The Science of Predicting the Weather and Climate





In our solar system of planets orbiting the sun, only the earth finds itself in the middle of a narrow “Goldilocks” zone.  We are not too close to the sun and thus too hot, nor too far away and too cold, but rather we are at just the right distance for balmy day trips to the beach.   Even more remarkable is the fine-tuning of our orbit given the chaotic chemistry of the earth’s surface namely the delicate influence of oceans and atmosphere on global temperature.  In any event, the net result of this providence is to provide exactly the right environment for the exquisitely sensitive carbon chemistry of life.


If we do a basic physics calculation combining the solar irradiance, or energy output of the sun, with the earth’s distance from the sun, the earth should have an average temperature of about 0 degrees Fahrenheit (-18 degrees Celsius).  If this were the entire story, the oceans would have long since frozen and life would probably never have evolved.   Fortunately this simplistic calculation ignores the earth’s atmosphere and so is not what we observe.


What saves the environment for life is that the world’s oceans supply copious amounts of water vapor to the atmosphere.   Water vapor is our “greenhouse” gas which is transparent to the sun’s incoming visible light but which re-radiates outgoing infrared light partially back to the earth’s surface rather than letting it escape entirely out into space.  While a more difficult calculation, this gives us an average temperature for the earth with its current concentration of water vapor of more than 153 degrees Fahrenheit (67 degrees Celsius) [1].   Again, this isn’t what we observe but then this calculation assumes a static atmosphere and ignores the heat loss caused by convection currents to include evaporation and condensation, more commonly called the winds and the weather.


Rather it is the continuous convective churning of the atmosphere in the form of jet streams, thunderstorms, hurricanes,  clouds and breezes of all descriptions, as well as interactions in the form of “negative feedback loops”, that reduces our average global temperature [by a massive 94 degrees Fahrenheit] to a comfortable 59 degrees Fahrenheit (15 degrees Celsius) [2].    Basically things like “Hadley cells” move hot equatorial air from the surface to high altitudes and to the north and south poles.  This allows the atmosphere to radiate away more heat chaotically cooling the planet.


Unfortunately this final result cannot be calculated, i.e. from first principles or “ab initio”, but rather only measured.    This failure of science is not the result of any current, and perhaps correctable, ignorance of atmospheric physics; nor can it be corrected with faster computers or larger research grants.   Rather this unfortunate state of affairs comes from a gradual understanding beginning in the 1960’s that the short term weather, and thus the longer term climate, falls into a class of problems now called “chaotic.”  And it is the theoretical unpredictability of “chaotic-systems” that cause any forecasts to run smack -dab into a brick wall.   Science has now demonstrated unequivocally we will never be able to predict the weather beyond about 10 to 14 days [3].   This is the infamous “butterfly” effect that makes prediction not just difficult but theoretically impossible as described below.


Note that if it is scientifically impossible to build predictive models of the weather in the short term, then it is logically impossible to build predictive models of these same identical weather systems, or climate, in the long term.   And of course, any imagined difference between a short term weather model and a longer term climate model, which has the same identical physical properties and is described by exactly the same identical equations [i.e. Navier-Stokes], is non-existent and meaningless.  Hopefully it is also obvious that approximating theoretically unpredictable quantities doesn’t mean unknowable inaccuracies or statistical bias will ever cancel out.




Please note that besides water vapor there are several other trace “greenhouse” gases, most notably carbon-dioxide and methane, of which CO2, though nearly insignificant in total effect, is the more important minority constituent.


It is interesting to note, if not ultimately definitive, that the heat retention effect of a single CO2 molecule is less than about one-eighth that of a single molecule of H2O.   In this sense, and to first approximation, CO2 is not really a greenhouse gas at all [5a].


This is because in H2O the atomic arrangement is asymmetric so that the center of negative charge is displaced from the center of positive charge.   This forms a permanent electric field called a “dipole moment” that strongly interacts with heat in the form of microwave electromagnetic radiation.   In dramatic contrast, the atoms in CO2 are collinear without a permanent dipole moment (only a much weaker quadrupole moment).   So it is only when transient thermal vibrations jostle the carbon and oxygen atoms temporarily out of place that there is any pitifully weak interaction with the outgoing microwave radiation.


