Study: Another failure of climate models – they can’t handle barometric pressure change

One of the most basic meteorological factors isn’t handled by climate models – and they run off in a “hockey stick” style.

From the UNIVERSITY OF LINCOLN

Climate models fail to simulate recent air-pressure changes over Greenland

Climatologists may be unable to accurately predict regional climate change over the North Atlantic because computer model simulations have failed to accurately include air pressure changes that have taken place in the Greenland region over the last three decades.

This deficiency may mean regional climate predictions for the UK and parts of Europe could be inaccurate, according to new research published today.

Researchers compared real data with simulation data over a 30 year period and found that the simulations on average showed slightly decreasing air pressure in the Greenland region, when in fact, the real data showed a significant increase in high air pressure – or so-called ‘Greenland blocking’ – during the summer months. These simulations are widely used by climate scientists worldwide as a basis for predicting future climate change.

The findings raise serious questions about the accuracy of regional climate projections in the UK and neighbouring parts of Europe because meteorological conditions in those regions are closely linked to air-pressure changes over Greenland.

Researchers warn that record wet summers in England and Wales such as those experienced in 2007 and 2012 could become more frequent if Greenland air pressure continues to strengthen over the next few decades, but such a trend might not be predicted due to inaccurate regional climate simulations.

The study, carried out by the University of Lincoln, UK, and the University of Liège in Belgium, also concluded that current models of melting on the Greenland Ice Sheet – a vast body of ice which covers more than 80 per cent of the surface of Greenland – may significantly underestimate the global sea-level rise expected by 2100.

Professor Edward Hanna led the study with Dr Richard Hall, both from the University of Lincoln’s School of Geography, and Dr Xavier Fettweis of University of Liège. Professor Hanna said: “These differences between the estimates from the current climate models and observations suggests that the models cannot accurately represent recent conditions or predict future changes in Greenland climate.

“While there is natural variability in the climate system, we think that the recent rapid warming over Greenland since the early 1990s is not being fully simulated by the models, and that this misrepresentation could mean that future changes in atmospheric circulation and the jet stream over the wider North Atlantic region may not be properly simulated.

“Until now, no-one has systematically examined the projections to see how they represent the last few decades and future changes – up to the year 2100 – from a Greenland regional perspective. Previous work reported a tendency for global warming to result in a slightly more active jet stream in the atmosphere over the North Atlantic by 2100 but our results indicate we may actually see a somewhat weaker jet, at least in summer.”

The research is the first to systematically compare global climate model data and observational data of air pressure changes for the Greenland region. The study, Recent changes in summer Greenland blocking captured by none of the CMIP5 models has been published in the European Geosciences Union journal, The Cryosphere.

###

The paper:

Brief communication: Recent changes in summer Greenland blocking captured by none of the CMIP5 models

Edward Hanna et al.

Abstract. Recent studies note a significant increase in highpressure blocking over the Greenland region (Greenland Blocking Index, GBI) in summer since the 1990s. Such a general circulation change, indicated by a negative trend in the North Atlantic Oscillation (NAO) index, is generally highlighted as a major driver of recent surface melt records observed on the Greenland Ice Sheet (GrIS). Here we compare reanalysis-based GBI records with those from the Coupled Model Intercomparison Project 5 (CMIP5) suite of global climate models over 1950–2100. We find that the recent summer GBI increase lies well outside the range of modelled past reconstructions and future GBI projections (RCP4.5 and RCP8.5). The models consistently project a future decrease in GBI (linked to an increase in NAO), which highlights a likely key deficiency of current climate models if the recently observed circulation changes continue to persist. Given well-established connections between atmospheric pressure over the Greenland region and air temperature and precipitation extremes downstream, e.g. over northwest Europe, this brings into question the accuracy of simulated North Atlantic jet stream changes and resulting climatological anomalies over densely populated regions of northern Europe as well as of future projections of GrIS mass balance produced using global and regional climate models.

Open access downloads:

 

Advertisements

121 thoughts on “Study: Another failure of climate models – they can’t handle barometric pressure change

    • Nice work Vuk’ , that is pretty convincing correlation, though if you are looking physical causation , I would suggest using SST or land SAT ( or may be both separately ) , not a bastardised chimera of the two.

      Here is a less heavily filtered look at the same data as is discussed in the article.
      https://climategrog.files.wordpress.com/2018/10/gbi-peak.png

      We can see that the latter half of the data period ( for which we have satellite sea ice data ) is not representative of the atmospheric changes in the region. Most of the ice loss that got everyone freaking out occurred between 1997 and 2007. This can be compared to the period of the rise in 500mbar isobar. Attempting to fit a straight line to the whole period of the ice record and call it a “trend” is misleading.

      This also shows where the atmospheric models are fatally flawed. They are based on models “tuned” around the idea of a declining “trend” + “noise”. This is simply not what climate is doing.

      • Hi Paul
        CRUTEM4 is a gridded dataset of global historical near-surface air temperature anomalies over land averaged on a 5 degree grid – red line (data source Met Office) .
        Geomagnetic (field) – GMF dipole is an overall strength of the Earth’s magnetic field calculated from data available for the two polar regions – blue line (data source NOAA).
        High correlation R^2 is obtained when GMF data is moved 10 years forward – green line.
        Assumption: if there is a causal direct or indirect relationship between two, the change in the global temperatures takes place about 10 or so years after the change in the Earth’s magnetic field occurred.
        Important stipulation: any correlation however high does not mean causation. However, natural variability of the global temperature currently is not fully understood, thus a high correlation to the another natural variables should not be dismissed as irrelevant.

  1. The press release – I assume that’s what it is – does not contain the word “grid” at all. With a typical grid size of 200×200 km, most meteorological phenomena like low/high pressure areas, fronts, even hurricanes, can’t be modeled with any accuracy.

    • Actually, George, these days the typical gridcell size is about 100×100 km (60 x 60 miles). That’s plenty to model high-low pressure. The problem isn’t that they can’t model pressure. The problem is that they do a lousy job of it.

      It does, however, prevent them from modeling the most crucial climate phenomenon, thunderstorms, as well as things like squall lines and dust devils, all of which are huge players in thermoregulation.

      Regards,

      w.

      • Hmmmm, 100x100km seems small enough to simulate a tropical cyclone with some degree of fidelity. But if you monitor noaa’s hurricane watch site https://www.nhc.noaa.gov/gtwo.php?basin=atlc&fdays=2 it seems that tropical cyclones in the Atlantic and Eastern Pacific typically start off as a region of low pressure with a lot of thunderstorms. There are many of those. Some develop into cyclonic storms. Most (?) don’t. My question would be whether the simulations are good enough to ever create a tropical cyclone

          • Thanks Kristi

            It’s a TED talk and appears to be quite general . Unfortunately, it suffers from problems that look to me like bad Javascript (Is there any other kind?) and therefore the video can’t be paused, the volume can’t be adjusted, and requests to see and transcript are ignored in my version of Firefox on Linux. None of that — other than the generality — is your or Gavin Schmidt’s fault, but it’s still highly annoying and I’m invoking my standard response to crummy software. I’m going to ignore the damn thing.

