Climate Change Science

Compiled by Ken Gregory,
Calgary, Alberta, Canada
email: kgregory@shaw.ca
Revised March 16, 2013  (Revision history)

Table of Contents

          Introduction
          The Science in Summary
          Greenhouse Gas Effect
          Climate Is Always Changing
          CO2 - Temperature Correlation
          CO2 Changes Follow Temperature Changes
          Sun Activity Correlates with Temperature
          Sun and Cosmic Rays
          Milankovitch Cycles
          Heating of the Troposphere
          Stratospheric Cooling
          Warming on Other Planets
          CO2 Versus the Sun/Cosmic Ray Warming Theories
          CO2 Greatly Increases Plant and Forest Growth
          IPCC and Model Projections
          Computer Models Fail
          Water Vapour Feedback 
          Cloud Feedback
          Aerosols
          Climate Sensitivity
          The IPCC Hockey Stick
          Urban Heat Island Effects
          Falsified Historical CO2 Measurements
          No Consensus
          Effects of Warming
              Sea Level Rise
              Severe Weather
              Warming Is Good for Your Health
              Agriculture and Climate Change
              Warming Effects on Animals
         Kyoto Protocol - Misallocation of Funds
         An Inconvenient Truth
         Warnings of Global Cooling
          
Introduction

One of the goals of the Friends of Science Society is to educate the public through dissemination of relevant, balanced and objective technical information on the scientific merit of the Kyoto Protocol and the global warming issue. The science of climate change is complex. Unfortunately, politics and the media has affected the science. Climate research institutions know that they must present scary climate forecasts to receive continued funding - no crisis means no funding. The media presents stories of climate disaster to sell their products. Scientific research that suggests climate change is mostly natural does not receive much if any media coverage. These factors have caused the general public to be seriously misled on climate issues resulting in wasteful expenditures of billions of dollars in an ineffective attempt to control climate. This document gives an overview of climate change issues as determined by a comprehensive review of the state of climate science.


The graph above shows the temperature changes of the lower troposphere from the surface up to about 8 km as determined from the average of two analyses of satellite data. The best fit line from January 2002 to February 2013 indicates a decline of 0.03 Celsius/decade.  The sharp temperature spikes in 1998 and 2010 are El Nino events. Surface temperature data is contaminated by the effects of urban development. The Sun's activity, which was increasing through most of the 20th century, has recently become quiet. The magnetic flux from the Sun reached a peak in 1992. The high magnetic flux reduces cloud cover and causes warming. Since then the Sun has become quiet, however it continues to cause warming for about a decade after its peak intensity due to the huge heat capacity of the oceans. So we expect the warming to peak at about 2002. The green line shows the CO2 concentration in the atmosphere, as measured at Mauna Loa, Hawaii. The ripple effect in the CO2 curve is due to the seasonal changes in biomass. There is a far greater land area in the northern hemisphere than the south that is affected by seasons. During the Northern hemisphere summer there is a large uptake of CO2 from plants growing causing a drop in the atmospheric CO2 concentration.

The data are obtained from microwave sounding units (MSUs) on the National Oceanic and Atmospheric Administration's satellites, which relate the intensity or brightness of microwaves emitted by oxygen molecules in the atmosphere to temperature. The MSU data set represent the temperatures of a layer of the atmosphere that extends from the surface to approximately 8 kilometres (5 miles) above the surface. The data is from the University of Alabama in Huntsville and Remote Sensing Systems. 


The Science in Summary

The history of the Earth tells us that the climate is always changing; from warm periods when the dinosaurs flourished, to the many ice ages when glaciers covered much of the land. Climate has always changed due to natural cycles without any help from people.

The United Nations Intergovernmental Panel on Climate Change (IPCC) is a political organization promoting a theory that recent minor temperature increases may be caused largely by man-made carbon dioxide (CO2) emissions. CO2 is an infrared gas, and increasing concentrations can potentially increase the average global temperature as the gas absorbs long-wave radiation from the Earth and emits the absorbed energy. However, the warming ability of CO2 is limited because much of the absorption spectrum is near or fully saturated. When CO2 concentrations were ten times greater than today the Earth was in the grips of one of the coldest ice ages. The climate system is dominated by strong negative feedbacks from clouds and water vapour which offsets the warming effects of CO2 emissions.

The history of climate and CO2 concentration shows that temperature changes precede CO2 changes and can not be a significant driver of climate. Temperature changes over different time scales have been well correlated to solar cycles, cosmic ray flux and cloud cover. Recent research shows that cosmic rays act as a catalyst to create low clouds, which cool the planet. When the Sun is more active, the solar wind repels the cosmic rays, reducing low cloud cover allowing the Sun to warm the planet.

Computer model results presented in the IPCC Fourth Assessment Report predict that global warming will cause a distinctive temperature profile in the atmosphere of enhanced warming rate in the upper atmosphere at 8 to 12 km altitude over the tropics. The predicted temperature profile is the result of an expected increase in water vapour in the upper atmosphere which would amplify a CO2 induced warming three fold. The computer models are programmed to forecast a constant water vapor relative humidity with increasing CO2 resulting in a large water vapor feedback. Actual temperature data shows no such enhanced warming profile. Therefore, the comparison of observed data to computer models proves that no such water vapour induced warming amplification exists, so CO2 is not the main climate driver. In atmosphere layers near 8 km, the modelled temperature trend from 1980 is 200 to 400% higher than observed. Weather balloon data shows that specific humidity has fallen 9% since 1960 in the upper troposphere (400 mbar pressure level) where the models predict the greatest feedback. Adding CO2 to the atmosphere replaces a significant amount of water vapour, the most important greenhouse gas, resulting in only a small increase of the greenhouse effect.

An analysis of satellite data shows that clouds cause a strong negative feedback on temperature, but climate models assume that clouds cause a positive feedback. Modelers assumed that all cloud changes are caused by temperature changes which results in them inferring a positive feedback. But changing cloud cover can also cause temperature changes. Scientists can now separate these two effects. The correct analysis shows that clouds cause a strong negative feedback, so if temperatures increase, cloud cover increases, reflecting solar energy back to space and greatly reducing the warming effect of CO2 emissions.

Several planets and moons have warmed recently along with the Earth, confirming a natural sun caused warming trend. Over longer time periods, as the solar system moves in and out of the galactic arms the cosmic ray flux changes, causing ice ages and warm ages. A comparison of temperature and solar activity proxy data suggests that solar effects can explain at least 75% of the surface warming during the last 100 years.

CO2 is plant food and the increase in the CO2 concentration may have increased the global food production by 15% since 1950 resulting in huge benefits for people. For Canada, any CO2 warming effect would also benefit us by reducing our space heating costs and making a more pleasant climate.

The IPCC predicts that global average temperatures will increase by 0.17 to 0.38 ^oC per decade to the end of the century depending on the rate of CO2 growth in the atmosphere and other assumptions. The projections assume that no action is taken to limit CO2 emissions. However, these predictions are unrealistic because they falsely assume that the recent temperature changes are driven solely by CO2 and that the Sun has little effect on climate. A recent study of past climate change used by the IPCC has been shown to be wrong due to the use of a faulty algorithm, and the inappropriate selection of data.

The land temperature record is contaminated by the urban heat island effect. Fully correcting the land temperature record would reduce the warming trend from 1980 to 2002 by half. The IPCC historical CO2 record may be incorrect due to inappropriate adjustments to the ice core data, and ignoring direct historical CO2 measurements. The IPCC selects and adjusts data to conform to its CO2 warming hypothesis and ignores alternative climate theories. This is the wrong way to do science. Many scientists strongly disagree with the IPCC conclusions.

The sea level data shows no increase in the recent rate of sea level rise, and no such increase is expected over the next hundred years. There has been no detected increase in severe storms and there is no reason to expect an increase in the number or intensity of hurricanes resulting from any warming assumed to be from human caused CO2 emissions.

Any increase in temperatures due to human caused CO2 emissions will likely be beneficial to human health. The CO2 fertilization effect will increase the rate of forest growth and CO2 induced crop yield increases will reduce the pressures to cut down forests for farmland expansion. This will greatly benefit animals by slowing habitat destruction.

The benefits of CO2 emissions greatly exceed any likely harmful effects. Several authorities who have studied solar cycles have warned that the Earth may soon enter a cooling phase as the Sun is expected to become less active. The atmosphere may warm because of human activity, but if it does, the expected change is unlikely to be more than 0.5 C, and probably less, in the next 100 years.

