that these three records in a row are due to remaining problems. I would not have emphasised it.

I do not recall any climate scientist making that claim. It may happen, but I am unaware of reasons why it should be more likely than otherwise to happen.

Where the Arctic ocean freezes the air can get very cold because the ice blocks the heat transport from the relatively warm ocean water. The temperature variability is naturally larger in the Arctic than in the mid-latitudes (USA, Europe).

Thus it was very warm in the Arctic. As Gavin Schmidt said:

“What’s going on in the Arctic is really very impressive; this year was ridiculously off the chart”

But 20 to 30 degrees Fahrenheit may not be as exceptional as it sounds to our ears.

If people refuse to solve the problem and civilization goes down, it should be noted that the rich have most to lose.

And this mostly just in the USA. In the rest of the world conservative parties accept the the science of climate change.

We are sure that natural factors alone could not have caused the observed warming, without the increase of greenhouse gasses there would not be so much warming, but they could have contributed.

For the period 1951 to 2010 our best estimate is that all of the warming was due to human activities. But natural factors could have warmed or cooled the Earth a bit. In the period around 1900 part of the warming was probably natural due to less volcanoes and a stronger sun.

It would have been more accurate to say that we can see that the CO2 increase is from burning fossil fuels by measuring isotopes, differences in the number of nuclear particles of atoms. Not all relevant isotopes are radioactive and the ones that are only a little.

Also important is that we have good estimates of how much fossil fuels we have burned, from which one can estimate the CO2 emissions. The increase in CO2 concentration in the atmosphere is about half of those emissions (the rest was taken up by the vegetation and the oceans).

Note that the warming over land is about twice as large as the warming over the ocean. Over the ocean more of the additional heat goes into evaporation of water rather than warming of the air. It furthermore takes time to heat up the oceans, just like a water kettle takes time to boil.

Because there is more land in the Northern Hemisphere it will warm more than the Southern Hemisphere.

A way to see how different two degrees are is to look at a region more towards the equator that is two degrees warmer. Nature will look very different.

In that case the 1981-2010 average should be zero, which it clearly is not. Likely, like the other NOAA figures on this page, the values are the departures from the 20th century average.

At the moment it is more likely than not that 2017 will become the second warmest year of the instrumental period. The warmest was 2016 with El Nino. So this claim is correct.

2016 would most likely also have been a record warm year without El Nino. Given the long-term warming trend due to global warming it was to be expected that 2017 would also be warm.

The rate of warming has been about the same since 1980.

The warming is only exceptionally strong if you were deceived by people claiming that global warming had/has stopped.

Right.

The temperature has dropped a bit.

It may be too early to say the industrial emission have peaked, but the last three years they were more or less stable. This was in large part due to strong emission reductions in China.

Most early scientists?

I would expect most see emission reductions as important and other options, like massive carbon capture, as quite unrealistic. My impression is that most are hoping geo-engineering will not be necessary.

I hope the author is not suggesting that aircrafts will not be able to fly in 2100. They also fly in hotter countries.

Why not simply use another energy source? The energy sector is about 6% of the economy. Even if alternative energy sources would be more expensive (and the difference is no large any more), I fail to see how that would be a major problem. Especially if with "steady state" these scholars are thinking of subsistence living.

The main problem is burning coal.

If this author already writes "presumed result", I would love to see a reference to the scientific literature. It would be quite surprising if the 1-1.5 °C warming we have seen up now makes such a differences.

Data Berkeley Earth

For context the reader needs to know that the seasonal cycle in temperature is much smaller in tropical countries than in the USA/mid-latitudes.

That also illustrates that many climate impacts are not about climate change alone. In this case people so poor that they are forced to work when it is too hot.

The climate just is.

Mitigation sceptics give the impression they are annoyed the climate does not negotiate with them, that is similar anthropomorphizing the climate as thinking it will make war on us.

Not knowing who the author talked to, selected to talk to, I cannot be sure, I would be highly surprised about this and it does not fit to what I and the colleagues I know are thinking.

It is easier to keep our climate habitable than to make the climate of another planet fit for human habitation.

It also reflects sun light. Clouds are the largest uncertainty in climate projections. Their effect will depend on cloud height, type, droplet concentration, region and the time of day.

Uncertainty goes both ways, mitigation sceptics are wrong to assume it will always be better, this article is wrong to assume it will be worse.

That is for the RCP8.5 scenario, the one mitigation sceptics are fighting to realise. In the light of the Paris agreement and recent price drops of renewable energy production and storage, this scenario has become less likely.

Absent aggressive action or absent action? Does the author mean the world we would get if the mitigation sceptics win politically and we burn as much fossil fuels as possible ignoring the costs?

Does the author see the pledges of the Paris Agreement as "aggressive" action? Or does the author think of a WWII mobilization as "aggressive" action and the outcomes described here also valid for a Paris world?

What "aggressive" action is makes a huge difference. Hopefully every reader has the same idea of this term than the author.

If carefully selected this can produce a wide range of outcomes the author would like and does not have to give a fair overview of the scientific state of the art.

Agree with the comment of pneff/Michael Mann.

Would like to add that this twice as fast is just for a cherry picked period. That is about as bad as mitigation sceptical claiming that global warming has stopped by cherry picking a specific period.

The real message was that satellite upper air warming estimates are unreliable (still are), but seem to confirm what we know from the surface warming.

This story was thus not alarming. If anything is was good news, it reduces uncertainties and as this uninhabitable-article already suggests: uncertainty is not our friend.

Does the author really want to suggest scientists expected the permafrost to stay frozen because it is called permafrost? Scientists are generally not stupid and a lot of work is done on permafrost. Work from which the numbers in the rest of the paragraph come from.

