Love him or hate him, it is worthwhile to understand where he is coming from, so I present this video: The emergent patterns of climate change
According to TED:
You can’t understand climate change in pieces, says climate scientist Gavin Schmidt. It’s the whole, or it’s nothing. In this illuminating talk, he explains how he studies the big picture of climate change with mesmerizing models that illustrate the endlessly complex interactions of small-scale environmental events.
Video follows, comments welcome.
The transcript is here: http://www.ted.com/talks/gavin_schmidt_the_emergent_patterns_of_climate_change/transcript
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
IanH says: “… they’ve never made a prediction that works, they have managed to *somehow* make their models predict the past. Oh I wish I could get paid to predict the past whilst claiming skill. …”
Exactly! They can predict the past and they can “adjust” the data sets, but they just can’t make their models predict the future. The hypothesis that CO2 will lead to catastrophic global warming is hogwash.
Most striking to me is Gavin’s presentation suggests that CO2 attribution studies are most likely very wrong. CO2 attribution studies argue that because “skillful” models can not explain current climate without adding CO2, then current climate must be due to added CO2. But by his own evidence the models fail miserably to capture the climate before the 1970s.
Stop his fast paced animation comparing observation and model results anytime from 1892 to 1970 and the patterns of warming and cooling are not in agreement. There is no skill. They have simply tweaked the models to match trends from the late 70s to now. Most likely, just as the ancient modelers added imaginary epicycles to match contradictory observations to advocate and protect their intellectual status, Gavin et al have simply added climate epicycles but failed to model global climate from 1900 to 1970.
“Skill is not measured that way.
Here is how you measure skill.
In 1938 imagine two people were asked the following question.
How warm will it be in 2014.”
@Steven Mosher,
Science doesn’t happen in a vacuum. It does have a history. And in 1938, atmospheric scientists said quite a bit – much of it was all over the place. Fast forward to the 1970s (the decade when scientists worldwide first began to accurately model the atmospheric as a whole). And many of them in places like NOAA and NASA projected cooling. Of course, they didn’t have the knowledge we have today. Yes, they knew about the Walker and Hadley Cells; they knew about various atmospheric pollutants, aereols, not to mention Greenhouse gases. Yet, there was no definitive understanding of the one of the globe’s primary atmospheric/oceanic oscillations -ENSO.
But, even with that knowledge, our main climate scientists insisted and continue to insist that ENSO and other oscillations are secondary to Greenhouse gases. Yet, their projections are not only off, but terribly so. Yes, they’ve to the trends correct. But, in weather forecasting that is just called persistence. You guys are just using persistence and then declaring victory. Based upon your model projections, it is fair to say that your understanding of our climate is poor. Someone has to write the code, and the code is just a reflection of your understanding.
Mosher,
What if your co2 guy was asked how warm will it get from 1997 to 2014? He would have failed. Give up on your crappy models.
Dr Norman Page says “The establishment scientists need take on board the fact that the Modeling technique is inherently useless for climate forecasting because models with such a large number of variables simply cannot be computed or indeed even initialized with sufficient precision and accuracy. “.
Dr Norman Page is absolutely correct. Climate is what happens over decades, centuries, millenia, and more. Think Ice ages. Think Al Gore’s famous graph showing peaks in temperature and CO2 about every 150,000 years. Think warming and cooling periods like the Roman and Medieval Warm Periods and the Little Ice Age. Think warm decades like the 1930s and cool decades like the 1960s.
There is nothing – absolutely nothing – in the climate models that explain any of the past climate events or predict the next one. The reason is that they work on the minutiae of weather. It is absolutely impossible to model a decadal or longer event whose cause is unknown by dividing the globe and time into small parcels and then trying to follow the short term impact of each parcel on its neighbours in space and time. Those models are not climate models, they are weather models (and not very good weather models at that). Even the best weather models cannot successfully predict more than a few days ahead.
Climate scientists need to admit : There are no climate models.