Greenhouse Gas Effect






This graphic from Trenberth et al 2009 here, shows the exchange of energy among Space, the Sun, the atmosphere and the Earth. Greenhouse gases are primarily water vapour, carbon dioxide and ozone. Greenhouse gases are mostly transparent to incoming solar radiation, but absorb outgoing long wavelength radiation. The absorbed energy is then transferred to cooler molecules or radiated at longer wavelengths than the energy previously absorbed. This process makes the Earth warmer than it otherwise would be without the greenhouse gases (but with the atmosphere and clouds) by about 33 degrees Celsius.

Water vapour and clouds together account for over 70% of the total current greenhouse effect. However, in terms of changes to the greenhouse effect due to human activities, water vapour is generally considered a feedback and not a forcing agent. Computer simulations show the a uniform 1.8% change in water vapour has the same effect on outgoing longwave radiation as a 10% change in CO2 concentration. (See the water vapour feedback section for further information.)

Optical depth is a measure how transparent the atmosphere is to longwave radiation. More greenhouse gases reduce the transparency of the atmosphere to longwave radiation from the surface.

See here for a discussion of CO2 versus water's contribution to the greenhouse effect.

The top panel of the graph above shows the absorption spectral intensity of the greenhouse gases. Most of the short wave length solar radiation in the visible part of the spectrum is transmitted to the surface. Most of the upward thermal long wave radiation from the surface is absorbed except in the atmospheric window indicated by the blue region. About 16% of the long wave radiation is transmitted directly to space and the rest is absorbed by greenhouse gases. The middle panel shows the total absorption bands by wavelength of downward solar radiation and upward thermal radiation. The gray shading at 100 percent indicates that the energy is fully absorbed at that wavelength. The lower panel shows the absorption of the major greenhouse gases. Comparing the CO2 and H2O absorption spectra shows that much of the CO2 spectrum overlaps with that of water. Parts of the CO2 spectrum are already fully saturated. Adding more CO2 will result in ever diminishing effects as more of the available wavelengths become saturated. The temperature response to adding CO2 to the atmosphere depends on the amount of positive and negative feedbacks from water vapour, clouds and other sources. The temperature effect of increasing CO2 concentration is approximately logarithmic. This means if doubling the CO2 concentration from 300 ppm to 600 ppm, a 300 ppm increase, causes the temperature to rise by 1 ^oC, it would take another 600 ppm increase to add a further 1 ^oC temperature gain. Methane has an absorption band (at 8 micrometres) that largely overlaps with water vapour, so an increase in methane has little effect on temperature.

The above diagram shows the upward radiation spectrum from the top of the atmosphere at 20 km with 300 ppm CO2 and 600 CO2 as calculated by the Modtran radiative code. (Note that the horizontal axis of this diagram shows wavenumber, or number of wavelengths per cm, which is the reciprocal of the wavelength in micrometers used in the previous diagram.)  This model calculated radiation is very similar to what is actually measured by satellites from space. The green curve shows the emissions spectrum with 300 ppm CO2 in the atmosphere and the blue curve shows the spectrum with 600 ppm CO2 with the same surface temperature and water vapour profile. The model shows that doubling the CO2 concentration changes the spectrum only at the edges of the main CO2 absorption band, at 600 and 740 cm-1.  The resulting forcing of 3.34 W/m2 would cause the surface temperatures to increase if not offset by negative feedbacks.

Climate Is Always Changing

The Earth's history shows that the climate has always been changing, over both short-term and long-term time scales. These changes have sometimes been abrupt and severe, without any help from humans. Climate temperature reconstructions are determined from a variety of sources, such as from tree ring width studies and ocean floor sediments.  During the last 2 billion years, the Earth has alternated between cool periods like today, and warm periods like when the dinosaurs roamed the planet. The figure below on the left is a temperature reconstruction of the Earth over 2 billion years. Temperatures over this time frame are determined by mapping the distribution of ancient coals, desert deposits, tropical soils, salt and glacial deposits, as well as the distribution of plants and animals that are sensitive to climate, such as alligators, palm trees & mangrove swamps. See here for further information.

Temperature Over Geological Time
   

The above chart from here shows that CO2 levels have been declining since the end of the Jurassic period. The change in CO2 as indicated by the red line in the red circle is the change in CO2 since the industrial revolution.

                                             Holocene Optimum

The graph above shows the northern hemisphere temperature history since the last ice age.


                         Temperature History from North Atlantic Ocean Sediments
     

The graph above shows temperature variations of the past 3,000 years (during recorded history), as determined from ocean sediment studies in the North Atlantic. [Keigwin, 1996]. Note the rapid variations, as well as the much warmer temperatures 1,000 and 2,500 years ago. See here for further information.

A new temperature reconstruction with decadal resolution, covering the last two millennia, is shown above for the extratropical Northern Hemisphere (90?30?N), utilizing many palaeo-temperature proxy records, from Ljungqvist 2010 here. The shading represents 2 standard deviation errors.

RWP = Roman Warm Period AD 1-300; DACP = Dark Age Cold Period 300-900; Medieval Warm Period 800-1300; LIA = Little Ice Age 1300-1900; CWP = Current Warm Period 1900-.  The proxy data shows the parts of the Roman Warm Period and the Medieval Warm Period were as warm as the late twentieth century.

Climate is always changing, as the history of Europe's temperature over the last thousand years shows in the graph below:
                               

                1000 Years Temperature History IPCC 1990


The temperature history shown at the left was published in the first IPCC report in 1990, based on Lamb's estimated climate history of Central England.
Clearly, human activity could not have had a significant effect on the temperature changes before 1900. These changes are the result of natural processes.
See here.

See here for NASA's GISS temperature graphs since 1880.

HadCrut3 is the global surface temperature index produced by the Hadley Centre and the Climate Research Unit, England. It combines land and marine temperature data. The best line from January 2002 to December 2012 indicates a decline of 0.085 C/decade. The IPCC projected that temperatures would increase by 0.2 C/decade during this period.

CO2 - Temperature Correlation

The temperature of the Earth has warmed slightly, about 0.7 degrees Celsius, over the last hundred years. Over this time, CO2 concentration in the atmosphere has increased, mostly due to the increased use of fossil fuels. However, the Sun has increased in intensity since 1900 which may have induced much of the observed warming since then. Scafetta and West estimate that the Sun may have caused 10 to 20% of the increase in CO2 during the last century. (See page 2 of their paper here.)  A short-term correlation does not imply that the CO2 increase caused the temperature increase. Causation can be inferred if there is a correlation over several cycles of CO2 concentration changes, with the CO2 change preceding the temperature change. The actual climate history shows no such correlation, and there is no compelling evidence that the recent rise in temperature was caused by CO2. Temperatures have been variable over time, and do not correlate to CO2 concentration. When CO2 concentrations were 10 times higher than they are now we were in a major ice age. As a greenhouse gas, CO2 is vastly outweighed by (natural) water vapour and clouds, which accounts for over 70% of the greenhouse effect. Human-related CO2 emissions soared after 1940. Yet most of the 20th century's world-wide temperature increase occurred beforehand. See here for a graphic of the carbon cycle.

The CO2 annual growth rate to December 2012 is given below.
                                    







The actual increase of CO2 concentration averaged 0.5% per year since 1990.

                        


CO2 Changes Follow Temperature Changes

Fischer et al. (1999) examined records of atmospheric CO2 and air temperature derived from Antarctic Vostok ice cores that extended back in time across a quarter of a million years. Over this immense time span, the three most dramatic warming events experienced on earth were those associated with the terminations of the last three ice ages; and for each and every one of these tremendous global warmings, Earth's air temperature rose well before there was any increase in atmospheric CO2.  In fact, the air's CO2 content did not begin to rise until 400 to 1,000 years after the planet began to warm. Ice cores provide a detailed record of local temperature and CO2 concentrations. A study by Caillon et al. (2003) finds that the CO2 increase lagged Antarctic deglacial warming by 800 +or- 200 years. The authors measured the isotopic composition of argon40 and CO2 concentration in air bubbles in the Vostok core during the end of the third most recent ice age (Termination III), 240,000 years before the present. The argon40 isotope is found to be an excellent proxy for temperature.