Regions with open water will stay relative warm due to the relatively warm water. As soon as ice limits heat flows from the ocean to the atmosphere it can get very cold. Also in general variability is much stronger in the Arctic, what sounds like a really big warm anomaly to an innocent reader in the USA is thus exceptional, but less so than the numbers suggest. That is context a reader would need to know.

That would be a design or construction error of the seed vault. That the Arctic would warm faster has always been known.

Depending on how you define "uninhabitable", parts of the Earth are already uninhabitable. Conveniently, they are the parts where nearly no one lives.

It is a pity the report only showed the average of UAH and RSS. Seeing both datasets, including some previous ones, had shown how large the differences are.

Armchair logic.

No idea why one would think this. Climate "sceptics" like to claim the raw data shows too much warming due to urbanization. (Detailed studies show this effect to be small for the global average warming.) If that had been the main problem, the adjustments should have had to reduce the warming and not " sometimes get warming and sometimes get cooling".

Why did you not show the new RSSv4? The large adjustments made to the satellite temperature could be seen if you have not only shown UAHv6, but also UAHv5.

I fail to find any analysis of the sort.

It is kind of ironic you like the upper air satellite temperatures so much. They are the dataset with the most adjustments.

Details about this graph can be found on the homepage of RSS.

Remember you wrote something in the beginning about a "tropical hotspot". That means that the upper air temperature and the surface temperature should not be the same. A hint is that they have different names.

You did no such thing. You did not even explain how the data adjustments were made, just complained you did not like the outcome.

That you personally would prefer the world to have more cyclical patterns is no reason why the world should have more cyclical patterns.

If you worry about the outcome, show where the methods are wrong. Just like scientists do.

The weather service of New Zealand made such a beautiful webpage explaining the reasons for the adjustments in so much detail. A pity you do not cite them.

Nice to see a citation of the scientific literature. Even if indirect and 30 years old.

That urban station have additional warming when the city around them develops is not disputed. The question is whether it matters for the global mean temperature. While many people live in cities only a small part of the Earth is urban. Furthermore, urban stations also tend to be relocated often to newer less urban locations. The empirical question is thus whether the original location was more urban than the current one.

There have been many studies, looking at the problem in many different ways. They show that the effect is small when it comes to the global mean temperature. Regionally, for example China, it does make a difference.

:-)

No way. NOAA does not make such unprofessional messy graphs.

Source: some blog of people making claims I like.

Ever heard of the Dust Bowl in America? (America is just 2% of the Earth's surface.) Hopefully the Dust Bowl is not cyclical.

No the reduction is not "targeted" to the WWII. It is a consequence of the different measurement methods for the sea surface temperature used during the war.

The methods are the same everywhere, the war caused a data problem.

A nicely red graph. That color is determined by the arbitrary zero point and only suggestive, not informative.

The difference between the two datasets is about constant, does not affect the trend much.

The difference in the trend is very small. The green line is similarly above the red line for most of the period.

The graph is about as red in the beginning as at the end. The long-term trends is just not affected much. Also look at the y-axis and compare this to the warming over this period of about 1 °C.

If the reason are adjustments, a likely reason would be the adjustments of the sea surface temperature that make global warming smaller. Older bucket measurements are cooler than modern buoy measurements because the water cools by evaporation between taking the bucket out of the water than reading the thermometer. In the past these adjustments were thought to be larger than modern estimates.

This could also be due to increases in the amount of data used and thus also increases in coverage. Data is especially scarce in the Arctic where the warming is strongest.

So the "scandal" here would be that scientists were conservative in their early estimate of global warming.

Hansen wrote about the absolute temperature: The "Elusive Absolute Surface Air Temperature (SAT)" is the title of the page.

The figures in the rest of this report are temperature anomalies used to study whether the temperature at a certain location is increasing.

Estimating the absolute temperature is difficult because it changes quickly in space, is very different on a mountain than in a valley. In a hot year both the mountain top and the valley are warm. The anomalies vary much less in space. Warming is thus easier to estimate than the absolute temperature.

Has been studied by the climate "sceptics" of BEST. Does not make a difference, as expected.

For the percentage of airport stations a scientific reference would be nice. The value is likely much too high. Airport stations actually show less warming. These stations were previously often in cities or villages and the airports have less urban heat island than cities.

Airports are also mostly very small and good places to measure, the ones everyone knows are the big ones.

These potential problems have all been studied. Studies the authors apparently did not read, did not engage with or at least did not cite.

It is evidence. If the three datasets would be hugely different, such as the two upper air temperature satellite datasets, I am sure this report would use that to claim there are large uncertainties.

And yes, it does not preclude that there are problems that are not taken into account in any of the three datasets. The difference is not an estimate of the uncertainty.

Uncertainties can go both ways. The uncertainty monster is not our friend.

Those are platitudes that can be written about any piece of science. Science keeps on making progress in understanding the world around us.

Feel invited to do it better. Everyone is.

That is no longer current. For its land data, GISS now uses the adjusted data of NOAA. As a consequence the adjustments GISS made for urbanization nowadays are no longer needed and hardly make any change any more. The main difference is thus the way the global mean is computed.

CRU gathers data themselves. This should be know among "sceptics". There was an action of the mitigation sceptics to harass Phil Jones of the UK Climate Research Unit (CRU) by asking for the data and contracts with the weather services for five random countries.

Naturally these are to a large extend the same stations and the data from the same stations is naturally largely the same.

GISS depends on NOAA data (and a few other sources). The other groups gather data themselves and naturally also include NOAA data for stations they do not obtain otherwise.