Pamela Gray says:
May 3, 2014 at 3:27 pm
Ok then – let me ask a really simple question – when you are in your garden, say, burning leaves, or a having a BBQ, how many times has the wind shifted? Was this predictable? Why not? Bear in mind, this is on a micro scale – i.e. your garden! – Now try and upscale this and imagine how multiplied any errors become. Sorry, but no GCM can model even the basic wind patterns as far as I am concerned – and I think the UK Metoffice proves this on a regular basis with it’s rubbish forecasting!. I’m not saying that GCM’s can’t have a use – but simply that computer modelling, as advertised, is not ANYTHING like it is hyped up to be – and WORSE – it is presented as being RELIABLE! All of which is false!
best regards, Kev
Anthony, when will you give a TED talk?
FUTURE temperature projection skill is all that matters and since the IPCC’s first report in 1990 they have failed time and again V OBSERVATIONS. Telling me how well they simulate anything in the past is garbage. They know what to tweak and voila!
Gavin:
This can also be described as a House of Cards.
Ross McKitrick after the 2011 Erice conference:
Source.
Gavin’s ‘big problem’ is that the central question to his view of climate change is HOW MUCH of a contribution are humans making to global warming.
As Gavin has noted on Real Climate, his answer to his central question of climate sensitivity has always been based on the response of the climate system to forcing that has occurred since the Last Glacial Maximum, some 20,000 years ago.
The question Gavin really wants to answer, therefore, is the physics of the interglacial cycles, or climate change over periods of > 20,000 years.
Yet, in his talk all he talks about is cloud resolution and ice crystals, mini tornados, dust from the Sahara, etc. All the stuff which has absolutely nothing to do with reducing the uncertainty of his central question.
It’s quite possible Gavin’s view of climate change is right, and also possible that we may never know how to predict “glacial maximum events”. Ignorance of the physics is not reason for inaction.
There is however, one good reason for inaction, and that is a cohort of over confident scientists playing down uncertainty, or even worse, not even talking about the real reasons why they could be wrong!
Gavin is a modern-day pseudo.
Here is a wonderful skill of the models. Back to the future!
Did you see that?
“The ability to predict retrospectively this slowdown not only strengthens our confidence in the robustness of our climate models”.
GIGO really does apply. Now I know why the computer guys came out with such a term. It all makes so much sense now.
These are the models that are so skilful that none of them projected the cessation of the Late 20th Century Warming Period, and when hindcast, fail horribly.
Odd used of the word “skilful”, it seems to me. Still, whatever gets Gavin through the night, poor lamb.
Kev, you bring to the table an inappropriate example (possibly from a misunderstanding?) of the purpose and output of general circulation models. They cannot predict or model the direction or strength of wind at ground level on a scale tiny enough to capture my garden. That is why their title includes the word “General”.
Steven Mosher May 3 1:50pm
What you say is silly nonsense.
First, if the same question had been asked only 17 years ago the naive answer would have been correct.
Second, long term trends were well known even back in 1939. It was recognized that the earth was coming out of the Little Ice Age. That the earth was warming was evident and because the earth had not yet warmed to the level of the Medieval Warm Period more warming would be indicated. A truly naive predictor exists only in your imagination.
Third, “skillful” is a sound byte word. It means “having or showing great skill”. Skill is “the ability to do something well”.So something that is skillful has shown the ability to do something well.
i suggest you compare the predictions made in the past using climate models and the real world events that occurred. They don’t match. Climate models are not skillful.
Eugene WR Gallun .
More modelling success!
The IPCC in AR5 has acknowledged this failure. Are we not wasting our time with the climate models? They are not fit for the purpose as shown since the IPCC’s first report. They are a load of dog’s dropping. Spend the money on the poor and homeless instead of rent seeking tricksters trying to sell us new electrickery.
Mosher,
Do you get paid in cold hard cash for advocating for the failed climate computer models. THEY HAVE FAILED! If these climate models were adapted and relied on by 10,000 investors in stocks then the fund management company would have gone bust. You can’t run an investment company since 1990 with a record of guaranteed failure. Yet this is what you are trying to sell me. I am not buying your stocks.
A lesson on humility and the ordinary man’s acute sense of a scam.