            Thanks for trying.

            My curiosity being aroused, I’ll go see if I can find something useful and informative on GCMs and tropical cyclone simulation.

          • Gav is deceptive in dismissing the failure to reproduce the early 20th c. warming as “it’s in the noise”. No, the models simply fail to produce something which was as large as the late 20th c. warming.

            He then says the model match the more recent warming without explaining that they have been “tuned” to do so by rigging a large number of parameters which are pre-established physics but frig factors.

            But then he has a dog in the fight. He is at the head of NASA GISS and is trying to sell his product.

          • I went off to do other stuff, and when I came back, the stupid TED talk had decided in my absence to acknowledge my request for a transcript. I read said transcript. It’s fluff. It does babble a bit about modelling “emergent phenomena” and I think it likely that climate models do model diurnal, and seasonal phenomena reasonably well. My understanding is that some sort of model ENSO. That’s nice. Really. It’s better than not modelling those things. Nothing about tropical cyclones.

            I went off and did some web searching. Given paywalls idiosyncratic security certificates, some utterly opaque writing, etc, it’s hard to be sure. But it looks like GCMs don’t do all that well with tropical cyclones. Insufficient/marginal resolution and maybe problems spawning them. There are of course, non-GCM “hurricane” models that specifically model these storms with reasonable resolution and presumably better results. The best article I found is at http://naturalhazardscience.oxfordre.com/view/10.1093/acrefore/9780199389407.001.0001/acrefore-9780199389407-e-22

      • Willis,

        They can’t predict when a storm is going to happen – that would be asking a little too much, wouldn’t it? Better than weather forecasting. Cyclones are an emergent property of simulations. See the one here:
        https://www.ted.com/talks/gavin_schmidt_the_emergent_patterns_of_climate_change/up-next#t-41995

        I don’t know why thunderstorms wouldn’t be possible, since models simulate precipitation events.

        Dust devils – not part of GCMs yet, but they are modeled
        http://adsabs.harvard.edu/abs/2015EGUGA..17.6191J

        • “They can’t predict when a storm is going to happen – that would be asking a little too much, wouldn’t it?”

          I agree that I wouldn’t expect them to model actual storms. Getting the average number and intensity roughly right would be the best I’d expect of a good model. However NOAA actually does make predictions about the likelihood of real, observed storm clusters developing into tropical storms in the next few days and, while not perfect, the predictions seem to me to better than useless. It’s a bit late in the season, and when I looked a few hours ago, NOAA was predicting no development in the next 48 hours in the Atlantic Basin, but they show one defunct storm, one probably developing storm off the coast of El Salvador, and a possibly developing storm 900mi SSW of Cape San Lucas in the Eastern Pacific.

          • The NOAA NHC forecasters use reports/observations of storms then assess the impact various weather features that have been observed such as shear and dry air will affect those storms.

            What the models should do is say when a tropical wave will appear in say the next month and how it will develop and what path it will follow. The closest I have seen to that capability is from Weatherbell and Joe Bastardi, and that appears to be mainly based on analogues and seems to beat models.

            The Greenland blocking high is really an emergent feature of the Rossby waves in the jet stream. So the models are unable to forecast the jet stream and the strength of the Ferrel cells which would appear to be a major failing.

        • “I don’t know why thunderstorms wouldn’t be possible”

          Because they happen on a smaller scale than the grid-size. Convection cells typically are from a few hundred meters to a few tens of kilometers large. And, by the way, you will need surface topography and albedo at the same 100-m scale to physically model convection. Which is the dominant mechanism controlling surface temperature, considerably stronger than radiative cooling.

          By the way, going from 100 km to 100 m cells will require (10^3)^4 = 1,000,000,000,000 times more computing power (1000 times finer resolution in three spatial and one time dimension).

          It’s gonna take a while I think.

  2. I noticed they still managed to get hysterical about sea level rise from Greenland. Too bad for them the last two years showed a reversal (mass balance increase in excess of melting/calving.

  3. For years I have been providing comments on various Climate web pages that there are just to many variable and more than half of them are unknown to even attempt to make a computer model of the global atmosphere and global temperature changes. Designed computer models for Coal/Nuclear Steam Power Plant Simulators forty years ago. It took years just to get one process modeled properly and we already had all of the math equations! We also had the real powerplant to play with and record the changes in system parameters and verify that the modeled process did exactly the same as the operating plant. The other problem with models is that we could build one for a 500 MW plant and when we built a 1,000 MW plant the old model was essentially useless. All it was good for is an outline of what we had to do. The other problem with the Global Warming Models is that they are predicting what is going to happen in the future, That means that the slightest error is compounded, just like interest in your retirement account and negatively like the interest on your bank loan. Then Minor errors that do not show up in short term projections will provide absolutely useless results. Those that were in college when the Digital Calculator came out (CA 1972) know full well that the answer given to Statics/Dynamics problems do not provide the same answer as a slide rule – for several various reasons. And th people making these Climite models don’t eve know what they don’t know that is causing the problems that make their models worthless let alone the paramaters that are important in the model.

    • Usurbrain, good comments about the complexities of modelling on computers. Then take a chaotic system with a lot of unknowns, as you point out, and there is not a good result. However, there is the hoped for result, because the computer models of climate catastrophies are not the end game, the end game is when the money changes hands.

      • Precisely, we need to remember that the climate BS is just a smokescreen for the real agenda which is to usurp world power to an unelected UN body.
        Maurice Strong et al openly admitted this multiple times and it is time this was made widely known.

    • Climate modelers do not understand that, “error is compounded,” Usurbrain. They do not understand propagated error, and in fact do not understand physical error at all.

      I have extensively discussed this problem on WUWT, for example, here.

      Since that post, I’ve had several more encounters with climate modelers. Not one of them understood physical error, or it’s analysis, or its impact. They are untrained, arrogantly dismissive, and hopelessly incompetent.

      • Pat Frank,

        I’m curious – how much do you know about the process and theory of building climate models? Have you read any books about them? Have you studied the literature as it relates to the way error is dealt with in the models in a theoretical sense? What is your area of expertise, if I may ask?

        It seems a little hard to believe that the dozens of climate modeling groups all over the world are universally hopelessly incompetent, and all in the same way, and that no one but you and maybe a few others have noticed. There are, after all, statisticians involved in these things.

        What’s you

        • Did you read the post I linked, Kristi? Had you, you’d not have asked that question.

