Why Climate Modeling Is Not Climate Science

Most everyone has heard the recent announcement that Global Warming has been put on hold for 20 to 30 years. Earth's variable climate continues to make fools of climate scientists, obstinately refusing to follow the IPCC's climate change script. Why? Because the climate change doomsayers put their faith in computer models, not in hard science.

Following a 30-year warming trend that started in the 1970's, global temperatures have leveled off and even declined since 2001 (see Greenland's Ice Armageddon Comes To An End). Despite rising greenhouse gas concentrations, and a heat surplus that should have caused global temperatures to continue rising, nature has pulled a thermal about face and sent temperatures downwards. Climate scientists are flummoxed.

“It is possible that a fraction of the most recent rapid warming since the 1970's was due to a free variation in climate,” Isaac Held of the US National Oceanographic and Atmospheric Administration (NOAA) wrote in an email to Discovery News. “Suggesting that the warming might possibly slow down or even stagnate for a few years before rapid warming commences again.”

“This is nothing like anything we've seen since 1950,” Kyle Swanson of the University of Wisconsin-Milwaukee said. “Cooling events since then had firm causes, like eruptions or large-magnitude La Niñas. This current cooling doesn't have one.” Why didn't the vast cadre of global warming researchers see this cooling trend coming? After all they have been predicting how Earth's climate will behave 100 years in the future, surely they can predict what will happen over the next decade or two?

The problem here is that all the predictions bandied about by the IPCC and various climate change celebrities are based on General Circulation Models (GCM). These are large, complex computer programs that attempt to model Earth's entire environment. A computer model is a simplified stand-in for some real system; a computer network, a protein molecule, the atmosphere, or Earth's entire climate. A modeler tries to capture the most important aspects of the system being modeled, while leaving off unnecessary detail. Of course, determining what's important and what isn't is the trick.

This means that a model is always less complicated than the thing being modeled, which in turn means that a model never behaves exactly like the thing being modeled. When you have a large complex system that can react in unexpected ways—called non-linear responses by mathematicians—the accuracy of models becomes even more suspect. In fact, it may be impossible to create an accurate model at all.

Modeling the atmosphere alone is complex, requiring knowledge of incoming solar radiation, the movement of air currents over the land and sea, heat convection, the amount of water vapor, the effects of clouds, and on and on. People have been trying to model Earth's atmosphere for decades, primarily to predict the weather. The weather forecasts you hear on your evening news are all based on computer models. How accurate are these models? In the near term, a few days from today, local weather forecasts are about 60% accurate when predicting high temperatures. For a GCM add in modeling of many more features as well. In fact, GCMs started off very simple and have continued to grow in complexity, if not accuracy, over the years.

Storm track prediction is an example of quantitative modeling, where the expected results of a model are hard numbers. In the case of hurricanes, a storm's track and changes in strength over time are the desired results. But Hurricane trackers get constant feedback from the real storms they track, allowing them to constantly correct their models when they go wrong. We have so little accurate climate data that such corrections are not really possible with GCM. When a GCM prediction diverges from reality is the model wrong, is it missing some critical factor that has been overlooked, or is the baseline data erroneous or non-representative? It is very hard to tell. Ignoring the known inaccuracies of GCM, scientists are still asking models to provide hard answers, something they just cannot do.

Climate change modeling is an example of qualitative modeling. These types of models result in general trends and overall effects of parameter modification. They are used to provide insight into processes where scientists' intuition fails. For example, a qualitative model can tell us that adding more CO2 to the atmosphere should cause warming. But qualitative models should not be used to make concrete predictions of future conditions, such as the global average temperature for the next 100 years. As Richard W. Hamming put it, “The purpose of computing is insight, not numbers.”

That, unfortunately, has not stopped the IPCC and affiliated climate alarmists from trotting out model predictions as science fact. In doing so they are committing the cardinal sin of computer modeling: believing that the model is the thing being modeled. Factually speaking, such numbers are at best a guess and at worst out right lies. So, why is Earth cooling down?

Russian scientists have been saying for years that we’re headed for a cold spell based on known solar cycles. Swanson credits the alignment of a series of climate processes for the cooling climate but what's actually causing the cooling is a mystery. Sinking water currents in the north Atlantic Ocean could be transferring heat into the ocean depths. Extensive low cloud cover in the tropics may be reflecting more of the sun's energy than usual back out into space.