BEST analyzed over 30 thousand stations, The ISTI has a similar number. NOAA's GHCNv3 only has about 8 thousand.

The report ignores the monthly datasets by BEST, the ISTI, JMA and the University of Delaware. Plus all the other climatic changes that can be used to estimate warming.

Monthly temperature datasets. There are many more for daily and sub-daily data.

A hypothesis must be formulated clearly so that it could be falsified. "Sufficiently credible" "for policy analysis purposes" are much too vague terms.

No, the evidence is not that fragile. There are many ways to estimate how much the Earth has warmed.

Also without climate models we would understand the greenhouse effect, observe the increases in CO2 and methane and observe warming. Also without climate models we would notice that we can only explain past changes in the climate by considering the influence of CO2.

Climate models are helpful in reducing the uncertainties, adding regional details, but are not necessary to make the basic case that global warming exists and is due to us.

More uncertainty increases risks of global warming. There are bound to be surprises when we take the global climate outside of known territories. The uncertainty monster is not our friend.

Wrong. The tropical hotspot (faster warming in the upper air than near the surface) is expect for warming of any cause, not just for warming due to greenhouse gasses.

Also without any thermometer it would be clear that the Earth is warming. Just look at sea ice, sea level rise, melting glaciers, later river freezing, earlier ice breakup, earlier spring and animal migration, movement of animals, plants, insects and ecosystems towards to poles and up mountains, etc.

These chances can be used for independent warming estimates. See for example this one for glaciers or this study estimating warming via the time rivers and lakes freeze up.

1. The adjustments are scientifically important, but rather modest for the global mean temperature.
2. The adjustments make the estimated warming smaller, not larger.

These are for example the raw and adjusted warming signals of the NASA GISS dataset. The how and why is discussed in this post.

Global warming is not US warming or New Zealand warming. Surface warming is not upper air warming.

The most important dataset this report forgets to mention is the one of Berkeley Earth (BEST project). A Red Team project partially funded by the Koch Brothers led by conservative physics professor Richard Muller.

This was a group of outside scientists who looked at the same objections of this report:

We carefully studied issues raised by skeptics: biases from urban heating (we duplicated our results using rural data alone), from data selection (prior groups selected fewer than 20 percent of the available temperature stations; we used virtually 100 percent), from poor station quality (we separately analyzed good stations and poor ones) and from human intervention and data adjustment (our work is completely automated and hands-off). In our papers we demonstrate that none of these potentially troublesome effects unduly biased our conclusions.

And found the main datasets to be accurate:

Our results turned out to be close to those published by prior groups. We think that means that those groups had truly been very careful in their work, despite their inability to convince some skeptics of that. They managed to avoid bias in their data selection, homogenization and other corrections.

The main problem of this report is that is just repeats decade old claims, but does not engage with the arguments in the scientific literature, for example those of the BEST Red Team.

The report wants to see the summer heat waves of the Dust Bowl in America in the 1930s in the global average annual mean temperature.

The figure shows the summer temperature difference between the Dust Bowl year 1936 and now. The affected area, which was warmer in 1936 than now (blue), is about 2% of the Earth's surface.

Hopefully the Dust Bowl is not a cyclical pattern.

The entire study could not have been more wrong in three clear ways.

1 The adjustments are scientifically important, but rather modest for the global mean temperature.

2 The adjustments make the estimated warming smaller, not larger.

These are for example the raw and adjusted warming signals of the NASA GISS dataset. The how and why is discussed in this post.

3 Also without any thermometer it would be clear that the Earth is warming. Just look at sea ice, sea level rise, melting glaciers, later river freezing, earlier ice breakup, earlier spring and animal migration, movement of animals, plants, insects and ecosystems towards to poles and up mountains, etc.

No idea why one would think this. Climate "sceptics" like to claim the raw data shows too much warming due to urbanization. (Detailed studies show this effect to be small for the global average warming.) If that had been the main problem, the adjustments should have had to reduce the warming and not " sometimes get warming and sometimes get cooling".

The second world war increased the percentage of American navy vessels, which make engine intake measurements, and decreased the percentage of merchant ships, which make bucket measurements.

(Figure from UK MetOffice)

Engine intake observations are typically warmer than bucket measurements because the water cools due to evaporation before the thermometer is read. That produces a spurious warm peak in the raw data.

That amounts to the claim that the Earth is not really warming. Even without looking at any thermometer data, even if we would not have invented the thermometer we would know it is warming

Glaciers are melting, from the tropical Kilimanjaro</a> glaciers, to the ones in the Alps and Greenland. Arctic sea ice is shrinking. The growing season in the mid-latitudes has become weeks longer. Trees bud and blossom earlier. Wine can be harvested earlier. Animals migrate earlier. The habitat of plants, animals and insects is shifting poleward and up the mountains. Lakes and rivers freeze later and break-up the ice earlier. The oceans are rising.

This chart does not show the adjustments. It shows an unsourced graph with the difference between two datasets. This is also due to increases in the amount of data used and thus also increases in coverage.

In as far as the reason are adjustments, a likely reason would be the adjustments of the sea surface temperature that make global warming smaller. Older bucket measurements are cooler than modern buoy measurements because the water cools by evaporation between taking the bucket out of the water than reading the thermometer. In the past these adjustments were thought to be larger than modern estimates.

So the "scandal" here would be that scientists were conservative in their early estimate of global warming.

The differences between the raw and the adjusted data are small for the global mean temperature. The warming estimated from adjusted data is smaller than it would have been without taking changes in the way observations were made into account.

Global warming is something else as US warming or New Zealand warming.

The temperature increase is not a peak. The warming will continue if we do not do more to solve the problem. We determine how much further. Without geo-engineering the CO2 concentrations, the warming and sea level rise will stay will us for millennia.