          Another of my posts further detailing the incompetence of climate modelers is here.

          Here’s what I know about climate models: the uncertainty in their air temperature projections reaches ±15 C after 100 years. Climate models have zero predictive value.

          I have demonstrated that. It’s a fact. The linked posts show it. Here’s yet another.

          I have now dealt with about 27 climate modeler reviews. Every single one of them is incompetent. Every. Single. One.

          My view of climate models is from error analysis, Kristi; analyzing the physical errors climate models make. I’m a physical methods experimental chemist.

          I sweat physical error all the time. I have direct and repeated experience that climate modelers know nothing of it. They are incompetent to evaluate the accuracy of their own models.

          So you go on ahead and wrap yourself in the warm swaddling of “not a statistician.” If you like, have some he’s not a climate scientist for dessert.

          Or, bite the bullet and do some ferschlunginer critical thinking for once.

          • Pat:
            “Here’s what I know about climate models: the uncertainty in their air temperature projections reaches ±15 C after 100 years. Climate models have zero predictive value.”

            “I have demonstrated that. It’s a fact.”

            You have demonstrated it to yourself – and this echo-chamber will agree anything that supports ABCD “science” – however your peers do not agree.
            That’s a fact.
            Just like the consensus of experts in the observations matching theory in climate science forces common sense to tell us why it is most likely correct. Conspiracy excepting of course.

            https://www.youtube.com/watch?v=rmTuPumcYkI&feature=youtu.be

            I fairness your rebuttals to the video are here ….

            https://patricktbrown.org/2017/01/25/do-propagation-of-error-calculations-invalidate-climate-model-projections-of-global-warming/

          • That describes the problem displayed by the early digital calculators when trying to calculate the stress on a cable suspended at each end with a heavy weight in the center. Angle is small, thus dealing with numbers like 0.000123. The six digit capability lost digits or created incorrect numbers that were used for the rest of the calculations. Users got answers different from the one in the book and even different calculators gave different answers.

          • Basically, they can make the models produce the results they are looking for, so therefore the models must be correct.

          • Anthony Banton, if you read the posted links you’ll discover that the demonstration of ±15 C uncertainty after 100 years is an objective fact.

            If you have some substantive criticism of the analysis, let’s see it. If not, then your unbuttressed opinion means nothing.

            Climate modelers are not my peers. They’re not even scientists.

            Had you read the debate below Patrick Brown’s critique, you’d have noted his argument did not survive the encounter.

            He showed no understanding of model calibration, calibration error, or the impact of such error on predictive accuracy.

            He supposed a ±uncertainty is a positive offset error, and at the end tried to eliminate the denominator that defines the locus of an average.

            So go ahead and comfort yourself. But your argument lacks any weight.

          • Anthony Banton, to put that ±15 C uncertainty factor into perspective that you might understand, that uncertainty is like boarding an airplane in NYC on a trip for LAX. The pilot puts it in “Autopilot” [a computer model] and when you land you discover you are in Sea–Tac (+ 15 degrees) OR it could be Mexico City (-5 degrees). Even their best prognostications you would land at John Wayne Airport, Orange County, OR San Diego, you won’t know till you get their.
            And the IPPC Working Group reports tell you this in the technical section. NOT in the SPM, Summary for Policy Makers, the section with the lies, propaganda and threats of disaster. I seriously doubt that you have ever read any other section, let alone all of the section and definitely not every one of the periodic releases. Or even have any idea how many pages there are in the other sections. But I do know you will fire back a page number count of each to bolster your lie that you have.

        • It seems a little hard to believe that the dozens of climate modelling groups all over the world are universally hopelessly incompetent, and all in the same way, and that no one but you and maybe a few others have noticed. There are, after all, statisticians involved in these things.

          Yup. Its a shocker, innit?

          Nevertheless, its true.

          In almost every area specialization blinds one branch to the workings of another.

          Climate modellers take the data on trust, the data people bend the data, and take the models on trust.

          And you trust all of them.
          Those of us who have bothered to do the deep investigations and the hard sums ourselves, know that teh answers they (climate scientists) give cannot be right. And its no different fir e.g. nuclear power, fracking or renewable energy. All of these are subject to mass hysteria and propaganda under which the truth gets buried.

          Why didn’t we and others speak out?

          Some did, and lost their jobs.

          Others buttoned their lips and kept theirs.

          This bandwagon of belief, suits almost everyone involved in it. Politicains use it to geain power. Commercial companies use it to make money. Greens have been bought totally, and have stopped bothering about real ecosystem damage and are now all funded secretely to wibble on about only climate change, fracking and nucklear energy.

          Energy is teh indispensable component of a post industrial world. It is IIRC the biggest global market there is.

          The rewards for bringing it under global control by a cadre of individuals are immense.

          If you look carefully at what is happening in the world today and work out who benefits – “Cui Biono” – You will find that the Arab states and Russia, which huge reserves of oil and gas – benefitr mightily from global islamification, from rising energy prices, from elimination of coal, nuclear power and fracked gas in the west, and all the little countries can claim cash from te West for te damage the West has caused from ‘Climate Change’.

          Kristi: Let the scales fall from your eyes. Thsi isn’t about the science: The scientists were bought with grants and career promises and fame and fortune decades ago.

          THis is about geopolitics, power and money and control. Russia learnt years ago that when you cant opr dren’t beat an enemy militarily, you subvert his nation economically and culturally. The lesson has been learnt. Liberalism is teh result – a massive subvesrion and corruption of the culture that actually made the West great, by telling it that it should feel guilty of success, and that the price of success is too high.

          All of the modern ‘Liberal’ movements – promoted as social justice – are in fact carefully designed to destroy the core commonality of culture of the West. LBGT rubbish cause people to question their own identities. Gay marriage drives a cart and horses through the idea of marriage and family as a protected space for children, Radical Feminism drives a wedge between the sexes isolating and confusing both sexes, vociferous minorities with chips on their shoulders are given preferential treatment whilst the tendency for ever higher taxation and public sector employment means that people are no longer rewarded for productivity, but for political correctness.

          World war three has been going on for several decades. Its juts a very unconventional and asymmetric war. There are people out there who want your land, your wealth and your lifestyle.

          And have deep enough pockets to buy most of the politicians they need. And then te politicians use tax dollars, raised for the most righteous of causes, naturally, to buy the scientists and academics whilst the business that support them buy the journalists and the media tarts.

          You live in a very protected bubble Kristi. People in Britain didn’t believe about the holocaust till someone sent back pictures of Belsen. All the people who believed that it was plain stupid to believe civilised people could do that, were faced with the horrifying thought that in fact they had.

          You find it hard to believe that you have been lied to comprehensively about almost everything for most of your life.

          Yup. We all felt that way.