                Vostok Ice Core Data over End of Third Ice Age BP
         The CO2 and Argon (Temperature) Age Scales are Shifted 800 Years










The CO2 concentration shown by the black line is plotted against age in years before present (BP) on the bottom axis, and the Argon40, a temperature proxy, shown by the grey line is plotted against age on the top axis. The age scale for the CO2 has been shifted by a constant 800 years to obtain the best correlation of the two data sets. The correlation shows that temperature changes precede CO2 concentration changes by about 800 years.

These findings confirm that an increase in CO2 has never initially caused an increase in temperature during a deglaciation. Temperature increases cause the oceans to expel CO2, increasing the CO2 content of the atmosphere. When temperature is at its maximum in each cycle and starts to fall, CO2 concentrations continue to increase for another 800 years!  As CO2 increases, temperatures fall. This is the opposite of what one would expect if CO2 were a primary climate driver. The ice core data proves that CO2 is not a primary climate driver. One must invoke reverse time causality to claim the ice core data shows CO2 causes temperature change, like suggesting actions taken today can affect the conquests of Mongol leader Genghis Khan. Logic demands that cause must precede effect. Increases in air temperature drive increases in atmospheric CO2 concentration, and not vice versa.

A more recent portion of the Vostok ice core record from Joanne Nova's Skeptics Handbook here is shown below.

See here for more information.  See here for a graph of Vostok ice core data.  See here for the Cailion et al (2003) paper.

Sun Activity Correlates with Temperature

Numerous papers published in major peer-reviewed scientific journals shows the Sun is the primary driver of climate change.  There is a very strong correlation between the Sun activity and temperature.

Early in the nineteenth century, William Herschel (1738-1822), discoverer of Uranus, found that five periods of low number of sunspots corresponded to high wheat prices when the temperatures were cold. (Cold climate reduces the supply of wheat causing its price to rise.) See here.

E. Friis-Christensen and K.Lassen have shown that the length of the mean 11 year Sunspot cycle correlates to the northern hemisphere temperature during the past 130 years.  The length of the Sunspot cycle is known to vary with solar activity, whereas high solar activity implies short sunspot cycle length. See here for further information.

See here for an updated plot based on Friis-Christensen and Lassen's methodology.

Here is a correlation of the sunspot cycle length, global temperature and CO2 concentrations.
                              Sunspot Cycle Length Temperature and CO2










The red squares on the graph represent the sunspot cycle lengths. One point is the cycle length from the time of the maximum number of sunspots to the time of the maximum number of sunspots of the next cycle, and the following point is the cycle length from the time of the minimum number of sunspots to the time of the minimum number of sunspots of the next cycle. The sunspot cycles are back filtered using weighting 1,2,3,4 applied to each cycle point, both min to min and max to max. This assumes that the current cycle has the most effect on temperature (weight 4), and previous half cycles affect current temperatures in declining amounts, but future cycles have no effect on the current temperature. The temperature curve in blue used the HadCRUT3 land and sea data to 1978, the MSU satellite data from 1984 to 2006, and the average of the datasets for 1979 to 1983. This eliminates much of the urban heat island effects. The temperatures are unfiltered annual. The CO2 concentrations (ppmv) from 1958 to 2007 are derived from air samples collected at the Mauna Loa Observatory, Hawaii. CO2 concentrations prior to 1958 are uncertain.

Note that there is a correspondence between sunspot cycle length and temperature. Both the temperature and the cycle length curves begin to rise at 1910, and temperatures fall after 1945 to 1975 when the cycle length curve falls, and both curves rise again after 1975.  Temperatures have been increasing since 1980 faster than can be explained by the sunspot cycle length, indicating a possible human CO2 contribution. The recent increase of the cycle lengths explains why there has been no warming since 2002. Temperature changes are expected to follow Sun activity changes due to a time lag resulting from the large heat capacity of the oceans.

N. Scafetta of Duke University, Durham, NC and B.J. West of the US Army Research Office, NC studied the solar impact on 400 years of the Northern Hemisphere temperatures since 1600. They find good correspondence between temperature and solar irradiance proxy reconstructions up until 1920 as shown on the graph below.

                    Northern Hemisphere Temperature vs Solar Irradiance 400 years











The temperature curve is derived from proxy records to 1850 by Moberg et al. [2005], and from instrumental surface temperature data from 1850 to about 1980. The surface temperature record includes the urban heat island (UHI) and land use changes effects. The Northern Hemisphere MSU lower troposphere record is shown from 1979 in blue, which eliminates most of the UHI effects. Two different solar irradiance proxy reconstructions are shown: Lean, 2000; Wang et al., 2005. Both curves merge the ACRIM satellite data since 1980 with the proxy data. By assuming ACRIM, the solar activity has an increasing trend during the second half of the 20th century.  This graph is modified from the version created by Scafetta and West, which uses the contaminated instrument record after 1979 instead of the satellite data.  See the original version here.

Note the low solar activity periods occurring during the Maunder Minimum (16451715, the Little Ice Age) and during the Dalton Minimum (17951825).

Note the excellent correlation from 1600 to 1900 when humans were unlikely to effect climate. During the 20th century one continues to observe a significant correlation between the solar and temperature patterns:  both records show an increase from 1900 to 1950, a decrease from 1950 to 1970, and again an increase from 1970 to 2000.

A divergence of the curves from the Scafetta and West original graph indicates that the Sun is responsible for 56% using Lean 2000, and 69% using Wang 2005, of the northern hemisphere warming from 1900 to 2005. The authors estimate the error at 20%.

There are two solar composites available from satellite data. The ACRIM is obtained directly from the satellite data, while the PMOD assumes that Nimbus7/ERB satellite data covering the ACRIM gap (19891992) are still significantly corrupted and require additional severe adjustments. The ACRIM data shows higher solar irradiance during solar cycle 22 - 23 than the PMOD data.  Using the PMOD data and the original graph, the Sun likely has contributed 50% of the surface warming from 1900 to 2005.

The authors did a similar analysis using the Mann and Jones 2003 temperature reconstruction. This temperature history shows little variation before 1900 and shows a hockey stick shape. This reconstruction has been severely criticized for several reasons. See The IPCC Hockey Stick section of this essay.  The authors found that the Mann and Jones 2003 reconstruction (when compared to the Lean 2000 data) results in an unphysical zero response time to solar forcing. The ocean's large heat capacity should result in a time lag of surface temperatures with respect to long time solar changes of several years, so this reconstruction cannot be correct.

The authors' analysis shows the Sun has contributed 50 to 69% of the surface warming depending on the reconstructions utilized. The remainder may be due to CO2, UHI and land use changes. The authors compare the Sun's irradiance to the Northern Hemisphere land surface temperatures, which are contaminated with the urban heat island effect. The global MSU satellite temperatures, which are not contaminated by the UHI effect, have increased by half as much as the North Hemisphere temperatures since 1980. If the Scafetta and West analysis used the uncontaminated satellite data since 1980, the results would show that the Sun has contributed at least 75% of the global warming of the last century. See more about the UHI effect later in this essay here. See here for the November 2007 article.

A group of NASA and university scientists have found convincing evidence of a link between the Sun activity and climate by comparing the records of the historical water level of the Nile River to the number of auroras observed in northern Europe and the Far East between 622 and 1470 AD. Auroras are bright glows in the night sky following solar flares, and are an excellent means of tracking solar activity.  See this link for further information.
      
A study by WJR Alexander et al, published June 2007 compared hydrometeorological data to solar variability. The study looked at rainfall, river flow and flood data. The authors conclude that there is "an unequivocal synchronous linkage between these processes in South Africa and elsewhere, and solar activity." The study included an analysis of the level of Lake Victoria, which has been carefully monitored since 1896. In the early 1960s a dramatic rainfall increase significantly raised the lake level, and the level since then has been falling at about 29 mm per year. The decline has been removed from the data plotted below.  The plot shows two periods of strong correlation between lake level and sunspot number, corresponding to periods of high levels of volcanic dust.

                                 Lake Victoria Water Level and Sunspot Number

See the paper "Linkages between solar activity, climate predictability and water resource development" here .

 Longer term, here is a correlation of a solar proxy to a temperature proxy for a period of 3000 years. Values of carbon-14 (produced by cosmic rays hence a proxy for solar activity) correlate extremely well with oxygen-18 (temperature proxy).  The lower graph shows a particularly well-resolved time interval from 8,350 to 7,900 years BP. 

The above graph summarizes data obtained from a stalagmite from a cave in Oman, as reported in the paper, Neff, U., et al. 2001.