This is wrong. The IPCC reviews the scientific literature. In my field, station temperature data, the review accurately represents the scientific literature. Several of my colleagues reported the same. These feedbacks suggest that if anything the IPCC has a tendency to present a too rosy picture of climate change.

John Christy provides no evidence that the discrepancy is due to the models and not due to errors in his dataset; see below. Or due to errors in the comparison.

Even if the reason were the climate models themselves, his claim that models are thus too sensitive to greenhouse gases is a second step into evidence-less territory. Without given any evidence for it, Christy seems to assume that the reason cannot be any other factor such as the historical increases in small airborne particles (aerosols) or historical land-use changes or how much the ocean delays warming.

For all the details provided, it is unfortunate that Judith Curry does not simply state that a warmer world will mean higher sea levels.

I would encourage Mr. Brooks to at least mention the name of his information and preferably add a scientific article to the Congressional record that would make the adventurous claim that sea level would go down when the Earth is warming. That goes against what we have observed in the past and against our understanding of the physical processes responsible for sea level changes.

Like I mentioned in more detail above, Christy provides no evidence that the discrepancy is due to the models and not due to problems with his dataset. If it were due to the models, he also provides no evidence that the reason is the sensitivity due to greenhouse gases.

There is clear evidence that weather extremes are getting worse. Especially heat waves and severe precipitation.

Roger Pielke has chosen to focus on four particular ones. For the evidence on hurricanes see James Elsner below.

Floods are not just influenced be precipitation, but also be changes in land-use upstream, in settlement in flooding areas and in flood defences. It is clear that severe precipitation is increasing in many areas and flood defences will thus have to become stronger and more expensive to keep the risks the same.

Droughts are rare and hard to measure, we do not have much high-quality long-term measurements of soil moisture. That the evidence is thus not as strong as one would like is no reason not to plan for increases in droughts especially in Mediterranean climates.

I am unaware of a consensus on changes in tornadoes due to climate change having ever existed. Thus I do not understand why Roger Pielke mentions this.

This sentence is correct.

This pattern is a fingerprint, it is unique for every possible cause of warming. That good match thus shows that we understood the warming due to greenhouse gases already quite well in 1989.

Exactly, it is sad that this needs to be stated in the US of America in 2017.

John Christy just made an update of his upper air temperature dataset, which changed the warming enormously.

https://twitter.com/klimaatVeranda/status/817121281267077121

The updated UAHv6 dataset of John Christy is quite an outlier. https://twitter.com/hausfath/status/846799655321067520

Originally he was even claiming that the upper air temperatures were cooling, which was due to large consequential errors in how he computed the warming.

In the light of the continual large changes John Christy makes to his dataset, it should be noted that he provided no evidence that the difference between models and his estimate of the warming is due to the models. I would argue that the "scientific method" indicates that one should first understand the problem before making strong claims.

The temperature of the upper air shows much more weather noise (due to, for example, El Nino) than the surface temperature, which has less spread.

Uncertainty is not our friend, it makes the risks of climate change larger.

The surface temperatures are expected to be more reliable, also by scientists who produce a satellite upper air temperatures dataset such as Carl Mears of RSS. The comparison of models and surface observations are very good, especially considering that these climate models are not made for decadal climate prediction, but for long-term projections.

https://www.climate-lab-book.ac.uk/comparing-cmip5-observations/

John Cristy has shown this plot for many years on his blog and in Congressional hearings. If it were a robust result, it would be scientifically interesting, however in all these years Christy has not published a scientific article on his claims and colleagues have had to reverse engineer how it was made.

One think that is clear is that due the 5-year averages the strong warming in the last 3 years is strongly suppressed in this plot. Due to the averaging of the balloon datasets and the satellite datasets, the large uncertainty in the observations is no longer visible. It is not good practise to align measurements and models by selecting only one year that happened to have been warm; to reduce the influence of such an arbitrary choice scientists normally use a longer period to align different datasets. Several of these problems have been discussed at RealClimate, but unfortunately Christy has not updated his graph.

A more honest graph would look like this (peak warming in 2016 not shown): http://www.realclimate.org/index.php/archives/2016/05/comparing-models-to-the-satellite-datasets/

Climate models inform us about long-term warming and have not been developed for short-term predictions, which are strongly influenced by fluctuations from, for example, El Nino. The timing of El Nino is not predictable more than a year in advance. This problem is even more the case for the upper air temperatures shown here by John Christy than for the surface temperature. The uncertainty of climate models results for short-term prediction is expected to be about twice the model spread of the CMIP models.

The scientific consensus on climate change is due the evidence that has been gathered. This was gathered over a period of much more than two decades.

There is nothing more beautiful for a scientists that to have strong arguments against a consensus. In the light of the overwhelming evidence that man is changing the climate and the multiple independent lines of evidence toppling the consensus will be hard and woudl thus be rewarding. Please, note that Judith Curry does not provide any such evidence.

Judith Curry provides no evidence that the larger uncertainties in this period are due to climate models.

Uncertainty is not our friend, it increases the risks of climate change.

Climate models can be validated by looking at how they perform for the state of the climate, recent changes and changes observed in the deep past.

The real world is complicated and one can claim for any personal plan and government policy that its future impacts cannot be "rigorously evaluated". That normally is not seen as a reason for inaction; people still get out of bed and do the hard work to make their future better.

Also without climate models it would be clear that global warming exists. There would be more uncertainty and uncertainty goes both ways, it could be better, but also worse. Thus without climate models, the risk of climate change would be larger and stronger action would be necessary to keep the risk the same.