          Welcome to the club.

          • “Russia learnt years ago…”
            Leo excellent summing up but perhaps we could add to the above as in “Russia and China learnt years ago …”

            I say that because the very idea of getting all your stuff made in China that burns coal to produce it instead of having your own country make it by burning coal to produce it seems completely illogical and unlikely to ‘save the planet’ even if all their models were perfect and showed that the planet does indeed need saving.

        • “There are, after all, statisticians involved in these things.”

          I rather doubt that given the egregious statistical errors that abound in climate science.

          Having done a fair amount of statistical analysis in my day (and published a few papers on the results) I usually check for about half a dozen common statistical bloopers whenever I run into another it’s-worse-than-we-thought paper. More often than not I find at least one.

    • Usurbrain, Good comments about climate models. Canadian Mathematician Dr. Christopher Essex does an excellent job covering, I think, what you are saying and goes into good detail about why climate models cannot EVER be correct (future quantum computing possibly withstanding) due to various computer errors, etc. This video is about one hour long, but has some surprising revelations about unintended computational results. (In case you haven’t seen it). https://www.youtube.com/watch?v=19q1i-wAUpY
      Clay

    • The oft-quoted response to this is “well it’s the best we’ve got so we had better run with it”. Some of us beg to differ.

    • Usurbrain,

      Well, there are a few key differences here.

      1. There are thousands of people collectively contributing information and expertise to climate modeling

      2. No one expects climate models to provide the kind of precision needed in coal/nuclear power plants.

      3. Models have actually predicted some things reasonably well, and they are improving. They have also been skillful in simulating the past.

      4. Prediction is not the only function of models. They also help understand climate.

      5. Scientists know there are weakness in their understanding of some parameters, but that doesn’t mean their understanding won’t improve

      6. Science is a process. Design is a goal.

      • First, point out the VERIFICATION of these models. Second, Any group of people that could do what you say do would also be able to do the same with the stock market. And thus be billionaires. Where are they?

        • Usurbrain,

          “The latest version of the National Center for Atmospheric Research (NCAR) Community Climate Model (CCM3) is described. The changes in both physical and dynamical formulation from CCM2 to CCM3 are presented. The major differences in CCM3 compared to CCM2 include changes to the parameterization of cloud properties, clear sky longwave radiation, deep convection, boundary layer processes, and land surface processes. A brief description of each of these parameterization changes is provided. These modifications to model physics have led to dramatic improvements in the simulated climate of the CCM. In particular, the top of atmosphere cloud radiative forcing is now in good agreement with observations, the Northern Hemisphere winter dynamical simulation has significantly improved, biases in surface land temperatures and precipitation have been substantially reduced, and the implied ocean heat transport is in very good agreement with recent observational estimates. The improvement in implied ocean heat transport is among the more important attributes of the CCM3 since it is used as the atmospheric component of the NCAR Climate System Model. Future improvements to the CCM3 are also discussed.”
          https://journals.ametsoc.org/doi/abs/10.1175/1520-0442(1998)011%3C1131:TNCFAR%3E2.0.CO;2

          https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/94JD01570

          https://link.springer.com/article/10.1007/s00382-011-1099-9

          https://www.nature.com/articles/nclimate2357

          This deals with some of the processes involved in verification and validation.
          http://www.climate.be/textbook/chapter3_node22.xml

          It’s important to remember that different models concentrate on getting different things “right” and that climate simulations are never going to be perfect fits to reality because they aren’t designed to project future time-specific states, but only averages, i.e., they can project that there will be more intense rain events in a given region, but they won’t predict when the rain events will happen.

          The stock market is subject to human reactions, news stories, acquisitions, all kinds of stuff that is impossible to predict from one year (or week, or day, or hour) to the next. It’s not about known (or estimated) physical/chemical/biological relationships. Totally different.

          • Would you volunteer to be the first person to drive a caravan of vehicles hauling the maximum design load across a suspension bridge in a storm with wind speeds at the maximum design wind speed that was designed with error margins as large as those explicitly stated in these Global Change Models?
            Before you say yes, think of the Hyatt Regency walkway collapse https://en.wikipedia.org/wiki/Hyatt_Regency_walkway_collapse
            or
            The day a bridge collapsed in Minneapolis and lives changed forever
            https://www.twincities.com/2017/07/29/the-day-a-bridge-collapsed-in-minneapolis-and-lives-changed-forever/
            Then consider that engineering practice Required allowing for 100% overload in the design calculation.
            The GCM show that there is as much as 50% uncertainty in their calculations. And you are proposing spending multiple $Trillions a year on this massive money sinkhole.

          • The chaos of the atmosphere is very similar to the chaos of “human reactions, news stories, acquisitions, all kinds of stuff that is impossible to predict from one year (or week, or day, or hour) to the next. “

          • Did you actually read (and understand) your own quote Kristi?

            It says “ changes to the parameterization of cloud properties, clear sky longwave radiation, deep convection, boundary layer processes, and land surface processes.”

            It means that NONE of these crucial processes are calculated from basic physical law, but are simply guessed at. I have actually developed and worked with computer models. They were a lot simpler than GCM’s (they dealt with logistical processes) and I can assure you that any model with more than three parameterized variables is utterly useless, because you can get essentially any result you want by fiddling the parameters, and usually without using unreasonable parameterizations too.

          • Kristi so they admit they had parts of it wrong “and they now agree”. They do this every assement. Saying yes the old one didn’t match but hey look the new one does. And, what do you know it produces the same hockey as all the other ones that were wrong. Why would we expect this one to be different?

      • Kristi, the only reason models can simulate the past with any accuracy is that they are designed precisely to do that. First the model is created, then, one by one, the dozens of parameters with assumed values are tweaked slightly, until such time as the model creates a reasonable simulation of the past. At that point the modellers consider that it is perfect, & start running it to create their predictions of the future – which have been proven for decades to be totally wrong.

        • No not true.
          Hansens’s 1988 paper just abut nailed it my friend.
          it’s the science behind the models that is sound (radiative physics of CO2 in the atmosphere).
          What modes struggle with (as they do with weather NWP) is the chaos within the system as heat is moved within it before exiting to space.
          Models means are just that – a mean of numerous runs in an ensemble – that smooths perturbations ….. even were it possible to project them.
          It is obvious how the PDO/ENSO cycle affects GMSTs – vis the prolonged -ve part of that cycle that slowed warming (aka the pause).
          In the end the noise of NV will be overcome by the overwhelming forcing of the GHE.
          That is GHGs reduce the Earth’s ability to cool.
          See Nick Stoke’s breakdown of Nansen 1988 …

          https://moyhu.blogspot.com/2018/06/hansens-1988-predictions-30-year.html

          “What the showing of combined temperature records shows is that Hansen’s 1988 prediction is about as good as it could be, because it sits within the scatter of modern records. The difference between GISS Ts and GISS land/ocean is comparable to the difference between GISSlo and scenario B. “

          • Gotta give credit to Kristi and Mr. Banton. They go in way over their heads, against people who know of what they speak (Pat Frank has reduced N. Stokes to spewing word salads on this subject, though he’s “not a statistician”; usurbrain no slouch, either), look foolish and wrong (Banton has seen Mr. Stokes say that Hansen’s ’88 is sooooo close, so Banton repeats it as if gospel, not caring that it’s been refuted by commenters who may “not be statisticians”). And when refuted, Kristi and Banton go away only to return for another lesson in humility (never takes, does it).