But this is not definitive because even the barely measurable trace amounts of CO2 now present absorb all of the earth’s outgoing infrared heat radiation that it is possible for them to absorb. And the massively greater amounts of water vapor perform the same function but in a much, much, shorter distance.   But what this does mean is that any infrared radiation that CO2 MIGHT have absorbed in various frequency bands is instead absorbed by water vapor.  This alone overwhelms CO2’s contribution making it even more insignificant, if that is still possible.


There are admittedly a few narrow absorption bands where only CO2 is active.  But these do not add more warming because heat absorption is not linear with concentration but rather logarithmic.   Fully half the total, but tiny, CO2 warming of the atmosphere comes from the first 20 parts per million (about 1/20th of current concentrations).  That is to say at each stage, to get the same heating as before, the amount of CO2 has to be repeatedly doubled.  Thus at this point adding more CO2 has almost no effect whatever. This is called “saturation” and is similar to applying more and more black paint to a window.  Beyond some point, more black paint can’t really make the room any darker.


In fact, the total effect of more CO2 is to slightly change the height at which water vapor, which is our only significant greenhouse gas, absorbs and retains heat.  This very slightly changes the area of the sphere which radiates heat away in to space.  And while we can make accurate predictive models of these effects for a STATIC atmosphere, this is BEFORE any considerations of the chaotic winds and weather which then massively reduce the temperature in theoretically unpredictable ways.


In rebuttal, alarmists wildly hypothesize that excess CO2 might have a significantly larger indirect effect by somehow triggering a POSITIVE feedback loop in the chaotic atmosphere, perhaps by increasing water vapor.  These imaginative musings, of course, are not supported by definitive calculations and, in fact, have not been observed by NOAA satellites over the last several decades.  Rather REAL WORLD observations unmistakably show less water vapor in the upper atmosphere, where it is more influential, again demonstrating the expected net NEGATIVE feedback loop [6].  This is in contradiction to all weather and climate model predictions whose wishful but arbitrary guess-work curve-fits, and arbitrary intentional tweaks to increase the predicted temperatures, are entirely divorced from the actual physics.


Please note that, because no transient temperature fluctuations have yet caused the earth to catch fire and melt, NEGATIVE feedback loops MUST be present as well as strongly influential, all other considerations aside.


Finally note that CO2 exists only in minute quantities, i.e. >0.04% for CO2 versus 1%-4% for water vapor, meaning there is as much as 100 times LESS CO2.    The combined result of these straightforward considerations is that CO2 from all sources only contributes about 2%-3% of earth’s total greenhouse effect.


That CO2 is such a rare trace gas in the atmosphere means it is one of the very few things humanity can influence but also to a barely detectable degree.  Indeed, the contribution from human activities, such as burning fossil fuels is swamped by natural forces acting across the globe.  Annual cycles of plant growth and decay, consuming and generating CO2, undeniably dominate all other effects.  Humans annually contribute about 4.4 ppm compared to natural emissions of about 114 ppm most of which from all sources is quickly re-absorbed by natural sinks for a net annual increase of 2 ppm. While these amounts vary from year to year, it is telling that net atmospheric CO2 increases do not track variations in human CO2 emissions but are strongly correlated to surface heating by El Nino events and to cooling by volcanic eruptions.  And indeed from the simplest of considerations, we can demonstrate that human contributions cannot contribute more than 1/3rd of the total annual CO2 increase[4].  And the more likely annual human contribution from a variety of calculations is less than 7%.


Also considering the diffusivity of CO2 in ice cores, so that spikes over only a few thousands of years are preferentially erased from the historical record, it is not possible to reasonably determine that current CO2 levels are unique or even unusual.  Nor does excess CO2 have a long residence time in the atmosphere before equilibrium is again achieved with the world’s oceans, having a decay time of less than about 6 years [4].