Without climate change we would know less about the changes in the water cycle and in regional climatic changes. This would thus need more funding for adaptation measures to protect communities because a larger range of possible outcomes would have to be planned for.

To members of the scientific community it should be obvious that there is a clear agreement in the scientific community that greenhouse gases are the dominant cause of global warming.

This kind of evidence-free statements that there is no such agreements has led to consensus studies that consistently find large agreements on such basics among scientists and scientific articles.

This first sentence is still correct. The Earth has warmed by about 1°C since the end of the 19th century.

http://globalwarmingindex.org/

Lamar Smith kept on asking about data and software that were on the internet and given to them.

Lamar Smith abused his personal subpoena power to also ask for nearly any email ever written. Producing this is a lot of work and seems to have been mainly intended to try to get text snippets that could be misquoted and misinterpreted like in case of Climatgate. This was clearly seen as harassing scientists as punishment for science that is inconvenient to the donors of Lamar Smith.

You should know better, Lamar Smith, politics does not make large decisions based on single scientific articles. It is the large body of scientific evidence on man-made climate change that underlies these policies and treaties.

Scientists are always working on improving the quality of climate data to be able to estimate more accurate trends. This is called homogenization.http://variable-variability.blogspot.com/p/homogenization.html

Also Dr Bates, much praised by the so-called Science Committee, homogenized his satellite datasets. http://orcid.org/0000-0002-8124-0406

Bates has the skills to write a scientific article. He opts not to do so but goes to a UK tabloid the Daily Mail. This should be a warning flag in itself. This has happened over and over again. If any of his complaints would hold up and be consequential he could have written a scientific article.

Lamar Smith nor Bates show that the temperature analysis is wrong. Much ado about nothing.

This is one of the parts he clearly did not understand. The update only brought the NOAA dataset more in line with the other global temperature datasets.

Later analysis with other datasets showed that the update was indeed an improvement. https://www.carbonbrief.org/factcheck-mail-sundays-astonishing-evidence-global-temperature-rise

A recent paper showed that the new sea surface temperature data was an improvement and fitted better to other datasets. http://advances.sciencemag.org/content/3/1/e1601207

One someone calls rushed. The update of the sea surface dataset, which was the main change for the period Lamar Smith finds politically interesting was done in 2013. Then no one complained, only after the Karl et al. paper in 2015. http://icarus-maynooth.blogspot.de/2017/02/on-mail-on-sunday-article-on-karl-et-al.html

This is nowadays just a personal subpoena of Lamar Smith. When this kind of harassment is done on the political friends of Smith, he does recognize that it is harassment and suppression of the freedom of science and freedom of speech to have Congress about its Washington power for its donors. Scientists have complained about this no matter which "side" was harassed by politicians. http://variable-variability.blogspot.com/2015/02/stop-all-harassment-of-all-scientists.html

The "hiatus" political hacks like so much is mostly well-understood year-to-year fluctuations in El Nino. If the influence of El Nino is removed, this apparent so-called "hiatus" disappears. For details, see this article. https://archive.is/nVqgk

Only 4% of Andy Biggs donations come from small donors.https://www.opensecrets.org/races/candidates.php?id=AZ05&cycle=2016

Darin LaHood is funded by the Koch Industries and ExxonMobile. This may have influenced his opinion. https://www.opensecrets.org/politicians/contrib.php?cid=N00037031&cycle=2016&type=I&newMem=N&recs=100

You are misinterpreting his claims. He wanted NOAA to follow his newer procedures, which he used for his satellite datasets. There was nothing wrong with the procedures NOAA followed. Normal science. http://icarus-maynooth.blogspot.de/2017/02/on-mail-on-sunday-article-on-karl-et-al.html

Who was not involved in the science of the surface dataset and clearly did not understand much of it. http://icarus-maynooth.blogspot.com/2017/02/on-mail-on-sunday-article-on-karl-et-al.html

The largest donor of chair Lamar Smith is the oil and gas industry. https://www.opensecrets.org/politicians/summary.php?cid=N00001811

For details on Smith war against science: https://www.desmogblog.com/2016/06/21/exxon-koch-lamar-smith-exxon-knew

"Unacceptably dangerous" is an opinion. If paraphrased right by Bob Ward they were thus expressing their personal opinion. For opinions I prefer to know who said it.

I agree with Chris Colose. If you mention a fraction (a quarter or a third) you need to mention relative to what. The total warming since 1900 is about 1°C. The influence of El Nino on the 2016 temperature was small compared to that.

The difference between NOAA and NASA-GISS is because of the way the Arctic is treated, where much of the warming was in 2016, but where we do not have much measurements. NOAA excludes much of the Arctic, which gives it a cool bias in 2016, which is accounted for by a larger uncertainty.

NOAA. The gray areas are where NOAA's methods do not have enough observations to compute the temperature. Not computing a temperature is similar to assuming these regions warm like the global average, but the Arctic warmed more.

GISTEMP of NASA-GISS uses other mathematical methods and does compute temperatures for the Arctic, which results in a more accurate global average.

Policy Foundation.

Not only warming. El Nino is associated with droughts in some regions and severe rain in others. It thus has strong impacts and is a large focus of climate research.

As far as i know it is not clear yet whether the El Nino / La Nina variations become stronger with climate change? Does anyone else know literature on this question?

Not based on any evidence. Just like the warming peak due to El Nino was higher in 2016 than in 1998, the period after 2016 will be warming than the years after 1998 due to the progress of global warming.

Earth science budget. It could be that Mr Walker also does not know the difference.

This image is deceptive as it only shows part of the data. With all the problems of this dataset, it also shows long-term warming if you look at the full period.