A team of researchers led by scientists from the Max Planck Institute for Solar System Research analysed radioactive isotopes in trees and has found that the Sun has been more active in the last half of the 20th century than in any time in the last 8000 years. This study showed that the current episode of high solar activity since about the year 1940 is unique within the last 8000 years. See a press release here. A graph from the study is below. The bottom chart is a detail of the shaded period of the top chart from 9300 to 8600 years before the present.

Recently, Tim Patterson, an adviser to the FOS,  has studied high-resolution Holocene climate records from fjords and coastal lakes in British Columbia and demonstrates a link between temperature and solar cycles.










The spectral analysis shown here is from sediment cores obtained from Effingham Inlet, Vancouver Island, British Columbia. The annually deposited laminations of the core are linked to the changing climate conditions. The analysis shows a strong correlation to the 11-year sunspot cycle.

See here for a powerpoint slide show by Tim Patterson.


N. Shaviv and J. Veiser using seashell thermometers shows a strong correlation between temperature and the cosmic ray flux over the last 520 million years.
               
       Cosmic Ray Flux and Tropical Temperature Variation Over the Phanerozoic 520 million years












The upper curves describe the cosmic ray flux (CRF) using iron meteorite exposure age data. The blue line depicts the nominal CRF, while the yellow shading delineates the allowed error range. The two dashed curves are additional CRF reconstructions that fit within the acceptable range. The red curve describes the nominal CRF reconstruction after its period was fine-tuned to best fit the low-latitude temperature anomaly. The bottom black curve depicts the smoothed temperature change derived from calcitic shells over the Phanerozoic. The red line is the predicted temperature model for the red curve above. The green line is the residual.  The top blue bars indicate ice ages.

A paper by Nicola Scafetta (2011) compares the historical records of mid-latitude auroras from 1700 to the surface temperature records. It shows that auroras record share the same ocsillation frequencies evident in the temperature record and in several planetary and solar records. The author argues that the aurora records reveal a physical link between climate change and astronomical oscillations. The abstract states "In particular, a quasi-60-year large cycle is quite evident since 1650 in all climate and astronomical records herein studied ... The existence of a natural 60-year cyclical modulation of the global surface temperature induced by astronomical mechanisms, by alone, would imply that at least 60 to 70% of the warming observed since 1970 has been naturally induced. Moreover, the climate may stay approximately stable during the next decades because the 60-year cycle has entered in its cooling phase."See here .


Sun and Cosmic Rays

During the 20th century the Sun has continued to warm and may have contributed directly to a third of the warming over the last hundred years. The change in solar output is too small to directly account for most of the observed warming. However, the Sun-Cosmic Ray connection provides an amplification mechanism by which a small change in solar irradiance will have a large effect on climate.

A paper by H. Svensmark and E. Friis-Christensen of the Center for Sun-Climate Research of the Danish National Space Center in Copenhagen has shown that cosmic rays highly correlate to low cloud formation. Changes in the intensity of galactic cosmic rays alter the Earths cloudiness. 

An experiment in 2005 shows the effect of cosmic rays in a reaction chamber containing air and trace chemicals found over the oceans. Electrons released in the air by cosmic rays act as a catalyst in making aerosols. They significantly accelerate the formation of stable, ultra-small clusters of sulphuric acid and water molecules, which are the building block for the cloud condensation nuclei.

Danish scientists reported in May 2011 that they have succeeded for the first time in directly observing that the electrically charged particles coming from space and hitting the atmosphere at high speed contribute to creating the aerosols that are the prerequisites for cloud formation. In a climate chamber at Aarhus University, scientists have created conditions similar to the atmosphere at the height where low clouds are formed. This artificial atmosphere was irradiated with fast electrons from ASTRID Denmarks largest particle accelerator. The experiments show that increased radiation from cosmic rays leads to more aerosols. In the atmosphere, these aerosols grow into actual cloud nuclei in the course of hours or days. Water vapour concentrates on the nuclei forming small cloud droplets. See the news release here.

A team of 63 scientists published results in August 2011 of a much more sophisticated experiment which investigated the effects of cosmic rays on cloud formation. The CLOUD (Cosmics Leaving OUtdoor Droplets) experiment at CERN (European Organization for Nuclear Research) in Geneva show big effects of pions from an accelerator, which simulate the cosmic rays and ionize the air in the experimental chamber. The CLOUD experiment is the most rigorous test of the Cosmic Ray hypothesis yet devised. The experiments show that cosmic rays strongly enhance the formation rate of aerosols by up to ten fold, and confirms the earlier results from the Danish experiment. The aerosols may grow into cloud condensation nuclei on which cloud droplets form. See the CERN press release here.

The graph below shows the aerosol particle concentration growth in the CLOUD chamber. In an early-morning experimental run at CERN, starting at 03:45, ultraviolet light began making sulphuric acid molecules in the chamber, while a strong electric field cleansed the air of ions. As soon as the electric field was switched off at 04:33, natural cosmic rays raining down through the roof helped to build clusters at a higher rate. When CLOUD simulated stronger cosmic rays with a beam of charged pion particles starting at 4:58 the rate of cluster production became faster still. The various colours are for clusters of different diameters (in nanometres) as recorded by various instruments. The largest (black) took longer to grow than the smallest (blue). See here. The CLOUD results also show that trace vapours assumed until now to account for aerosol formation in the lower atmosphere can explain only a tiny fraction of the observed atmospheric aerosol production.

Coronal mass ejections from the sun cause a large decrease in the cosmic ray count, which are called Forbush decrease. These dramatic, short term cosmic ray decreases can be used to confirm the cosmic ray effects on clouds. The magnetic plasma clouds from solar coronal mass ejections provide a temporary shield against galactic cosmic rays.

See here.

Data from the International Satellite Cloud Climatology Project and the Huancayo cosmic ray station shows a remarkable correlation between low clouds (below 3 km) and cosmic rays. There are more than enough cosmic rays at high altitudes, so changes in the cosmic rays do not effect high clouds. But fewer cosmic rays penetrate to the lower clouds, so they are sensitive to changes in cosmic rays. 

                   Cosmic Rays and Low Clouds







The blue line shows variations in global cloud cover collated by the International Satellite Cloud Climatology Project. The red line is the record of monthly variations in cosmic-ray counts at the Huancayo station.

Low-level clouds cover more than a quarter of the Earth's surface and exert a strong cooling effect on the surface. A 2% change in low clouds during a solar cycle will change the heat input to the Earth's surface by 1.2 watts per square metre (W/m²). This compares to the total warming of 1.4 W/m² the IPCC cites in the 20th century. (The IPCC does not recognize the effect of the Sun and Cosmic rays, and attributes the warming to CO2.)

Cosmic ray flux can be determined from radioactive isotopes such as beryllium-10, or the Suns open coronal magnetic field. The two independent cosmic ray proxies confirm that there has been a dramatic reduction in the cosmic ray flux during the 20th century as the Sun has gained intensity and the Sun's coronal magnetic field has doubled in strength.

                      Cosmic Ray Flux Since 1700











Changes in the flux of galactic cosmic rays since 1700 are here derived from two independent proxies, 10Be (light blue) and open solar coronal flux (dark blue) (Solanki and Fligge 1999). Low cloud amount (orange) is scaled and normalized to observational cosmic-ray data from Climax (red) for the period 1953 to 2005 (3 GeV cut-off). Both scales are inverted to correspond with rising temperatures. Note that high cosmic ray flux around 1700 is at the end of the Little Ice Age. Also note the increase in cosmic ray flux after 1780 at the time of the Dicken's Winters.





The graph below shows a correlation between the cosmic ray counts and the global troposphere temperature radiosonde data. The cosmic ray scale is inverted to correspond to increasing temperatures. High solar activity corresponds to low cosmic ray counts, reduced low cloud cover, and higher temperatures. The upper panel shows the troposphere temperatures in blue and the cosmic ray count in red. The lower panel shows the match achieved by removing El Nino, the North Atlantic Oscillation, volcanic aerosols and a linear trend of 0.14 Celsius/decade.

The negative correlation between cosmic ray counts and troposphere temperatures is very strong, indicating that the Sun is the primary climate driver. H. Svensmark and E. Friis-Christensen published the above graph in a paper October 2007 in response to a paper by M. Lockwood and C. Frohlich, in which they argue that the historical link between the Sun and climate came to an end about 20 years ago. However, the Lockwood paper had several deficiencies, including the problem that they used surface temperature data that is contaminated by the urban heat island effect (see below). They also fail to account for the large time lag between long-term solar intensity changes to the climate temperature response.