Again in rebuttal, global warming whacos and whacketts necessarily claim CO2 is not naturally eliminated from the atmosphere but instead remains and thus accumulates for hundreds to thousands of years in a hysteric rejection of the actual science.  A favorite mantra is that absorption into rocks is a slow process requiring millennia ignoring the uptake by vegetation and the oceans which is on the order of a few months to years.


Mankind’s total CO2 contribution since the dawn of time is thus about one molecule in 100,000 or roughly 10 parts per million (7% of the current 120 ppm increase from pre-industrial levels of 280 ppm).   The maximum upper limit of 1/3rd would only increase this to 40 ppm.   This gives us a range from 2.5% to 10% of the current extremely tiny trace amount of 400 parts per million of CO2, which in itself and in total contributes almost NOTHING to global warming.


These straightforward considerations and the time needed to double CO2 [more than 150 years but actually impossible because we will first run out of fossil fuels], mean that any putative human-caused temperature increase [i.e. anthropogenic] on a global scale is almost certainly TOO SMALL TO MEASURE [5b] (from the raw data made and published by the IPCC itself).


 Methane is a much more potent greenhouse gas but is so extremely rare (in parts per billion) as to be significantly less important than even carbon dioxide.




Modeling chaotic systems is impossible mostly because large increases in the accuracy of the initial conditions only slightly extend the time before the calculations, and chaotic reality, diverge.   The net effect is that any window of predictability asymptotically approaches a finite limit.  Thus there is a fixed time beyond which no prediction is possible regardless of the computing resources, regardless of the quality of the model, and even regardless of one’s knowledge of the starting conditions.  This coupled with the extreme sensitivity of chaotic systems to initial conditions with exponentially diverging trajectories, i.e. the oft-quoted “butterfly” effect [7], makes simple arithmetic averages useless.


One erroneous but oft expressed thought, is that if a model faithfully represents most of the physical sub-processes, then even though its trajectory through phase space diverges from what actually happens, e.g. model thunderstorms are ignored or go in the wrong direction, model daily temperatures are grossly inaccurate, model cloud cover is laughably off the mark, and so forth, nevertheless enough of phase space might be sampled to calculate about the same average values as for the real weather.


To some extent this forlorn hope is born out of those calculations of gases and gaseous mixtures which are possible. We can for instance calculate equilibrium relationships between static pressure, temperature, volume, and number of molecules for a gas.   We can and do model the smooth flow of air over an airplane wing and can even approximate turbulent flow.


The problem with the weather is that its non-linear and complex feedback loops trigger chaotic non-repeating trajectories through phase space which are in principle unpredictable.   This fact invalidates the possibility that simple arithmetic averages of temperature differences in any conceivable model with exponentially increasing errors will faithfully mimic the same non-linear processes in the real world.  That is to say, small differences in phase space trajectories cause diverging simple arithmetic averages over all time scales.


Aside from undeniable theoretical limits, all existing climate models are also unable to model, and thus choose to ignore, such effects as clouds, thunderstorms, hurricanes, and so forth, as well as their associated feedback loops.  These ignored effects, especially of clouds, amount to some 60%-80% of the earth’s total energy flow [8], which is what actually determines global temperatures.  Processes acting over short times at millimeter lengths are approximated by grids with spacings of many kilometers and time steps of many hours.  This is a problem for alarmists because predictions of catastrophic global warming assume model accuracies of many orders of magnitude (i.e. frequently calculating differences between incoming and outgoing energy flows which differ by as little as a part per million) beyond what any computer resources currently provide or are likely to provide in any foreseeable future.


Nevertheless, the alarmist argument is that non-linear, dynamic, and strongly interacting weather processes, e.g.  advection because of cloud cover or thunderstorms as a function of excess CO2, can be approximated by curve-fitting polynomials which are entirely divorced from the physical processes themselves.   Instead of using fine scale grids and the Navier-Stokes equations of basic physics, which would numerically create winds, clouds, rain, hurricanes, and temperature changes in the computer, INSTEAD entirely arbitrary [“informed” guess-work] polynomials are substituted to estimate the relationships between literally thousands of different weather processes each with its unknown adjustable magnitude.