Cloud cover variation would be in line with the historically low solar activity according to the predictions of Henrik Svensmark and Nir Shaviv. In 1997, Svensmark and Eigil Friis-Christensen popularised a theory that linked galactic cosmic rays and global climate change mediated primarily by variations in the intensity of the solar wind, which they have termed cosmoclimatology. This theory was later expanded on by Shaviv. The influence of cosmic rays on cloud cover has been ignored by the traditional climate science community since it downplays the impact of CO2 on global temperature.

In 2008, a new computer model developed by German researchers, reported in the journal Nature, suggested cooling will counter greenhouse warming for the next decade. However, this model predicts that temperatures will again be rising by 2020. More recently, the study in Geophysical Research Letters predicts global cooling for the next 20 to 30 years. But, that report also predicts that once the hiccup is over, “warming will return and be very aggressive.”

Why do all of the “main stream” climate researchers always add, “global warming will be back”? Because they are still using the outputs of the same models that didn't predict the current cooling trend. The same models that have never correctly predicted Earth's climate. Climate modeling has become a crutch for a previously ignored scientific community that has been thrust into the public light by the global warming scare. They don't have real scientific answers so they use their wonky models, hope for the best and keep asking for more grant money. It is time the public said enough.

Another article

If you found this article interesting you might want to check Extinction, Climate Change & Modeling Mayhem, which contains several examples of how seemingly minor changes to models radically change their results.

I know nothing about science.

I know nothing about science. I said all along, though, that there was nothing to global warming. Just a bunch of geeks jumping on a cause that never was.

Absence of falsifiability

There is a more fundamental reason why "climate modeling" is not "climate science." The reason that is offered for belief in anthropogenic global warming is the climate models that are referenced by the United Nations Intergovernmental Panel on Climate Change. According to an author of the IPCC's 2007 report, Kevin Trenberth "http://blogs.nature.com/climatefeedback/recent_contributors/kevin_trenberth/" 0, the IPCC models do not make predictions. It follows that: a) the IPCC's models are not falsifiable and b) the IPCC's models are not scientific models, by the definition of "scientific."

An appearance is created that the IPCC's models make predictions by the fact that they make what the IPCC calls "projections." A "projection" is a mathematical function that maps the time to the computed global average temperature. A prediction is a proposition about the outcome of a statistical event and a different kind of entity. While a projection supports computation of the error, it does not support falsification of the model.

An interesting line of reasoning

I have never quite thought of it in those terms. For myself, I have always thought that refusing to call the output of GCM predictions intellectually dishonest. The whole purpose for constructing a model and then running it forward in time is to gain insight into the modeled system's future behavior—in other words, a prediction. To call predictions projections and argue that you can not falsify the model based on them is semantic subterfuge and further indication of the fundamental weakness of the IPCC's case.

Climate Modeling

Hi Doug -

I came across your site by accident and have found it
most informative. Trained as a physicist, but working as
a software engineer, I find your approach to this problem
strongly resonant.

As I have only recently become more aware of the issues behind
the claims of "global warming" - curiously now being called "climate change" - I am not expert in the literature. Do you know if the various climate models do any back testing ?
Surely it is a necessary though not sufficient condition
that you be able to predict the past before one can have any
confidence that you can predict the future. Any models I have
worked on have had a "calibration" period and then a predicted
period both forward and backward in time. Even then, a model
can fall into the trap of being in effect a curve fitter
with no predictive value.

Regards,

JG Lipham

The two most common arguments

The two most common arguments against warming theories seem to be (1) local temperature variations (or mutually-inconclusive data) disprove global warming itself; and (2) models aren’t real science, anyway, so we don’t need to worry about them.

So, to be clear on this, are there *any* validated models that show significant deviation of climatic conditions from the consensus predictions regarding the effect of anthropogenic forcings?

Validation

In the correct sense of the word there are no validated climate models at all.

Following a 30-year warming

"Following a 30-year warming trend that started in the 1970's, global temperatures have leveled off and even declined since 2001 (see Greenland's Ice Armageddon Comes To An End). Despite rising greenhouse gas concentrations, and a heat surplus that should have caused global temperatures to continue rising, nature has pulled a thermal about face and sent temperatures downwards." - Never heard of this before, and I had Environmental Science last semester! We had number of examples from different parts of the world, Atlantic Ocean, Asian Coast etc. It is so confusing, and it is obvious that interest and politics is influencing science. Thank you for this awesome article.