If you want to cherry pick 1998 as begin year, a year in which the temperature was high due to El Nino, it would at least be fair to stop in 2016, not 2015, because 2016 is the year that the temperatures were high due to El Nino. if you do so, there is clearly no "slowdown".

Better would be not to cherry pick a specific period. The long-term trend is clearly warming and statistics does not indicate any decline in the warming rate. Since 2016, I would argue you no longer need statistics to see this. Jus look at the complete graph.

Note Dr Schmidt said 2015. El Nino did start in the ocean that year, but 2016 is the year where the air temperature is most influenced by El Nino. That takes some time.

Dr Gavin Smith, who is head of NASA's Goddard Institute for Space Studies and not any of the NASA divisions, is right. The long-term warming of about 1°C globally and more over land and more in Northern Hemisphere is much larger than the El Nino fluctuations. See graph of the Global Warming Index above.

It is known that El Nino, an osculation in the tropical Pacific is a cause of short-term fluctuations of the temperature and precipitation. This should not be confused with long-term warming. This is clearly illustrated by this graph below that shows the warming of El Nino years, La Nina years (the opposite of El Nino) and neutral years individually.

The graph also takes out volcanoes, which is another source of short-term fluctuations. The graph is from the Texas State climatologist John Nielsen-Gammon.

Thus going from an El Nino to a La Nina some cooling is expected. To make the mislead claim that this is 1°C of cooling David Rose had to ignore a large part of the evidence and select only what suits his narrative.

1.He selects monthly temperature data, which is more noisy than annual average temperatures.

2.In this noisy data he selects the highest and the lowest value. If you do this you amplify the noise, the highest value is more likely than not high in part due to the noise and the low value low due to the noise.

3.Rose not only selected specific months, but also specific variables and datasets. He shows the temperature of the troposphere, several km up in the air, not the temperature at the surface. He shows one specific dataset (out of a dozen) of the tropospheric temperatures, while all other datasets show a smaller temperature drop.

4.The temperature of the troposphere is influenced more strongly by El Nino, that the surface temperature, where we live. This dataset is thus more noisy. It shows about the same long-term trend in warming, but this trend is harder to see due to all the noise.

5.Rose selected only the land temperature, which is only 1/3 of the Earth's surface and thus more noisy than the global temperature itself. And the land temperature is by its nature more noisy. While the temperature in the Arctic is extremely high, the land temperature in Siberia was very cold in October.

6.The dataset Rose selected has known problems. To quote its producer Carl Mears:

The lower tropospheric (TLT) temperatures have not yet been updated at this time and remain V3.3. The V3.3 TLT data suffer from the same problems with the adjustment for drifting measurement times that led us to update the TMT dataset. V3.3 TLT data should be used with caution.

We had 3 record warm years, 2014, 2015 and likely 2016. Also when Bayern Munchen loses a game, they are still be best team in Germany. Also when 2017 is not again a record year, global warming will continue.

The temperature increase can be seen in this graph of the Global Warming Index with data up to September.

It is virtually certain that 2016 will be warmer than 2015. See for example this estimate of the 2016 average temperature based on the data up to up to October.

The temperature estimates come from passive microwave radiometers. These instruments prime task is estimating air humidity for meteorology, weather prediction. Thus most of the these instruments actually flew on NOAA satellites. It should be noted that they were not designed to be used for long-term climate monitoring.

Tony Heller is a blogger who is best known by his pen name Steven Goddard and known for fabrication climate data. Some of these cases are listed on his Wikipedia page. https://en.wikipedia.org/wiki/Steven_Goddard

His claim that NASA fudged US temperature data was assessed by PunditFact as [Pants on Fiire].(http://www.politifact.com/punditfact/statements/2014/jun/25/steve-doocy/foxs-doocy-nasa-fudged-data-make-case-global-warmi/)

He has no scientific publications on his name.

Tim Ball not only denies that humans amplify the greenhouse effect due to our CO2 emissions, he denies the fact that CO2 is a greenhouse gas. A position that even most people who call themselves "climate sceptics" find to be extreme. https://en.wikipedia.org/wiki/Timothy_Ball

There is a report that provides the evidence, by a large international body of scientists called the IPCC. It is quite informative.

Science is consists of both understanding (theory) and observations (measurements). Both cannot do without each other.

Proof is for mathematics and alcohol. Science is not religion, nothing is ever proven. There is a wealth of (empirical) evidence for climate change.

There are many toxic gasses that are odorless and colorless. The best known one is like CO2 also related to combustion: Carbon Monoxide.

Strange side-argument that is unrelated to global warming, but actually, CO2 is toxic, that is why we exhale it to get rid of it. https://en.wikipedia.org/wiki/Carbon_dioxide#Toxicity

Global warming due to the amplifying of the greenhouse effect is, however, the more immediate problem.

This link, expectedly, does not show any observations for the future or anything else that could be seen as explaining what this claim was intended to mean.

There are no observations for the future century.

For the people who do not click on the link, James Hansen's problem is the reverse of Roy Spencer's. For Hansen the Paris climate treaty as strong enough: "no action, just promises". Since then many of these promises have already resulted in political changes.

The satellite record is climatologically very short, which is one of the reasons its trend has a large uncertainty. Furthermore, upper air temperatures are more noisy, because of the larger influence of El Nino. If you see the figure below, the "half" no longer looks that impressive. (The peak in 1998 is El Nino; the figure stops before the current peak due to a strong El Nino in 2015.)

Here it should be remembered that the spread of a large number of climate models (the yellow area) is an underestimate of the actual uncertainty. In case of decadal variation in the surface temperature, the uncertainty is about twice are large as the model spread.