Most of these relationships are not thought to be important and thus are omitted.   But not only are interaction strengths continuously debated, being increased ten-fold in some models and eliminated entirely in others, but entirely new weather PROCESSES themselves are being continuously introduced.   With the underlying principles of physics eliminated from weather and climate models, intuition rules and anything goes.


This technique of starting with unrealistically large scale Navier-Stokes [and neglecting clouds and hurricanes and most everything else which typically results in obviously nonsensical behavior], and then inserting adjustable but entirely arbitrary “fudge factors” for the most of the physical processes is called “parameterization.”  Repeatedly tweaking the necessarily large number of parameters will eventually cause the model to match historical records.  But if the CO2 concentration in the curve fitting exercise were replaced by a different independent variable such as the number of cheeseburgers consumed annually (which has also been increasing), the effect would be exactly the same.  And since some government weather stations, which were found in farmer’s fields decades ago,  are now located behind the exhaust outputs of fast food restaurants, cheeseburger counts might actually provide a better pseudo-linear fit to real physical processes (i.e.  the CHEEZEBURGER effect J).


The point is that ANY ARBITRARY continuous curve (to include increasing CO2 or cheeseburgers) can ALWAYS be mapped via Kalman-like filters or Monte Carlo approximations into ANY other arbitrary curve (by tweaking the parameters of an arbitrary polynomial).  This manifest and undeniable fact makes the assertion, i.e. that increasing CO2 levels are the only possible cause of changing temperature records, a really sick and expensive joke or, more to the point, totally and laughably and monstrously WRONG. 


Since computer models are in theoretical sense at best bad curve fits to CHAOTIC data and since curve fits are ARBITRARY in the extreme being entirely divorced from the actual physics and science of the phenomena, and since this is well known to every undergraduate majoring in science or engineering, it is difficult to understand why those who practice such deceptions to include the news reporting to the general public continue to parrot contrary non-sense.   This fact is so well known that even many decades ago John von Neumann made a crude joke about it noting “With four parameters I can fit an elephant and with five I can make him wiggle his trunk” [Attributed to von Neumann by Enrico Fermi, as quoted by Freeman Dyson in “A Meeting with Enrico Fermi” in Nature 427, (22 Jan. 2004), page 297.]


An associated problem is that there is no mathematical certainty any particular set of fudge factors is the best possible guess; rather there will likely be a very large number of such sets (i.e. an astronomical number of local minima in phase space).


But the definitive consideration is that historical temperature records show NO CORRELATION at all to observed CO2 levels; whereas other factors, e.g. sunspot cycles, show very strong correlations.  And this is true over geological time scales using proxy data as well as over the last two centuries from surface measurements as well as over the last 3-4 decades of satellite measurements.   Over geologic time, CO2 lags every major temperature change by up to 1000 years as warming oceans grudgingly give up dissolved gasses.  In real science, cause cannot come after effect.  Just like now, most of the CO2 increase [perhaps 93%] is because warmer oceans gradually release CO2 AFTER temperatures rise [i.e. NO causal correlation] [9].


The net result is that curve fits, which ignore the underlying physics, have no predictive power as inputs to the model change.   In any event, and as one should expect, all climate models created to date have been unable to predict temperatures beyond a few days [10], i.e. beyond the range used for “curve fitting” and either into the past or future.  Indeed for the entirety of the 21st century (since about 1998), no climate model has predicted the slight decline in global temperatures in the face of dramatically rising CO2 concentrations.   In real science, but perhaps not in DEMOCRAT-SCIENCE (DS), such predictive failures usually mean the models and the theories supporting them are WRONG. 




The public is not usually informed that the typically exaggerated global warming versus CO2 bandied about by alarmists in the popular press, only have any validity whatever for a STATIC atmosphere.   So to maintain both the hysteria and government funding, computer models of dynamic effects are tweaked to greatly magnify the initially insignificant temperature increases.