      • The models are initialised to steady state. After that they apply the forcings and then claim they models accurately predict temperature evolution 1861-present. They show this with the average of the model ensemble as compared with HadCrut4 temperature from 1861. However, the model ensemble mean can also be reproduced very closely from just a linear regression of the forcing parameters (Willis demonstrated this some time ago here at WUWT, I have done independent work on this too). This means that the models do nothing except transform input forcings (in W/m^2) into temperature change (in degK). They add no other physical insights or emergent properties.

        Furthermore, if the model ensemble mean is subtracted from the individual climate models the individual model residuals are not correlated with temperature AND the distribution of correlation coefficients vs temperature is exactly what you would expect from random errors. In other words, we find:

        CMIP5 climate model = forcings + random noise

        No physics is required, only linear regression. The random part arises, I suspect, because climate models are sensitive to small changes in the initial conditions. This makes them behave like computer random number generators.

        The spread of the temperatures from the different models (there are 39 individual models in the official CMIP5 ensemble) is MUCH larger than the claimed error bars. Normal science, when calculating the mean of a set of results, would calculate the standard deviation from the same approach. The average annual SD over the period 1861-2011 from CMIP5 39 model ensemble is 0.71 degK. So for 95% CI that is +/- 1.4 degK, a range of 2.8 degK ie MUCH larger than the HadCrut4 warming over the period. But climate modellers don’t calculate the SD consistent with the mean, oh no. What they do is normalise all the models to the period 1961-1990 first, and then do it. This gives an average annual SD from the 39 ensemble models of 0.17 degK, or +/-0.33 degK for 95% CI. This is not correct and hides the true error with a number that is over 4x smaller.

        Finally, the climate model residuals over time show no emergent features and they clearly cannot simulate multi-decadal oscillations nor short period events such as El Nino. They are basically random. So climate models are basically window dressing on the published climate forcings.

      • Models are skillful in predicting the past only because they have dozens to hundreds of parameters to tune. A model that can’t predict the past is proven to be useless from the get go. However merely being able to predict the past is not proof that the science in the model is correct.

        The fact that the scientists admit that they don’t understand all the processes is sufficient in and of itself to argue that we don’t need to do anything about CO2 until the scientists do understand.

    • What happens in these numerical models is called NMUERICAL INSTABILITY and has to do with multiplying small numbers thousands of times (iterations). So, the result will either be a very large number, or it will be a very small number (trending to zero).
      So either the pressure will build up wildly with time, giving a massive high pressure (over Greenland), or it will trend towards a low number (i.e., a very redicoluously low pressure over Greenland (which will be obviously wrong).

  4. “We find that the recent summer GBI increase lies well outside the range of modelled past reconstructions and future GBI projections (RCP4.5 and RCP8.5). The models consistently project a future decrease in GBI (linked to an increase in NAO), which highlights a likely key deficiency of current climate models if the recently observed circulation changes continue to persist. ”

    I note that they say “the recent summer GBI increase”.
    Climate projection is about more than “recent summers”, and the “recent increase” may or may not be an ongoing feature.
    Climate models may say not but we don’t know.
    It is what we have models for they are a tool … to learn from stuff such as this.

    I also note that they also say …..

    “The study, carried out by the University of Lincoln, UK, and the University of Liège in Belgium, also concluded that current models of melting on the Greenland Ice Sheet – a vast body of ice which covers more than 80 per cent of the surface of Greenland – may significantly underestimate the global sea-level rise expected by 2100.”

    • These “tools” (and I use that term very loosely) are not ready for prime time. Their current development stage can’t even deliver the correct sign for many processes (as seen in this post, for one) let alone the right magnitude to the nearest power of 10. They are laboratory curiosities and should not be used to disassemble the world’s economy in an effort to avert a level of warming that barely exceeds the error band for most temperature measuring devices.

    • https://www.livescience.com/63423-lost-squadron-unearthed-greenland-glacier.html

      The Greenland ice sheet melts from below because of a volcanic ridge. In World War 2 in July 1942, bomber airplanes were forced to land in a storm over Greenland but the pilots and crew survived and made it to the coast. 70 years later some of the planes were found and just this year another one was found. The planes have been covered with an average of 1.7 metres of ice per year that was snow which turned to ice. So since calving of glaciers on the coast is a natural process and 1.7m of ice is added each year to the top, global warming has not made a dent in the melting of Greenland. So Mr. Banton I call bullshit to you and any computer climate modeler of Greenland.

    • …I note that they say “the recent summer GBI increase”.
      Climate projection is about more than “recent summers”, and the “recent increase” may or may not be an ongoing feature…

      You failed to note that by “recent,” they are talking about since 1990.

      …It is what we have models for they are a tool … to learn from stuff such as this…

      What have you learned exactly? You’re still suggesting this may be temporary.

      The only thing we learned was

  5. Al Gore was complaining how loopy the jet stream is getting lately.

    Gore: Jet Stream ‘Getting Loopier and Wavier,’ So ‘We Have a Global Emergency’ (12 Oct)

    Mr Gore seems to think CO2 causes this, which is incorrect. Unsurprisingly the recent stories about loopy jet streams and blocking events mirror the same sorts of stories in 2010. Remember when the UK was entirely covered in snow in winter 2010 and the great 2010 Moscow heat wave occurred? That year was the low point of the solar cycle, and we’re now back into similar solar activity conditions as SC24 winds down.

    There are a number of papers around demonstrating the link between low solar activity and loopy Rossby waves. Prof. Mike Lockwood, who is a warmist, even told the BBC this back in 2010 during the UK white-out.

    I suspect the problem the models are having is a symptom of their general antipathy for anything solar. If they acknowledge solar activity is a strong driver of climate then they might also have to acknowledge that the rising solar activity during the 20thC naturally caused about half of the warming. Which would end the CAGW scam…and their modelling budgets.

    • Is he simply projecting himself into the jet stream? We’ll know for sure if he says it’s getting fatter.

  6. The study, … also concluded that current models of melting on the Greenland Ice Sheet … may significantly underestimate the global sea-level rise expected by 2100.