See the Svensmark rebuttal of the Lockwood paper here, and a critique by myself here.

Over the 20th century the Sun has increased activity and irradiance intensity, directly providing some warming. The graph below from here shows the rising solar flux during most of the twentieth century.

                                  Open Solar Flux

Dr. U.R. Rao of Bangalore, India, shows that galactic cosmic rays, using 10Be measurements in deep polar ice as the proxy, has decreased by 9% during the last 150 years. The decrease in cosmic rays cause a 2.0% decrease in low cloud cover resulting in a radiative forcing of 1.1 W/m2, which is about 60% of that due to the CO2 increase during the same period. See here.

In the top panel showing cosmic ray intensity, the continuous line represents estimated Climax neutron monitor counting rate (1956-2000), open circles denote ionization chamber measurements during (1933-1956) and filled circles represent cosmic ray intensity derived from 10Be (1801-1932). 10Be is a long-lived radioactive beryllium isotope produced by cosmic rays. The middle panel shows the near-Earth helio-magnetic field and the lower panel shows the sunspot number.

A reconstruction of the near Earth heliospheric magnetic field strength from 1900 through 2009 from here by Svalgaard and Cliver (2010) is shown below.

The red curve are satellite direct measurements of the near-Earth heliospheric magnetic field (HMF) strength resulting from the solar wind. The blue curve is the Inter-Diurnal Variability (IDV) index calculated from the geomagnetic field observations one hour after midnight. The IDV is highly correlated with the near-Earth HMF. The green curve are estimates of HMF by Lockwood et al 2009.

When the Sun is active it has a higher number of sun spots and emits more solar wind - a continuous stream of very high-speed charged particles. The increased solar wind and magnetic field repels cosmic rays that otherwise would hit the Earth's atmosphere, resulting in less aerosols in the lower atmosphere thereby reducing low cloud formation. The low clouds have a high reflectivity and have a strong cooling effect by reflecting sunlight back into space.

In summary, the process is:
More active Sun  -->  more Sunspots --> more solar wind --> less cosmic ray --> less aerosols --> less low clouds --> more sun light to the surface --> global warming.

The theory of CO2 warming implies that the arctic and Antarctica should be warming about the same, and the polar regions should be warming more that the rest of the Earth. However, Antarctica has not warmed since 1975, which is a big problem for the CO2 theory.  The ice covering Antarctica has even higher reflectivity than low clouds, so fewer low clouds cools Antarctica, while fewer low clouds warms the rest of the planet. (Greenland's ice sheet is much smaller and is not so reflective.) This Antarctica temperature trend is strong evidence that the Sun, not CO2, is the primary climate driver.

          Antarctica and North America Temperature Trends











The top curve is the North American surface temperature and the bottom curve is the Antarctica (64 S - 90 S) surface temperature over the past 100 years. The Antarctic data have been averaged over 12 years to minimize the temperature fluctuations. The blue and red lines are fourth-order polynomial fits to the data. The curves are offset by 1 K for clarity, otherwise they would cross and re-cross three times.




The cosmic ray flux is not only influenced by the solar wind, it also varies with the position of the solar system in the galactic arms. The solar system passes through the arms of the Milky Way galaxy roughly every 140 million years. When the solar system is in the galactic arms the intensity of cosmic rays increases, as we are closer to more supernovas that give off powerful bursts of cosmic rays. The variations of the cosmic ray flux due to the solar system passing through four arms of the Milky Way galaxy during the last 550 million years is ten times greater than that caused by the Sun. The correlation between cosmic rays and temperatures over 520 million years by N. Shaviv and J. Veiser was shown previously. Below is a similar graph based on their work, but with the times of the galactic arm crossings shown.

    Cosmic Ray Flux and Temperature Changes with Galactic Arm Crossings



















Four switches from warm hothouse to cold icehouse conditions during the Phanerozoic are shown in variations of several degrees K in tropical sea-surface temperatures (red curve). They correspond with four encounters with spiral arms of the Milky Way and the resulting increases in the cosmic-ray flux (blue curve, scale inverted). (After Shaviv and Veizer 2003)



Temperature changes over this time range can not be explain by the CO2 theory.

                   CO2 Concentrations 500 Million Years





The graph shows CO2 concentration over the last 500 million years. The CO2 does not correlate with temperature. Note when CO2 concentrations were more than 10 times present levels about 175 million years ago and 440 million years ago, the Earth was in two very cold ice ages.




See here for a paper on CosmoClimatology by Henrik Svensmark.
See here for a discussion of the Shaviv and Veizer 2003 paper by Tim Patterson. See here for their paper. 

Milankovitch Cycles

The Earth-Sun orbital changes are the principal causes of long term climate change.  During the last 800,000 years, eight periods of glaciations have occurred. Each ice age lasts about 100,000 years with warm interglacial periods lasting 10,000 to 12,000 years. Milutin Milankovitch (1879-1958) identified three major cyclical variables which became recognized as the major causes of climate change. The amount of solar radiation reaching the Earth depends on the distance of the Earth to the Sun and the angle of incidence of the Suns rays upon the Earths surface. The Earths axis tilt changes on a 40,000-year cycle, the precession of the equinox changes on a 21,000-year cycle, and the eccentricity of the Earths elliptical orbit changes on a 100,000-year cycle.

The Earth's axis tilt (also known as obliquity of the ecliptic) changes from 22 to 24.5 degrees over a 40,000-year cycle. Summer to winter extremes are greater when the axis tilt is greater. The precession of the equinox refers to the Earth's wobble as it spins on its axis. Currently, the north axis points to the North Star, Polaris. In 13,000 years it would point to the star Vega, then return to Polaris in another 13,000 years, creating a 26,000-year cycle. When this is combined with the advance of the perihelion (the point at which the Earth is closest in its orbit to the Sun), it produces a 21,000-year cycle. The variation of the elliptical shape of the Earth's orbit around the sun ranges from an almost exact circle (eccentricity = 0.0005) to a slightly elongated shape (eccentricity = 0.0607) on a 100,000-year cycle.  The Earth's eccentricity varies primarily due to interactions with the gravitational fields of other planets. The impact of the variation is a change in the amount of solar energy from closest approach to the Sun (perihelion, around January 3) to the furthest distant to the Sun (aphelion, around July 4).  Currently the Earth's eccentricity is 0.016 and there is about a 6.4 percent increase in incoming solar energy from July to January. In the Northern Hemisphere, winter occurs during the closest approach to the Sun. The graph below shows the three cycles versus time. The vertical line represents the present, negative time is the past and positive time is the future. See here.

Analysis of deep-sea cores shows sea temperature changes corresponding to these cycles, with the 100,000-year cycle being the strongest.

These solar cycles do not cause enough change in solar radiation reaching the Earth to cause the major climatic change without an amplifier effect. A plausible amplifier is the Suns varying solar wind that modifies the amount of cosmic rays reaching the Earths atmosphere.

The rate of change of global ice volume varies inversely with the solar insolation due to orbital changes. The graph below compares the June solar insolation anomaly north of 65 degrees latitude to the rate of change of global ice volume over the last 750,000 years. Reconstructions of global ice volumes rely on the measurement of oxygen isotopes in the shells of foraminifera from deep-sea sediment cores. The records also in part reflect deep ocean temperatures. Two ice records are shown; SPECMAP and HW04.

The ice melting and sublimation rates are very sensitive to summertime temperatures. The strong correlations and the absence of a large time lag demonstrate essentially concurrent variations in the change of ice volumes and summertime insolation in the northern high latitudes. Both ice volume reconstructions therefore support the Milankovitch hypothesis and show that the Sun is the dominant climate driver. The graph is from a paper by G. Roe here.

Heating of the Troposphere

Computer models based on the theory of CO2 warming predicts that the troposphere in the tropics should warm faster than the surface in response to increasing CO2 concentrations, because that is where the CO2 greenhouse effect operates. The Sun-Cosmic ray warming will warm the troposphere more uniformly.

The UN's IPCC fourth assessment report includes a set of plots of computer model predicted rate of temperature change from the surface to 30 km altitude and over all latitudes for 5 types of climate forcings as shown below.

                           Computer Model Predicted Temperature Change














    The six plots show predicted temperature changes due to:
           a) the Sun
           b) volcanic activity
           c) anthropogenic CO2 and other greenhouse gasses
           d) anthropogenic ozone
           e) anthropogenic sulphate aerosol particles
           f) all the above forcings combined

The rate of temperature change is shown by the colour in degrees Celsius per decade.