Also the observations have large uncertainties (the blue area). The satellites were launched to measure humidity for meteorology. They were not designed to be used to estimate temperature, and certainly not to do so for climate change research, which places much stronger demands on temporal stability. Changes in the orbit and design of the satellites and sensors make it difficult to compare temperature estimates in time.

The article by Peter Thone and colleagues (2011) thus states:

"Overall, there is now no longer reasonable evidence of a fundamental disagreement between models and observations with regard to the vertical structure of temperature change from the surface through the troposphere"

As a scientists who works on observational data, I would like to stress that even if there were significant differences, that could just as well be due to the observations. The reasons for the differences need to be understood before one can blame the models, the observations or the comparison method.

Different from the global picture, the differences in the tropics are larger may be significant, they may also not be. That is a real ongoing debate.

Nitpick: Also satellites are not global.

The current promises of governments around the world (INDCs) are not enough to stay below a warming of 2°C, but they do make a difference. In the article "Paris Agreement climate proposals need a boost to keep warming well below 2 °C" , Joeri Rogelj and colleagues wrote in Nature this June they state:

"The INDCs collectively lower greenhouse gas emissions compared to where current policies stand, but still imply a median warming of 2.6–3.1 degrees Celsius by 2100"

Sea level rise started before we had good instrumental observations, which makes it harder to see the acceleration. If you look sea level rise for longer periods based on indirect (proxy) evidence, it is clear that the current sea level rise is much faster than it was the last 1000 years

See for example this article on RealClimate on a recent study for the last 2000 years.

It will depend on our actions whether sea level rise will accelerate much more.

This is an article on the expected sea level rise for the coming century.

Natural variability has been studied by science since the beginning. That El Nino fluctuations can give the appearance of a "hiatus" is something scientists have warned about since the likes of Robert Bradley Jr. have started their "hiatus" meme, ignoring all those warnings. Without any need for statistical expertise, this can be seen in the plot below from the Texas State Climatologist. http://blog.chron.com/climateabyss/2012/04/about-the-lack-of-warming/

The classical because-x-did-not-happen-yet-thus-x-will-not-happen fallacy. A fallacy that is especially problematic in case of a problem like man changing the climate, which goes slowly, like an oil tanker slowly changes direction. In addition: We did something. Most of the new installed power generation capacity is free of CO2 emissions nowadays, we are now paying the price for past too slow actions and the consequences of past inaction will still become larger.

These predictions of Peter Wadhams have been opposed by all climate scientists I know of. This picking quotes that are convenient for Robert Bradley Jr.'s narrative while ignoring what most climate scientists say is one of his most used rhetorical tools.

The uncertainty monster is a reason to act faster. Uncertainties how strongly the climate will respond makes the risks of climate change larger. The surprises are what I fear the most. http://variable-variability.blogspot.com/2015/12/judith-curry-uncertainty-monster-high-risk.html

Climate change is the abstract equivalence of dumping your garbage on your neighbours laws and violating your neighbours property rights.

Robert Bradley Jr. probably makes the typical mistake of equating the model ensemble spread with the model uncertainty. If climate models are used for decadal climate prediction, like Robert Bradley Jr. does here, the uncertainty is twice as large as the model spread. For details and explanations see: http://variable-variability.blogspot.com/2015/09/model-spread-is-not-uncertainty-nwp.html

Note 50% of the surface was melting, not 50% of the ice mass. Melting in summer is naturally normal.

The sea surface was 2°C warmer, not the complete ocean (where the 90% of extra heat from the beginning of the sentence was about).

The influence of El Nino was likely not as large as it was in 1998. A large part of the warming was also in the Arctic, as lenae101 also argues below.

The temperature is 1°C warmer than in the beginning of the instrumental temperature record (1880). The 19th century was likely a little colder than the pre-industrial period. This difference is not large and currently being studied, but likely we did not cross the 1°C warming relative to pre-industrial yet.

Only mentioning the deaths in two cities understates the danger of heat waves. The number of people who died in the 2003 heat wave is estimated to be in the tenths of thousands. This article estimates it was 70 thousand additional deaths.

Heat waves are rare hot periods. This does not only depend on the mean temperature ("when a planet warms"), but also on the variability around the mean.

It is theoretically possible that the mean temperature increases, but that the variability decreases so much that heat waves become less common. Thus, if I may nitpick, it is not "just plain logic", but an empirical question. This comment is nitpicking because the evidence suggests that heat waves are indeed increasing.

This reference goes to a New York Times article by the authors of this scientific article. I feel that is not appropriate for this scientific claim.

Normally I would not nitpick, but Michaels calls himself a "climatologist". The (lower) troposphere is not the same as the free atmosphere.

The troposphere is the layer between the surface and the tropopause. Near the surface it has a boundary layer, which is typically about 1 to 2 km thick during the day. The free atmosphere is above the boundary layer.

Scientists always aim to quantify how accurately we know what we know by computing uncertainties. I hope the reader will read the article by John Kennedy of the UK MetOffice one of the most respected scientists in this field. The article does not support the claim of Michaels that there was a "hiatus".

In fact, if Michaels would like to make a big thing out of the small deviations that caused him to see a "hiatus" he should actually claim that the temperature measurements are extremely accurate. Uncertainties make it harder to determine with confidence that a minimal deviation exists.

Lakes and rivers are warming faster than the regional air temperatures.

The Arctic sea ice is disappearing faster than modelled. As well as snow cover.

Precipitation is getting stronger faster than models predict.

Would Michaels conclude from that that it would be prudent to increase the "modeled temperature forecasts" (actually projections)? I would personally argue that we should first understand the reasons before we would do that, but maybe I am too much of a scientist.