    Jordan Peterson talks about the ideologically possessed. link He points out that you can easily predict what they’re going to say. They are governed by a few simple algorithms.

    In the case of the above quote, the message is “It’s worse than we thought” … entirely predictable.

    BTW, we’re being bombarded on the CBC (Canadian Broadcorping Castration) 24-7 about the most recent IPCC report. Has CAGW jumped the shark? The claims have become way more extreme than even most warmist leaning scientists would be comfortable with. Is CAGW on the brink of imploding?

    • commieBob

      “Has CAGW jumped the shark?”

      The BBC is also doing round the hour climate scare stories on radio and TV.

      Astonishing really, the public are being panicked into inaction but they don’t care any more because the targets are so unachievable.

    • CommieBob,

      What a coincidence. I listened to an hour and forty minute lecture by Peterson just last night that a friend sent me.

      The ideology thing holds true for lots of people, not just liberals or alarmists, but skeptics and conservatives, too.

      The title of this post is, I think, a pretty good example. “Can’t handle barometric pressure change” is hardly what this study says – it’s talking about a specific region for which the simulations have been erroneous in the last 30 years. No one claims they are perfect.

      • “No one claims they are perfect.” Thought it was worth repeating, a true statement by Kristi. What has been said is that they are “not fit for purpose.” You seem to think they are, so let’s find out. Have the models convinced you (Banton can join in anytime) to stop burning fuel? Stop using electricity produced by burning fuel? Stopped air conditioning in your residence? IF you find the models fit for purpose, then you and Mr. Banton are making the necessary sacrifices, right?

        When did you last travel by air?

      • “it’s talking about a specific region for which the simulations have been erroneous in the last 30 years. No one claims they are perfect”

        This specific region is very important to all the Northern Hemisphere. If you get it wrong here this means it is also wrong for significant portion of the NH.

        Low pressure over the region generally means the NH jet stream is zonal with mild and often wet weather for most. In this situation the PV (polar vortex) is likely to be stationed here. High pressure over this region generally means the NH jet stream is meridional with often cold and dry weather for most.

  7. Userbrain,
    You said, “And th people making these Climite models don’t eve know what they don’t know.” +1

    I have a little experience with programming deterministic models, and even with only simple physics, well documented relationships, and a small number of variables, it it not trivial to get the models to work well.

    • That describes my problem, We had the Exact volume of a tank (boiler), the exact surface area of the tube sheet in the tank both inside and outside, the exact mu (heat transfer value) of the tubes in the tube sheet, the same for all of the factors effecting the heat loss of the tank. The model for this process was a simple heat in vs temperature rise for the tank (Boiler) and the steam output. We also had the engineering calculations used by the designer for this simple process which were verified during plant startup as part of the acceptance testing. Simple problem, right? In less than a week we had a simple model. However, with a five point verification against the design engineers calculations (which as I said above were proven during plant startup) the curves did not match. Took another month of computer runs and “Tweaking” to get the Model to match the actual. That is why I do not believe these unverified (against the real world) models.

      • As engineers know, the devil is in the details. I was the facility manager at a semiconductor tool maker. We wanted to install a closed loop cooling system for the tool test bays, allowing them to run their paces as in real life. We hired a mechanical firm to do the engineering and installation. Since ChemE was my major, I back-checked their calculations, and based on the proposed route, pipe sizes, fitting and valve counts, it looked like they had sized the pump just fine, and then some. On start up, it barely made 50% of the expected flow.

      • Userbrain,
        90% of the effort in finishing a computer program is the last 10% of the coding — tuning it and getting the bugs out.

  8. What if the string of high pressure occurrences in Greenland is something temporarily like the Dust Bowl heatwaves and droughts were? Did anything like that happen the previous time the Atlantic Multidecadal Oscillation was high, or is there no data from back then? (I know that jet streams were only beginning to be well known back then.)

  9. Climatologists “””””may””””””” be unable to accurately predict..

    Pompous BS…they can’t predict anything now

    • It’s about perception these days.
      If your public image is good, everything you proclaim is gospel (no matter what weasel words you employ).
      The media educated public was largely never taught the concept.
      That’s why they try to destroy the public image of those who disagree.

  10. “This deficiency may mean regional climate predictions for the UK and parts of Europe could be inaccurate, according to new research published today.”

    Why is there an assumption that the regional climate models are accurate in the first place? Climate models are the predictions derived from a climate change hypothesis; the first two steps of the scientific method. At this point, there should be no assumption that the hypothesis and the predictions derived from it are correct, but the above statement clearly indicates the these climate change scientists are shocked that the climate models ‘could be inaccurate!

    The next step is to test those predictions against actual observations. Once again, the climate models have failed this test, showing no more skill than a blind squirrel has at acquiring acorns. Yet, climate change scientists continue to talk and act like the model predictions are ‘settled science’. Until the models can show serious skill in this all-important part of the scientific method, they do not represent science, much less ‘settled science’.

    If predictions do not match observations, then the hypothesis is wrong! Will Edward Hanna et al. become real scientists and call for a reconsideration of the hypothesis? Doubtful. They do call for a tweak here or there. Perhaps this is how the whole paradigm collapses. All the tweaks that will be required to make the models fit the observations add up, until someone with the proper gravitas says what is clearly obvious: “Never mind!”

    • The ‘Head of the College of Science’ at the University of Lincoln (a botanist) is a true believer. I can’t instantly see where Hanna and Hall, who are geographers, sit in the academic structure, but they may need to mind their backs.

  11. “This deficiency may mean regional climate predictions for the UK and parts of Europe could be inaccurate, according to new research published today.”

    Try to say that statement like you mean it, without cringing or smirking. It would take a true believer to say that seriousness.

    • But at least this is a non-self-declared Sceptic funding that the settled Climate Science is flawed.

      Long may it continue.
      The more it becomes acceptable to point out the folly the wiser we become.

  12. Been saying this for a decade… and I’d like to also point out that they also can’t deal with DENSITY changes either. So the two actual physical forces that DRIVE CLIMATE are not accounted for in the “models”.

    Temperature is regulated by gravity at sea level. Fact.