It is apparent that plot c) of warming caused by greenhouse gasses is strikingly distinct from other causes of warming. Plot f) is similar to plot c) only because the IPCC assumes that CO2 is the dominant cause of global warming.

The computer models show that greenhouse warming will cause a hot-spot at an altitude between 8 and 12 km over the tropics between 30 N and 30 S. The temperature at this hot-spot is projected to increase at a rate of two to three times faster than at the surface.

However, the Hadley Centre's real-world plot of radiosonde temperature observations shown below does not show the projected CO2 induced global warming hot-spot at all. The predicted hot-spot is entirely absent from the observational record. This shows that most of the global temperature change can not be attributed to increasing CO2 concentrations.

                         HadAT2 Radiosonde Data 1979 - 1999

               
The left scale is atmosphere pressure in hPa. The right scale is altitude in km.
Source: HadAT2 radiosonde observations, from CCSP (2006), p116, fig. 5.7E
See Greenhouse Warming? What Greenhouse Warming? by Christopher Monckton

The graph below compares the global annual temperatures of the troposphere to the surface measurements. The lower troposphere measurements from the University of Alabama in Huntsville (LT UAH). It measures the temperature of the troposphere up to approximately 8 km. The HadCRUT3 curve is the Land and Sea-Surface Temperatures data set from UK Met Office. The GISS curve is the surface temperatures from the Goddard Institute of Space Studies. The three curves are scaled so that the average temperature of the first 5 years equals 0 degrees Celsius.

The graph below compares the annual temperatures of the troposphere to the surface measurements in the tropics. The lower troposphere measurements from the UAH is from 20 degrees North to 20 degrees South, and the surface temperatures from HadCrut4 is from 30 degres North to 30 degree South, and the surface temperatures from GISS is from 24 degrees North to 24 degrees South.


A comparison of the records show that the surface has warmed faster than the troposphere, the opposite of what is predicted by the theory of CO2 warming. Observations agree with the Sun-Cosmic ray warming theory.

The response of the troposphere temperatures in the tropics is sometimes called the fingerprint of the CO2 contribution to warming.


This graph shows two analyses of Microwave Sounding Unit (MSU) satellite temperature measurement data of the troposphere over the tropics from 20 degrees North to 20 degrees South. The UAH analysis is from the University of Alabama in Huntsville and the RSS analysis is from Remote Sensing Systems. The two analyses use different methods to adjust for factors such as orbital decay and inter-satellite difference. The overall trend lines to December 2012 show increasing temperatures at 0.06 C/decade for UAH and 0.11 C/decade for RSS. However, since January 2002, the temperatures have been declining at 0.16 C/decade for UAH and 0.24 C/decade for the RSS data. The IPCC projections do not agree with the data.

Stratospheric Cooling

The graph "HadAT2 Radiosonde Data 1979 - 1999" in the previous section shows that the stratosphere (above 16 km) has cooled, which might appear to indicate a greenhouse gas effect. However, stratospheric cooling is predicted to occur due to both greenhouse gasses and ozone depletion. The ozone concentration in the stratosphere has declined from 1970 until 1995, and has not declined at all since then due to the implementation of the Montreal Protocol, which limits the emission of ozone reducing CFCs. See here. The stratosphere temperatures are given below from here.


The lower stratosphere temperature has not declined at all since 1995 (when the ozone levels are stable or slightly increasing), so the data does not indicate any greenhouse gas cooling of the stratosphere. In fact, it appears that there has been a slight warming of the lower stratosphere since 1995, the opposite of what is predicted by computer models of the greenhouse gas effects. The stratosphere cooling indicated by the radiosonde data is caused by the changing ozone concentration, not by greenhouse gasses.


Warming on Other Planets

If the Sun is the primary driver of climate change, one should expect to see evidence of recent warming on other planets. As the Earth has warmed over the last 100 years, so too has Jupiter, Neptune, Mars and Pluto.

Jupiter is the largest planet in the solar system. Its most distinctive feature is the great Red Spot, which is a huge storm that has been raging for over 300 years. A new storm, called Red Spot Jr. has recently formed from the merger of three oval-shaped storms between 1998 and 2000. The latest images from the Hubble Space Telescope suggests that Jupiter is in the midst of a global change that can modify temperatures by as much as 10 degrees Fahrenheit on different parts of the globe. The new storm has been rising in altitude above the surrounding clouds, which signals a temperature increase. See here from Space.com.

Neptune is the furthest planet from the Sun (Pluto is now a dwarf planet) and orbits the Sun at 30 times the distance from the Sun to the Earth.
  
                                                   Neptune Warming













In the recent article, Hammel and Lockwood, from the Space Science Institute in Colorado and the Lowell Observatory, show Neptune has been getting brighter since around 1980; furthermore, infrared measurements of the planet since 1980 show that the planet has been warming steadily from 1980 to 2004.

In the figure, (a) represents the corrected visible light from Neptune from 1950 to 2006; (b) shows the temperature anomalies of the Earth; (c) shows the total solar irradiance as a percent variation by year; (d) shows the ultraviolet emission from the Sun. All data has been corrected for the effects of Neptune's seasons, variations in its orbit, the apparent tilt of the axis as viewed from the Earth, the varying distance from Neptune to Earth, and changes in the atmosphere near the Lowell Observatory.

See here for more information.

There is also strong evidence of global warming on Neptune's largest moon, Triton, which has heated up significantly since the Voyager spacecraft visited it in 1988. The warming trend is causing Triton's frozen surface of Nitrogen gas to turn into gas, making its atmosphere denser. See here.

A recent study shows that Mars is warming four times faster than the Earth. Mars is warming due to increased Sun activity, which increases dust storms. The study's authors led by Lori Fenton, a planetary scientist at NASA, says the dust makes the atmosphere absorb more heat causing a positive feedback. Surface air temperatures on Mars increased by 0.65 C (1.17 F) from the 1970s to the 1990s. Residual ice on the Martian south pole, they note, has steadily retreated over the last four years.  Thermal spectrometer images of Mars taken by NASA's Viking mission in the late 1970s were compared with similar images gathered more than 20 years later by the Global Surveyor.

Mars polar ice cap

See here or here or here for more information.

The demoted planet Pluto is also undergoing warming according to astronomers. Pluto's atmosphere pressure has tripled over the last 14 years, indicating rising temperatures even as the planet moves further from the Sun. See here for further information.

CO2 Versus the Sun/Cosmic Ray Warming Theories

The following table sets out a comparison of the predictions of two climate theories - the CO2 warming theory and the Sun/Cosmic Ray theory - and actual real world data.

CO2 Greatly Increases Plant and Forest Growth

CO2 is a major plant fertilizer. The increase in CO2 emissions have caused increased crop yields and faster growing plants and forests, thereby greening the planet. Estimates vary, but somewhere around 15% seems to be the common number cited for the increase in global food crop yields due to aerial fertilization with increased carbon dioxide since 1950. This increase has both helped avoid a Malthusian disaster and preserved or returned enormous tracts of marginal land as wildlife habitat that would otherwise have had to be put under the plow in an attempt to feed the growing global population. Commercial growers deliberately generate CO2 and increase its levels in agricultural greenhouses to between 700 ppm and 1,000 ppm to increase productivity and improve the water efficiency of food crops far beyond those in the somewhat CO2 starved atmosphere. CO2 feeds the forests, grows more usable lumber in timber lots meaning there is less pressure to cut old growth or push into "natural" wildlife habitat, makes plants more water efficient helping to beat back the encroaching deserts in Africa and Asia and generally increases bio-productivity.  See here

Bigtooth Aspen Growth Response to Enhanced CO2 and Temperature


Jurik et al. (1984) exposed bigtooth aspen leaves to atmospheric CO2 concentrations of 325 ppm and 1935 ppm and measured their photosynthetic rates at a number of different temperatures. At 25C, where the net photosynthetic rate of the leaves exposed to 325 ppm CO2 is maximal, the extra CO2 of this study boosted the net photosynthetic rate of the foliage by nearly 100%; and at 36C, where the net photosynthetic rate of the leaves exposed to 1935 ppm CO2 is maximal, the extra CO2 boosted the net photosynthetic rate of the foliage by a whopping 450%. These results are similar to studies of many other plants.

             Young Eldarica Pine Tree Growth Response to CO2


Young Eldarica pine trees were grown for 23 months under four CO2 concentrations and then cut down and weighed. Each point represents an individual tree. Weights of tree parts are as indicated. See here.