Global Dimming

The main reason of cooling effect is global dimming

The incomplete combustion of fossil fuels (such as diesel) and wood releases black carbon into the air. Though black carbon, most of which is soot, is an extremely small component of air pollution at land surface levels, the phenomenon has a significant heating effect on the atmosphere at altitudes above two kilometers (6,562 ft). Also, it dims the surface of the ocean by absorbing solar radiation.[26]
Experiments in the Maldives (comparing the atmosphere over the northern and southern islands) in the 1990s showed that the effect of macroscopic pollutants in the atmosphere at that time (blown south from India) caused about a 10% reduction in sunlight reaching the surface in the area under the pollution cloud — a much greater reduction than expected from the presence of the particles themselves.[27] Prior to the research being undertaken, predictions were of a 0.5–1% effect from particulate matter; the variation from prediction may be explained by cloud formation with the particles acting as the focus for droplet creation. Clouds are very effective at reflecting light back out into space.
The phenomenon underlying global dimming may also have regional effects. While most of the earth has warmed, the regions that are downwind from major sources of air pollution (specifically sulfur dioxide emissions) have generally cooled. This may explain the cooling of the eastern United States relative to the warming western part.[28]
However some research shows that black carbon will actually increase global warming being second only to CO2. They believe that soot will absorb solar energy and transport it to other areas such as the Himalayas where glacial melt occurs. It can also darken Arctic ice reducing reflectivity and increasing absorption of solar radiation.[29]
Some climate scientists have theorized that aircraft contrails (also called vapor trails) are implicated in global dimming, but the constant flow of air traffic previously meant that this could not be tested. The near-total shutdown of civil air traffic during the three days following the September 11, 2001 attacks afforded a rare opportunity in which to observe the climate of the United States absent from the effect of contrails. During this period, an increase in diurnal temperature variation of over 1 °C (1.8 °F) was observed in some parts of the U.S., i.e. aircraft contrails may have been raising nighttime temperatures and/or lowering daytime temperatures by much more than previously thought.[25]
Airborne volcanic ash can reflect the Sun's rays back out into space and cool the planet. Dips in earth temperatures have been observed from large volcano eruptions such as Mount Agung in Bali that erupted in 1963, El Chichon (Mexico) 1983, Ruiz (Colombia) 1985, and Pinatubo (Philippines) 1991. But even for major eruptions, the ash clouds remain only for relatively short periods.

Source http://en.wikipedia.org/wiki/Global_dimming

The Importance of Aerosols

Some studies have attributed as much as 50% of total global warming in Asia to the infamous "brown clouds" that blanket much of that continent. As your comment shows, airborne particulates constitute a mostly unaccounted for factor in the global warming equation. This is another example of why the IPCC models are not accurate: they simply do not account for all of the factors we know about, much less the ones we have yet to discover.

impetus

If IPCC models are inaccurate, isn't that an impetus to push for better standards and better funding for scientific organizations? Criticizing without constructive plans for innovation them wouldn't do much good in modernizing the science of climatology.

shinn

The Constructive Plan is to do Real Science

As a practicing Computer Scientist, I have constantly reminded my students and colleagues that a model is never the same as the thing being modeled. This is a fact either unknown to or ignored by most climate scientists. If you want constructive advise here it is: do real science! Turn off your computers, get out of the lab and into the field. That's the only way to learn the facts about nature.

Why Climate Modeling Is Not Climate Science

Hello Doug,
I am trying to find out what "shape" is actually modeled by GCMs.
The first figure in the article interests me in particular.
Where I have got to, is probably best described on this thread, post 20.
http://www.globalwarmingskeptics.info/forums/thread-1039-post-8076.html#...
Any help / information greatly appreciated.

My "best guess" as to the "shape" modeled at present appears to be a disc shaped half cylinder.
I realise that the models are always depicted as spheres, but you can never get to the dark side of these depictions..
Are they really modeling 1/4 sunlight corrected for latitudinal inclination ?
ie, no night, no seasons, and much more omitted besides.

dalker at btinternet dot com

yours,
Derek.

Choatic Systems

Is it not true that the climate is a mathematically chaotic system? I believe it was Edward Lorenz who showed this in the 1950s ( the famous Butterfly Effect). As such only relatively small changes in any of the hundreds of climate variables can cause major changes in the climate. So it should not be a mystery as to why wierd weather occurs without apparent explanation. It is also proves why climate models that can at best be described as gross representations of climate reality have no predictive power. They are nothing more than tools to help scientists understand the interaction of variables.

Chaos rules

You are correct, when it comes to climate chaos rules. See my previous posts on chaos and related subjects: Tiptoeing Through The Tipping Points and Climate Models Irreducibly Imprecise.