  13. The 25 reasons why computer climate models will never get it RIGHT.

    1) Closure problem with equations (too many unknowns with not enough equations).
    2) Ambiguous definition of forcing, where 1 W/m^2 of incremental solar input is claimed to
    have the same effect as a 1 W/m^2 decrease in the size of the transparent window.
    3) Assuming linearity between temperature and forcing when their relationship isn’t even
    approximately linearity across the range of T found on the planet.
    4) Assuming the atmosphere exhibits active gain, when the requisite implicit, internal
    and infinite source of Joules to power the gain does not exist (the Sun is the forcing input
    and not the internal power supply).
    5) Errors 3) and 4) led to misapplying Bode’s feedback analysis to the climate system which
    was the primary theoretical justification for a sensitivity as high as the IPCC required to
    justify its formation and continued existence.
    6) Assuming the normal distribution of sites that’s required to apply homogenization.
    7) Conflating the energy transported by photons with the energy transported by matter
    relative to the radiant balance of the planet.
    8) Accepting a sensitivity that requires 1 W/m^2 of forcing to increase surface emissions by
    4.3 W/m^2 in violation of Conservation of Energy.
    9) Ignoring both the Sun and natural variability as the primary causes of climate change.
    10) The broad application of unsound statistical analysis.
    11) Time averaging of fluid dynamics equations doesn’t work correctly
    12) Ensemble averaging is ridiculous
    13) Numerical method will never be as good as solving the partial differentials of Navier Stokes equations but solution impossible
    14) floating point and rounding errors are too large
    15) Representation of infinite process is necessarily cut short and doesnt represent reality
    16) Errors are not random(ex: fake clouds) grid resolution only to 1.5km
    17) Actual greenhouse effect is a fallacy
    18) Water cannot be modeled unless you model it down to a raindrop which the climate models cant do.
    19) There is non existent long term variability because otherwise the simulation would become chaotic so the model has to be tuned to flatten the variability
    20) Not enough time to run long term calculations therefore a fake method is used to speed it up
    21) Models are not empirical. They don’t have all the earth’s data inputted.
    22) Averaging of field commodities doesn’t make sense because temperature is a local phenomenon.
    23) In fact the modellers actually cheat on future scenarios because they have to shortcut the time taken to model the future because we don’t want to have to wait a 100 years running the simulation just to get an answer a 100 years from now.
    24) Because the models don’t model convection well enough they all predict tropospheric hot spot which doesnt exist.
    25) Climate models do not adjust the barometric pressure changes as they happen in reality.

  14. According to Pierrehumbert they don’t understand relative humidity either.

    https://www.whoi.edu/fileserver.do?id=21420&pt=10&p=17292

    Pg 73
    “…Hsun = 340 W/m^2…”
    Only in that really stupid and unrealistic K-T model.

    Pg 77
    “…surface temperature (soil or ground) is approximately the same as the surface air temperature.”
    USCRN data show this is absolutely false and in a big way.

    “In simple models it is usually acceptable to equate surface temperature with surface air temperature.”
    This completely ignores USCRN reality!!!

    Pg 78
    “…relative humidity remains constant as the temperature increases.”
    USCRN data show this to be totally incorrect. As temperature rises RH falls – and in large amounts. Check psychrometric properties of moist air.

    The albedo/atmosphere (with no atmosphere, albedo is like moon as is temperature – 390 K on lit side, 190 K on dark.) cools the earth, i.e. reduces the solar heating, by reflecting away 30% of the incoming irradiation. The atmosphere cools (ok, actually reduces the heat load) the earth which is contrary to RGHE theory.

    As the sun rises the soil and air temperatures rise together, as the sun sets and during the night the low density air cools off rapidly to below the soil temp while the high density soil holds the energy it collected during the day and remains warmer & cools slower than the air throughout the night. This process is contrary to RGHE theory. This is a common and plainly evident in USCRN data sets.

    Energy moves, i.e. heat, from surface to ToA per Q = U A dT just like the insulated walls of a house with U the complex combination of conduction/convection/advection/radiation until 32 km where molecules cease and radiation alone takes over.

    • It gets worse. Back in 1976 James Hansen was a coauthor on a study named ” Greenhouse Effects due to Man-Made Perturbations of Trace Gases. The study was dealing with the trace GHG’s not the main ones of H2O,CO2,CH4 or N3. However it did describe some key details of how Hansen’s 1 dimensional computer model at that time did its’ calculations. I quote “The relative humidity of the atmosphere is kept fixed; thus if a change in the abundance of an atmospheric constituent increases the temperature, the absolute humidity also increases , causing a substantial positive feedback effect……….indeed it is possible that the relative humidity could increase somewhat with an increase in the average temperature of the earth.”

      So here we are 51 years after Manabe et al (1967) first proposed that relative humidity of the atmosphere remains constant as the temperature increases. Hansen incorporated it into his 1st computer model and it seems to have always been assumed.

      • And the odd thing is that everyone with any knowledge of meteorology knows that it is wrong. While the total amount of water in the atmosphere on average increases with temperature the relative humidity on average decreases.

  15. When we go to a fair and enter the fortune tellers tent, we don’t really expect to be told anything accurate, its a fun thing thats all.

    So why do our politicians expect anything different from the modern day climate scientists predictions, with their Crystal ball. Remember that computers are just number crunchers, and that the programmers are hoping for a result that will confirm their particulars beliefs.

    Of course if the result keeps the grant money coming then that is s a bonus.

    MJE

  16. What their plots are showing is not GB2, but a normalized GB2 adjusted to units of standard deviation (the vertical axis). The visual representation of their GB2 index in Figures 1 and 2 have been normalized 0 and Std dev set to 1, as indicated of course, in the Figure Legends. This normalization really alters how the geopotential height anomaly data actually looks.

    And then they apply smoothing.

    Figure 1 and 2, they use a 20-year smoothing filter.
    In the Supplement they show a variety of other smoothing periods and also no -smoothing.

    When you look at Figure S3, there is sharp downward spike in 2012. That also coincides with the sharp Arctic Sea Ice minimum that year. What the unsmoothed data tells me, is that in the larger context of the Arctic, 2007 was the transition year to recovering sea ice, and 2012 was a 1 off anomaly in that 10 years between 2007 and 2017.

    • The low years of 2007, 2012 and 2016 all suffered August Arctic cyclones. In 2016, two.

      The storms piled up and scattered the floes.

      But I agree that 2007 was the switch year. Summer sea ice minimum has been flat since then, and the trend is up since 2012. The six years after 2012 have averaged higher than 2007-12. No new lower low has been made since 2012. By contrast, between 1979 and 2012 in the dedicated satellite record, a new, lower low was recorded at least every five years.

  17. Determining for the weather of the UK and continental Europe are the Azores high, west of North Africa, and Iceland deep, northwest of the UK.

    the respective locations can be seen 10> <days around dormouse day ( Siebenschläfertag ).

    This year the Iceland low is closer to Ireland and the UK.

    The Azores high is closer to Africa and just south of the Azores.

    gives:

    – Constantly warm, dusty air from the Sahara that absorbs moisture above the Mediterranean, raining off while ascending the Alps and invading Central Europe as a dry "Foehn".

    – Constantly cold air from the northeast taking off moisture over the warm Gulf Stream.

    – where the two meet: hailstorm + ongoing heavy rains.

    It stays that way until the next dormouse day.