                                    Wheat Yield Response to CO2


















This graph shows the response of wheat grown under wet conditions and when the wheat was stressed by lack of water. These were open-field experiments. Wheat was grown in the usual way, but the atmospheric CO2 concentrations of circular sections of the fields were increased by means of arrays of computer-controlled equipment that released CO2 into the air to hold the levels as specified. Average CO2-induced increases for the two years were 10% for wet and 23% for dry conditions.

Since atmospheric CO2 is the basic "food" of nearly all plants, the more of it there is in the air, the better they function and the more productive they become. For a 300 ppm increase in the atmosphere's CO2 concentration above the planet's current base level of slightly less than 400 ppm, for example, the productivity of earth's herbaceous plants rises by something on the order of 30% (Kimball, 1983; Idso and Idso, 1994), while the productivity of its woody plants rises by something on the order of 50% (Saxe et al., 1998; Idso and Kimball, 2001). Thus, as the air's CO2 content continues to rise, so too will the productive capacity or land-use efficiency of the planet continue to rise, as the aerial fertilization effect of the upward trending atmospheric CO2 concentration boosts the growth rates of nearly all plants.

A 2003 study using 18 years (1982 to 1999) of satellite observations shows that global net primary plant production increased 6% over 18 years. The largest increase was in tropical ecosystems. Amazon rain forests accounted for 42% of the global increase in net primary production. See here.

The world's population is 6.6 billion and increasing at 1.18% per year. People will require increasing quantities of food and more natural ecosystems will be lost to crops and pastures. The resulting loss of habitat may result in species extinctions if crop yields are not significantly increased. Unfortunately, the rate of increase of crop yields is declining as crops are approaching the genetic yield limits. Increasing crop yields on existing farmlands would help to save lands for nature. If crop yields fail to increase, humans will suffer more frequent famines. Fortunately, the increase in CO2 concentrations will substantially enhance crop yields and is essential to prevent or delay the destruction of habitat and animal species, and may allow us to produce sufficient agricultural commodities to feed the growing population. Any action taken by us to slow or reverse the increase in CO2 concentration in the air may result in more frequent famines and species extinctions.

See here from CO2Science.

IPCC and Model Projections

Intergovernmental Panel on Climate Change (IPCC) presents projections of climate change, which are based on computer models. The projections given in the Summary for Policy Makers are based on six scenarios, which include different assumptions of population growth, economic growth, technological change and CO2 emissions. The scenarios assume that no climate change mitigation actions are taken, and they do not assume implementation of the Kyoto protocol. The IPCC does not assign any probability or likelihood to any of the scenarios, and the middle scenarios should not be interpreted as the most likely. The Fourth Assessment Report (AR4) used the Special Report on Emission Scenarios (SRES) projections of greenhouse gas emissions that was used in the Third Assessment report (TAR). The emission projections are converted to CO2 concentration projections in the TAR using the Bern climate-carbon cycle model (BERN-CC; Joos et al., 2001) that accounted for the climate-carbon cycle feedback (AR4 Ch 10, pages 750 and 790). The Bern-CC model gives a range of CO2 concentration at the beginning of 2100 for the A2 emissions scenario of 735 ppm to 1080 ppm, with the best estimate reference value of 836 ppm. Coupled climate-carbon cycle models simulate a range of CO2 concentrations for the A2 emissions scenario of 730 to 1020 ppm by the beginning of 2100. These results shows that there is a large uncertainty of the projected CO2 concentrations even for a given emissions scenario due to uncertainty in future changes in the carbon cycle.

The CO2 concentrations from the Bern-CC model were used as inputs to the IPCC AR4 climate models to calculate the projected warming of each emission scenario. The initial growth rate of the projected CO2 concentrations range from 0.48 %/year to 0.56 %/year. The CO2 concentrations of the six projections increased from 367 ppm at the beginning of year 2000 to a range of 540 ppm to 958 ppm at the beginning of year 2100. The table below shows the AR4 projections. The CO2 concentrations are the reference Bern-CC model values given here.

The temperature changes "Best Estimate" given in the second column are from the average surface temperatures in the period 1980 to 1999. The "Best Estimate" from 2006 given in the fourth column is reduced by 0.3 ^oC to account for the actual temperature change to 2006 from the average of 1980-1999. The average CO2 growth rates of the last two scenarios at 0.82 and 0.96 %/year appears to be unrealistic considering the actual CO2 growth rate 1990-2006 is 0.5%/year, and fossil fuels are expected to become more expensive as it becomes increasingly difficult to replace depleting oil and gas reserves. Note that the CO2 growth rate of the A1FI scenario increases from 0.51%/year in the 2000-2010 decade to 1.20%/year by 2060. The CO2 concentration projections corresponding to the six emission scenarios are shown below.

The graph below compares the actual CO2 concentrations as measured at Mauna Loa, Hawaii to the IPCC CO2 reference projections of the Burn-CC model for the period January 2000 to January 2010. The uncertainty in the CO2 projections is greater than the range shown here. The Low IPCC projection is from the B2 emission scenario, and the High IPCC projection is from the A1B emission scenario. The average IPCC projection is the average of the six scenarios. The actual annual CO2 growth rate in 2007 and 2008 were both 0.49%.

The IPCC temperature projections for the A2, A1B and B1 scenarios are displayed below.



Kevin Trenberth is head of the large US National Centre for Atmospheric Research and one of the advisors of the IPCC. Trenberth asserts ". . . there are no (climate) predictions by IPCC at all. And there never have been". Instead, there are only "what if" projections of future climate that correspond to certain emissions scenarios. According to Trenberth, GCMs ". . . do not consider many things like the recovery of the ozone layer, for instance, or observed trends in forcing agents. None of the models used by IPCC is initialised to the observed state and none of the climate states in the models corresponds even remotely to the current observed climate." However, Scott Armstrong and Kesten Green audited the relevant chapter in the IPCC's latest report. They find that "in apparent contradiction to claims by some climate experts that the IPCC provides 'projections' and not 'forecasts', the word 'forecast' and its derivatives occurred 37 times, and 'predict' and its derivatives occur 90 times" in the chapter. Consequently, it is not surprising that the public has this misimpression that the IPCC predicts future climate.

Computer Models Fail

The computer models predict that the 20th century temperatures should have increased by 1.6 to 3.74 Celsius, while the actual observed 20th-century temperature increase was about 0.6 Celsius. A model that fails to history match is useless for predicting the future.

The chart below compares the surface warming projections of the 2007 IPCC report to the actual global temperatures as represented by the HadCrut3 index. The red, green and blue curves are temperature projections from the A2, A1B and B1 emission scenarios. The orange curve is the temperature projection assuming the CO2 levels stay constant at the year 2000 value. The pink curve is the annual HadCrut3 actual temperature measurements. The black curve is the Fast Fourier Transform (FFT) best fit to the data.

See here. The quoted error on a single measurement is 0.05 C. The probablility that the IPCC projections overstate the warming in greater than 90%.

The IPCC Fourth Assessment Report projected a surface temperature increase from 1990 to 2100 of 1.4 C to 5.8 C, corresponding to 0.13 C/decade to 0.53 C/decade. The IPCC low estimate corresponds to the actual temperature warming rate as measured by satellite data.

Dr. John Christy presented to the US Senate on August 1, 2012 the following graph of the results from 34 climate models that will be used in the IPCC's fifth assessment report. The thick black line is the mult-model mean hindcast and projection from 1975 to 2020. The graph also shows the surface temperature observations and satellite observations adjusted to surface temperatures.

The graph shows that these new climate model results are wildly different from the observations. The satellite observations show the temperature has increased by 0.2 Celsius from 1980 to 2010, but the climate models mean increase is 0.6 Celsius. Surface observations show a 0.35 Celsius increase from 1980, but as explained elsewhere in this document, the surface temperatures are contaminated by urban development. The model mean temperature increase from 1980 to 2010 is three times higher than the satellite observations so the forecasts are useless for making policy decisions. Dr. Christy's presentation is here.

The IPCC assumes that the Sun has little effect, even though observational evidence clearly shows the Sun has a significant effect on climate.

The models assume the 20th century temperature rise is caused by CO2 increases, and parameters are set in the models to make the temperature rise in response to the CO2. The direct effect of increasing CO2 concentration on global warming is very small. All the models amplify an initial increase in temperature due to CO2 by employing water vapour and clouds as a large positive feed back. However, there is no evidence that water vapour and clouds provides a large positive feed back. They may provide a negative feed back.