Basically the sign of CO2 feedback-loops is assumed to be positive despite the fact that the exact same feedback loops for H2O are negative and massive.  Please note that these are the models which consistently and universally fail to predict anything across any and all time scales.  [This is perhaps because the models have NO theoretical validity in the chaotic real world, because they are heavily parameterized and thus DIVORCED from the basic physics, and because they cannot even approximate what they do claim to calculate within several orders of magnitude. ]


Rather the definitive scientific truth is, that if one effectively takes a ruler and draws a straight line through chaotic data, be it historical temperatures or stock prices, that trend line has no predictive power either in theory or in practice.  In light of this cold hard fact, claiming climate models have any validity beyond numerical experiments to estimate coarse relationships between different crudely modeled weather processes is not only unscientific but willfully ignorant.


Claims to the contrary lack both theoretical and empirical foundation, except of course in politically inspired demagoguery.  Thankfully in a free country, one might still profess a blind faith in liberal politicians (full of DS, i.e. “democrat-superstition”) promoting human caused global warming for a variety of even good reasons [11-12], but it is not science.




1.    Our Affair With El Niño, chapter 7: Constructing a Model of Earth's Climate, page 105.



A major component is hot air rising at the equator which pulls in cold air from the poles which more effectively radiates heat into space.  These “Hadley cells” coupled with the “Coriolis effect” cause Northern hemisphere winds to be mostly from the north-west and Southern hemisphere winds to come from the south-west. 






5.    (a)




[ This data comes directly and without modification from IPCC Published Report TAR3, Chapter 6. Radiative Forcing of Climate Change: Section 6.3.4 “Total Well-Mixed Greenhouse Gas Forcing Estimate” ]


The precise calculations for global temperature increases in degrees Centigrade and Fahrenheit for cumulative increases in CO2 in parts per million for a STATIC atmosphere without weather and BEFORE  convection and before the observed negative feedback effects are included, are as follows [5b]:


CO2 ppm

Total °C

Extra  °C

Total °F

Extra °F
































Note that mankind’s total contribution is about 7% of the net increase over the last few centuries as CO2 amounts rose from about 280 ppm to 400 ppm.   The rest is due to outgassing from the world oceans (and only a little by democrat politicians).   This means mankind’s total maximum increase of global temperatures since the dawn of time is about [7% * 0.16°F] or 0.01°F.  If instead people contribute 1/3rd of the total CO2 increase, this would only be 0.05°F.  And this is before the observed negative feedback loops of the weather/climate causing conductive heat loss are considered.


Please recall that the negative feedback convective cooling of real world weather reduces STATIC atmosphere calculations of greenhouse heating for H20 from 153 [calculated] -> 59 [measured] degrees Fahrenheit. Similar effects MUST apply to CO2 and REDUCE the above heating even more and by about the same ratio.


Does anyone really think any of these putative temperature increases can be separated from the measurement noise of the last several centuries?


In rebuttal, alarmists claim the observed NEGATIVE feedback temperature reduction due the weather for H2O is somehow massively POSITIVE for CO2 for exactly the same processes.  This scheme increases the global temperature for a DOUBLING of CO2 [280 to 560 ppm] from a calculated 0.27 degrees Fahrenheit for a static atmosphere to a full ONE DEGREE and even more as they imaginatively tweak their weather-climate models.  In contradiction to ALL real world evidence, it’s hard to say something nice about this hallucinogenic fraud considering the massive waste of monies by government bureaucracies going to the veritable industry of fat cat politicians and friends.






8.     “The Physics of Atmospheres”, by John Houghton, Cambridge University Press, 3rd Edition, March 25, 2002 , ISBN-13: 978-0521011228 and ISBN-10: 0521011221, page 41;

which states in part:  “Clouds are, in fact, probably the dominant influence in the radiative budget of the lower atmosphere but adequately taking them into account raises many problems …”.

which states in part: “The cloud forming processes take place on fractions of a millimetre, while global climate models typically operate with a grid size of 50-100 km…  and the total global cloud cover reached a maximum of about 69 percent in 1987 and a minimum of about 64 percent in 2000…  These observations leave little doubt that cloud cover variations may have a profound effect on global climate and meteorology on almost any time scale considered “


“We are nowhere close to knowing where energy is going or whether clouds are changing to make the planet brighter ... We are not close to balancing the energy budget.”  Kevin Trenberth, National Center for Atmospheric Research, USA


“Basic problem is that all models are wrong - not got enough middle and low level clouds”.  Phil Jones, Director of Climate Research Unit, UEA, UK.