    __________________________________________________

    has nothing to do with climate change – that's weather!

    __________________________________________________

    hope you can cope with my dinglish.

  18. — SUBJECT- THE FLAWED MODELS.

    . The models are all flawed because they do not incorporate solar/geo magnetic effects. The only reason why this is not more apparent yet is because thus far the solar/geo magnetic fields have yet to reach threshold values of weakness which would result in a major rather then a minor climatic impact.

    Look at how wrong they have thus far and tis will only be getting worse.

  19. Study: Another failure of climate models – they can’t handle barometric pressure change

    Study: Another failure of climate alarmists – they can’t handle the truth

  20. Until now, no-one has systematically examined the projections to see how they represent the last few decades and future changes – up to the year 2100 – from a Greenland regional perspective.

    Well we know the models fail globally, that’s why IPCC AR5 set them aside in favour of “expert opinion” so why anyone would expect them to be accurate at the regional level is beyond me. But hey, let’s keep going:

    Hey Hanna et al! Could you do a few more regions “systematically”? Let’s see what happens…

  21. “…This deficiency may mean regional climate predictions for the UK and parts of Europe could be inaccurate, according to new research published today…”

    Wait…someone though regional climate predictions were going to be accurate in the first place?

  22. corrected northeast –> northwest:

    Determining for the weather of the UK and continental Europe are the Azores high, west of North Africa, and Iceland deep, northwest of the UK.

    the respective locations can be seen 10> <days around dormouse day ( Siebenschläfertag ).

    This year the Iceland low is closer to Ireland and the UK.

    The Azores high is closer to Africa and just south of the Azores.

    gives:

    – Constantly warm, dusty air from the Sahara that absorbs moisture above the Mediterranean, raining off while ascending the Alps and invading Central Europe as a dry "Foehn".

    – Constantly cold air from the northwest taking off moisture over the warm Gulf Stream.

    – where the two meet: hailstorm + ongoing heavy rains.

    It stays that way until the next dormouse day.

    __________________________________________________

    has nothing to do with climate change – that's weather!

    __________________________________________________

    hope you can cope with my dinglish

  23. “Dormouse day”? Not certain if this refers to putting dormice in teapots or extracting them. Or freshly roasted dormice as per QE1’s feasts.

    Perhaps someone could explain?

    • It’s the German version of the US Groundhog Day and a parallel to the English St Swithin’s Day

  24. As long as I have known, the weather models have been notoriously poor at forecasting the pressure over Greenland. They seldom properly represent the Greenland high pressure area. Now that hand analysis is pretty much a lost art, I suspect many meteorologists do not even look at the actually Greenland pressure pattern.

  25. The models do not calculate volcanic activity , solar activity ,geo magnetic field strength in other words they do not calculate the items that influence the climate therefore they are USELESS!

    In the meantime the global temperatures continue in a down trend. It is unfortunate we had a rise in overall oceanic sea surface temperatures but that seems to be subsiding some after a rapid blip up or at least I hope it was just a blip.

    El Nino not looking nearly as healthy as it was even as recently as a week ago. SOI INDEX should hold positive territory for at least the next week or so. I think -8 is the El Nino threshold which I say is 100% sure to hold over the next week

    As the magnetic fields continue to weaken and the duration of time lengthens I think this is going to translate more and more into a more dynamic effect upon the climate through an increase in major volcanic activity, global snow, cloud coverage increasing , overall oceanic sea surface temperatures cooling and a more meridional atmospheric circulation pattern.

    The thing that I can’t quantify is the magnetic threshold levels of weakness needed which would translate into a more dynamic effect upon the climate. However, I know it is out there.

    Let me try an analogy, I do not know if this is good or not, but say we are new to this planet and saw water in a liquid state for the first time. Say someone comes along and said if this liquid water gets cold enough it is going to reach a threshold level and turn into a solid substance. So the water starts to cool but nothing dramatic happens and the person that predicted something would happen to the water says it will happen but the threshold levels(32f) have not been reached yet. Everyone doubts it because it has not happened . He says in vain (he is right but does not know it) but the threshold level is out there.

    I think that is similar to what I am saying about the weakening magnetic fields , yes they are weakening but not to much has happened because the threshold levels have yet to be attained.

  26. You get hockey sticks because the predictions are all linear and based on the last observation. If there was an upward trend, all future predictions will be upward. Likewise a downward trend.

  27. “Previous work reported a tendency for global warming to result in a slightly more active jet stream in the atmosphere over the North Atlantic by 2100 but our results indicate we may actually see a somewhat weaker jet, at least in summer.”

    The science is settled then.

  28. The models are wrong with atmospheric pressure because the programmers think the jet steam will strengthen with warming. Also expected with global warming to develop increasingly positive NAO. (both wrong) These increasingly place the PV (polar vortex) over Greenland with low pressure. Large differences in temperature between polar air and sub-tropical air strengthens the jet stream. This tends to happen increasingly during a cooling planet and covers more of the planets surface because it is further away from the poles.

    Weak solar activity causes indirectly pressure to rise over Greenland because during these periods the jet stream weakens and becomes more meridional.

    The increase in pressure over Greenland occurred with weakening solar activity and had nothing to do with CO2 and/or Arctic sea ice because this would had happened well before. When these had for many years been increasing and decreasing accordingly. Though as usual with AGW anything happening even many years later is blamed on it. Nothing changed before, during or after it other than a decline in solar activity.

  29. Modellers’ fixation with radiation ignores the dominant energy flows towards Space. That is by Convective/latent heat uplift. Because they are not practical Physicists or Engineers neither they, nor the trolls here, can understaand that energy as with water, always must find the easiest pathway.
    The Ideal Gas Law nullifies any supposed special effects of Polyatomic molecules. This because of the relatively large space between gas molecules in the relevant operative PVT ranges.
    Warmista, and even Willis, would rather play around with models. They never make data, they just cannot. But as with Willis Thunderstorm idea, the Universe does “Integrate Data Empirically” (h/t Einstein) for open eyes to see. Recent bloggers above have been rightly pushing empiricism and I thank them.

  30. The importance of getting Greenland right for weather prediction was well known as far back as WW2. The Germans put remote weather stations in place at considerable effort.

    Why wasn’t this a priority for the modelers?

    • Because you talk of weather over a few days and GCM’s handle climate, over decades.
      A region of high pressure, an Anticyclone, especially one that is not semi-permanent does not affect climate.
      It affects weather.
      Weather is the movement of heat within the climate system.
      It is not the overall balance of it between absorbed solar SW and outgoing terrestrial LWIR.
      Climate deals with that.
      The Greenland high is not semi-permanent, else there would be no snow build-up on its ice-cap such as we’ve seen trumpeted here this last summer.

Comments are closed.