Ultimately the video is showmanship, pure showmanship to “sell” the climate model story. He ticks off on a bunch of successes modelling phenomena that occur at a large scale and the graphics coming from this sort of stuff is captivating and, when the numbers are good, compelling and inspiring. Trouble is for the past couple of decades the temperature numbers have not been good compared to the measurements, the measurements have a bit of a question mark over them ( as having been tweaked upwards) and it is well understood that the evaporo-transpiration system, formation of cumulus clouds, tropical and subtropical storm formation in particular happen at a relatively small, virtually local scale. This is true of most natural evaporation/precipitation phenomena that rely on extremely localised initiation to get going and at that point the models need fudge factors.
The models can handle the physics of most of the phenomena with reasonable and useful accuracy but not all. In other fields other methods are resorted to at this point because it is known that fudge factors do not deliver realistic let alone reliable results. It requires an arrogance driven by influences on the human condition other than objective, scientific enquiry for its own sake to present these models as a ‘one stop shop’ for your climate reality. It seems to me that at this point in time we are in a search for honesty and credibility much more than scientific knowledge on this subject.
Mr Schmidt’s talk was generally informative and one would only have to change the last few minutes to give a proper state of the science to a lay audience but ultimately, without the necessary caveats and admissions of work still to be done, it misrepresents the models and their accuracy and simply seeks to blind the audience with “science” and mere spectacle.
Steve McIntyre says:
May 3, 2014 at 2:33 pm
Steve, do I understand you correctly? You seem to me to be saying that modelers generally don’t investigate low to no feedback because of their parameter choices. I wasn’t under the impression that this was a parameter, I thought the feedbacks in the models arose somehow from the algorithm being iteratively applied.
I’d love to hear more about this. Thanks all.
Here are a few model results and climate predictions from global warming. They show our current state of understanding is very high indeed for the past, present and future. We are really doomed.
Could someone ask Gavin where the model-predicted ‘hot spot’ is?
If this is a wrong prediction, of what value are the models?
Jimbo says:
May 3, 2014 at 4:28 pm
Here are a few model results and climate predictions from global warming.
====
I love that 😉
You should have posted that next after Mosh……
You guys seem to all be forgetting…..all of these computer games are based on the “official” temperature history reconstructions….
…and they were all jiggered
Jiggered in…..jiggered out
Even their obvious trend can only reflect the trend they fed into it…
Steven Mosher says:
May 3, 2014 at 1:50 pm
Skill is not measured that way.
Rubbish. All of it.
Skill is not measured by the artifice of setting up a straw man argument and then knocking it down.
Let’s try a realistic perspective instead. You can call it ‘skeptical’, if you like:
“The earth has been warming in fits and starts for approximately the last 11,700 years, allowing mile thick glaciers to slowly melt and recede world wide. Realizing atmospheric CO2 gas concentrations in the range of 250 – 1000ppm has only a tertiary influence on ‘global warming’, we can anticipate that the earth’s atmosphere will continue to warm in fits and starts until the true drivers of this Holocene interglacial warming (natural drivers as yet undetermined with certainty) wane and the planet again starts the inevitable slide into another massive 100,000 year long glaciation period again. This is a well established natural cycle which atmospheric CO2 has little ability to prevent, more is the pity! Would that it could, for it will be woe unto mankind, when the cold embrace of glaciation again asserts itself on the vast areas that are currently inhabited by a panoply of all God’s flora and fauna. ”
That is the ‘skeptics’ prediction, based on real observations of our planets natural cycles.
Nevermind my question, AR4 WGI 8.6.2.3 looks like it talks about this. Different models exhibit different feedbacks, water vapor and lapse rate and cloud feedback and so on.
What I didn’t realize (but probably should have) is that modelers haven’t explored low feedback parameters. That’s just sad.
Steve McIntyre said:
“it’s unfortunate that the modeling community have failed to fully map the parameter space and left low-to-no feedback largely as a terra incognita.”
When skeptics, convinced that H2O feedback is terra incognita, and that the climate sensitivity is low, look at the scenarios painted by the successive IPCC model runs, we see the general problem and conclude the models are crap.
Steven Mosher says:
May 3, 2014 at 1:50 pm
=================================
Couldn’t a similar argument made about speaking to a Viking in Greenland in 1033, with regards to the temperature in 1700 and be “not skillful” (layman here – go easy please)
TR