One wonders where this number, "a third", comes from. The global mean temperature trend of the tropospheric temperatures (estimated by satellite) is very similar to that of the surface temperature over the short period for which we have satellite data.

It is expected that the tropospheric temperature would rise a bit faster than the surface temperature, but this is a small deviation. Far from the completely false claim of Michaels that the warming is only a third of what it should be.

As in any case, this deviation could be due to the models, the measurements, their data processing or the comparison. One would need to understand the reasons for this small deviation before one can claim that the models would need to be a little less sensitive.

Also before 1950 there was warming due to greenhouse gasses. Part of this warming was compensated by increases air pollution that also comes from burning fossil fuels. However, CO2 builds up, while air polution has a short life time. In the end, CO2 dominates. Before 1950 there was also warming due to natural causes due to less volcanoes and a stronger sun.

After 1950 we can say that our best estimate is that all warming was man-made, but also before 1950 part of the warming was man-made. All forcings (physical causes that change the temperature) are shown in this graph from the latest IPCC report.

In fits and starts

There are also other reasons why the global mean temperature changes from one year to the next, such as volcanoes and El Nino. You see these changes on top of the long-term trend due to increases in greenhouse gases.

It is strange that Michaels emphasises this, because that is exactly the main reason why his idea that there was a "hiatus" in man-made global warming due to greenhouse gasses is so terribly wrong. Does anyone see a "hiatus" in the graph below? You no longer need statistics to tell you that this meme was wrong.

I would not call the temperature increase "slightly". But am happy to let the reader decided on this matter of taste by simply showing the graph rather than Michaels' complicated description.

It is heart warming to hear Mr Taylor describe the temperature reconstructions of scientists such as Michael Mann as a "reliable indicator".

The UAH6.0 dataset has not been published yet and scientists have thus not had the opportunity to check these results. UAH6 used very different methods to estimate the temperatures and having a much smaller recent trend than the previous UAH5 dataset.

An awkward choice to use UAH6.0 to contrast to the well vetted and more reliable trend estimates of the surface temperature by six different groups world wide. A balanced journalist would at least also show these results.

The last record of the satellite tropospheric temperatures in 1998 was in the year after an El Nino. This may well be the case again. Climate predictions on the short-term are difficult, but it is expected that 2016 will again be a record year.

One wonders if James Taylor has his mansion 4 km up in the air, which is where this satellite estimate comes from. It is a pity that the author does not indicate why this temperature should be the arbiter of everything, while it has no impact on human society.

Wrong, 2015 was the hottest year in the global instrumental record.

When "journalists" think the understand science better than scientists and feel the need to call them names that does not bode well for the quality of the information coming.

Next to these four groups, there is also the privately funded Berkeley Earth dataset and the Cowtan & Way dataset. They all show similar results.

Exactly, before you may have needed good statistics to protect yourself against the human tendency of being a greedy pattern recognizer. Now it is hopefully seen by eye for everyone that there never was any trend change and that it is a good idea to use objective statistics.

Selecting to start your period at a peak is a terrible way to compute a trend and will by definition bias your result.

There was no change in trend, just variability around a long-term trend.

It is a good idea to check the source. In this case the link goes to an unpublished manuscript.

The manuscript only cites one study with positive impacts. The one of Mr. Tol himself for a 1°C temperature rise, which is a point we have just reached. That impacts for this study become statistically significant only at 3.5°C is because the economic uncertainties are so large. Uncertainties go both ways.

The introduction gives another impression as this op-ed:

For such a fractious discipline, there has been remarkable agreement amongst economists on first-best climate policy. Ever since the writings of Nordhaus (1977), d'Arge (1979) and Schelling (1992) , it has been widely accepted that climate change is, on balance, a negative externality and that greenhouse gas emissions should be priced, preferably taxed.

It is deceiving to call these estimates "based on actual observations". Also comprehensive physical climate models reproduce the warming in the last century.

The real difference is that these estimates are based on simple statistical models rather than on comprehensive physical models.

This is beyond the literature I normally read, I hope the others are able to provide references, but I have the impression that we are starting to understand why these simple statistical models are biased too low and that the uncertainty of these estimates is very large because the estimates depend strongly on the assumptions made and the datasets used.

It should be noted that the estimates of the climate sensitivity are not only based on comprehensive physical climate models, but also on our physical understanding of the climate system (radiative transfer, the water vapour feedback and the albedo/snow feedback). And the climate sensitivity estimates are also based on the past behaviour of the climate system, the magnitude of the temperature changes during ice ages and the fact that Earth has recovered from being an ice ball. (I wanted to remind the others of these arguments and hope that they have references to back this up.)

This is the range for the Equilibrium Climate Sensitivity, not for the temperature in 2100. This temperature depends on how much we will mitigate climate change for which there are no predictions.

This is a wide range of climate sensitivities, but not a sign of disagreement in the scientific community as the op-ed suggests. This range is a fair summary of the evidence for most scientists.

That clouds and aerosols are an important reason for the large confidence interval for the climate sensitivity is generally accepted. However, it is an exaggeration to claim that Bjorn Stevens found this. He presented limited evidence that makes a case for future research into this topic. more in general, one should never judge the state of the art based on single papers, especially new papers.

It would have been nice to get an argument why the best estimate that all warming since 1951 is man-made is wrong.

I do not like focussing on single years, but 2015 will indeed likely become very hot.

The second sentence is also true, climate change is real and ongoing.

The implied link between sentence one and two is okay, without global warming a temperature record this year would be much less likely.

Catastrophic is a political category. Thus by definition there is no "convincing scientific evidence" for it. But the impacts and adaptation will be costly.