 “We only understand 10 percent of the climate issue. That is not enough to wreck the world economy with Kyoto-like measures”.  Henk Tennekes, former research director, Dutch Royal Meteorological Institute


“The fact is that we can't account for the lack of warming at the moment and it is a travesty that we can't”.  Kevin Trenberth, National Center For Atmospheric Research, USA


 “We can no longer absolutely conclude whether globally the troposphere is cooling or warming relative to the surface”.  Thorne et al, BAMS Oct 2005


“It is interesting to see the lower tropospheric warming minimum in the tropics in all three plots, which I cannot explain. is remarkably robust against my adjustment efforts”.  Leopold Haimberger, Department of Meteorology and Geophysics, University of Vienna


“But it will be very difficult to make the MWP [Medieval Warm Period] go away in Greenland”.  Henry Pollack, University of Michigan


 “The data doesn't matter. We're not basing our recommendations on the data. We're basing them on the climate models.”  Prof. Chris Folland, Hadley Centre for Climate Prediction and Research.

“I doubt the modeling world will be able to get away with this [tuning] much longer.”  Tim Barnett, Scripps Institution of Oceanography, USA


“The models are convenient fictions that provide something very useful.”  Dr. David Frame, climate modeler, Oxford University








“No matter if the science is all phony, there are collateral environmental benefits... Climate change provides the greatest chance to bring about justice and equality in the world”.

Christine Stewart, former Canadian Minister of the Environment as reported in Calgary Herald, December 14, 1998.


 “We've got to ride the global-warming issue. Even if the theory of global warming is wrong, we will be doing the right thing, in terms of economic policy and environmental policy.”

Timothy Wirth, former Democrat Senator from Colorado and Clinton Administration Under Secretary of State for global issues; quoted in Science Under Siege by Michael Fumento, 1993 and National Journal interview, 1990. .  Wirth now heads the U.N. Foundation which lobbies for hundreds of billions of U.S. taxpayer dollars to help underdeveloped countries fight climate change.


“A global warming treaty [Kyoto] must be implemented even if there is no scientific evidence to back the [enhanced] greenhouse effect”.

Richard Benedick, deputy assistant secretary of state, USA and representative to the IPCC Conference in Rio.


“We need to get some broad based support, to capture the public's imagination...  So we have to offer up scary scenarios (about global warming), make simplified, dramatic statements and make little mention of any doubts...  Each of us has to decide what the right balance is between being effective and being honest.Stephen Schneider, Stanford Professor of Climatology and lead author of many IPCC reports.


“Some people will do anything to save the earth . . . except take a science course”.  P. J. O'Rourke, author, journalist.


12.  “Hubris: The Troubling Science, Economics, and Politics of Climate Change by Professor Michael Hart.


“Again, it will take a determined effort by people of faith and conscience to convince our political leaders that they have been gulled by a political movement exploiting fear of climate change to push a utopian, humanist agenda that most people would find abhorrent. As it now stands, politicians are throwing money that they do not have at a problem that does not exist in order to finance solutions that make no difference. The time has come to call a halt to this nonsense and focus on real issues that pose real dangers. In a world beset by war, terrorism, and continuing third-world poverty, there are far more important things on which political leaders need to focus.”




Despite wild alarmist claims, no model has yet been able to predict global temperatures, especially for the last two decades.  Nor do measured global temperatures show any striking recent increase.   It is almost as if (but as one might expect) excess CO2 doesn’t correlate with global temperatures because the effect is too small to measure.   Perhaps the massive temperature increase predicted by alarmist models was stolen by garden/wood nymphs, or suddenly went into hiding at the bottom of the ocean and is still undetected [see graph of ACTUAL ocean temperatures below], or maybe, just maybe, was never there to begin with….  



Ipcc climate models versus actual temperatures





Global Surface Temperatures & Model Projections from 2002



Mauna Loa CO2