The amount of solar energy the Earth recieves depends on the Earth's albedo, or reflectivity. The greater the albedo, the more sunlight is reflected and the less solar energy is absorbed by the Earth. Project "Earthshine" being done at the Big Bear Solar Observatory measures the Earth's albedo by observing the amount of sunlight reflected by the Earth to the dark side of the Moon and back to Earth. The process is shown below.



The results show that the Earth albedo has gradually fallen up to 1997, likely causing most of the global warming through 1998. Since 2001 the albedo increased rapidly, which has stopped the warming and resulted in the current global cooling. The recent dimming of the Earth is likely due to increased low cloud cover. The albedo is shown below.














The blue lines are the observed earthshine data for 1994-1995 and 1999-2003. The black line is the reconstructed albedo from partially overlapping satellite cloud data with respect to the mean of the calibration period 1999 to 2001. The vertical red line shows the cumulative climate forcing of the increase in greenhouse gases over the 20th century of 2.4 W/m2 according to the IPCC. Note that the change of the albedo's climate forcing in W/m2 is much greater than that due to greenhouse gases. Current climate models do not show such large albedo variability.  See an article by Anthony Watts here for further information.  See the project Earthshine site here.

Climate models utilize large grid blocks to simulate climate, which are too large to include thunderstorms or hurricanes, so they use parameterization to account for these. These parameterizations ignore real-world transfers of energy, moisture and momentum that could significantly alter the results and severely limits the usefulness of climate model projections. Computer models employ approximations to represent physical processes that cannot be directly computed due to computational limitations. Because many empirical parameters can be selected to force a model to match observations, the ability of a model to match observations cannot be cited as evidence that the model is realistic and does not imply it is reliable for forecasting climate.  See the Fraser Institutes Independent Summary For Policy Makers.

Atmospheric methane concentrations have been declining in recent years. Methane is a significant greenhouse gas. Climate models assume that methane concentrations increase with temperature, and it is not known why its concentration is declining. Aerosols play a key role in climate, with a potential impact of more than three times that of CO2 emissions, but their influence is very poorly understood. Aerosols exert an overall cooling effect on climate but estimates of the effect vary by a factor of ten. Models used in the IPCC Fourth Assessment Report assume aerosols have a large cooling effect, thereby attributing a large warming effect to CO2.

Only 2 of the 23 models used by the IPCC account for varying Sun intensity, and these models do not assume the Sun affects the cosmic ray flux and cloud formation. Only 2 of the models account for land use changes.

Computer models predict warming at the north and south poles to be symmetrical, but there is a warming trend at the North Pole but not at the South Pole. They also predict that the polar surface regions will warm more than the surface at the tropics. Winter temperatures will warm more than summer temperatures; night-time temperatures will warm more than day-time temperatures. Therefore, according to the CO2 warming theory, winter nights in the arctic will warm, but there will be little summer day time warming in the tropics.

A team of four researchers from three American universities led by David Douglass compared the troposphere temperature trends in the tropics predicted from climate models to actual satellite and radiosonde observations. In a paper published in December 2007 by the Royal Meteorological Society, Douglass et al analysed the simulation results from 22 climate models at the surface and at 12 different altitudes. The simulation results were compared to the temperature trends determined from two analysis of satellite data and four radiosonde datasets for the period January 1979 through December 2004.

                   Computer Model Temperature Trends versus Observations

The above diagram shows the comparison of temperature trends from 1979 through 2004 of climate models and actual satellite and radiosonde observations, expressed as degrees Celsius per decade versus altitude and atmospheric pressure. The left panel shows four radiosonde results as IGRA, RATPAC, HadAT2 and RAOBCORE. The thick red line shows the mean of the 22 computer model results, and the models' 2 times standard error of the mean are shown as the two thin red lines.  Temperature trends from three surface measurement datasets are identified in the legend by Sfc and are plotted on the left axis. The RSS and UAH analysis of satellite data are plotted on the right panel at two effective layers: T2lt represents the lower troposphere with a weighted mean at 2.5 km, T2 represents the mid troposphere with a weighted mean at 6.1 km altitude. A trend is the slope of the line that has been least-squared fit to the data. Synthetic model values corresponding to the effective layers of the satellite data are shown in the right panel as open red circles.

An essential place to compare observations with greenhouse computer models is the layer between 450 hPa and 750 hPa atmospheric pressure where the presence of water vapour is most important, and is called the "characteristic emission layer". In this layer, the observations are all outside the 2 times standard error test. The radiosonde and satellite trends are inconsistent with the model trends at all altitudes above the surface. Douglass et al. conclude that Model results and observed temperature trends are in disagreement in most of the tropical troposphere, being separated by more than twice the uncertainty of the model mean. In layers near 5 km, the modelled trend is 100 to 300% higher than observed, and, above 8 km, modelled and observed trends have opposite signs. Therefore any projections of future climate from the models are very likely too high, and these projections should not be used to form public policy. See the paper "A comparison of tropical temperature trends with model predictions" here.

A technical paper published by R. McKitrick, S. McIntyre and C. Herman in Atmospheric Science Letters, August 2010 shows that the climate model temperature trends of the mid-troposphere, using 57 runs from 23 climate models, are four times larger than observations from satellites and weather balloons.

See here for a discussion by D. Stockwell and see the technical paper here.

While air temperature may fluctuate from year to year as heat is transferred between the air and oceans, if CO2 is causing global warming by the IPCC hypothesis, the ocean heat content must increase monotonically provided there are no major volcanic eruptions. Ocean heat content is a much more robust metric than surface air temperature for assessing global climate change because the ocean's heat capacity is greater than that of the atmosphere by many orders of magnitude. For any given area on the oceans surface, the upper 2.6 m of water has the same heat capacity as the entire atmosphere above it! According the IPCC models, all major feedbacks are positive so there is no mechanism that would allow the heat content of the Earth to decline.     

Dr. Craig Loehle, Ph.D. has analyzed the ocean heat content for a linear trend over 4.5 years of data from mid-2003 to the end of 2007. The data shows an annual variation because most of the oceans are in the southern hemisphere. To eliminate the annual cycle, a model was fit with slope, intercept, and sinusoidal (1-year fixed period) terms using nonlinear least-squares estimation. The linear component of the model shows a decline of 0.35 x E22 Joules/year. (The graph shows the recalibrated data, after the data from certain instruments with a cool bias were removed. Initial Argo results showed strong cooling.) The Argo heat content is shown below. See his paper here.

William DiPuccio compared the projected ocean heat content of the GISS climate model to two analyses of the ARGO heat content data. The projected heat content of the GISS model was adjusted to include only the upper oceans for comparison to the ARGO actual data. He also calculated a lower limit by scaling the net global anthropogenic radiative flux to ocean surface area.  The observed ocean heat content trends were calculated by Josh K. Willis of NASA's Jet Propulsion Laboratory and Craig Leohle of the National Council for Air and Stream Improvement, Inc. Loehle's calculations have a smaller margin for error than Willis, because Willis only uses annual average data.

The heat deficit shows that from 2003-2008 there was no positive radiative imbalance caused by anthropogenic forcing, despite increasing levels of CO2. Indeed, the radiative imbalance was negative, meaning the earth was losing slightly more energy than it absorbed. The figure reveals a robust failure on the part of the GISS climate model.

William DiPuccio says, "Since the oceans are the primary reservoir of atmospheric heat, there is no need to account for lag time involved with heat transfer. By using ocean heat as a metric, we can quantify nearly all of the energy that drives the climate system at any given moment. So, if there is still heat in the pipeline, where is it? The deficit of heat after nearly 6 years of cooling is now enormous. Heat can be transferred, but it cannot hide." See his paper here.

Below is a graph which compares ARGO era (2003 to Q1 of 2011) the ocean heat content of the top 700 m from the National Oceanographic Data Center to the projections of the GISS climate model. The NODC OHC dataset is based on the Levitus et al (2009) paper which describes various adjustments and corrections to the data. The NODC data includes the ARGO data as described above and data from expendable bathythermographs. The GISS model projection is discussed here. The NODC data is here, and the graph is from here.

Note the enormous discrepancy between the measurements and the climate model projections.

The graph below shows th GISS-ER climate model 20th century hindcast (9 runs) and the projections (5 runs) compared to the NOAA observations. See here.

One of the most important parameters in determining climate sensitivity in climate models is the amount of heat they transfer to the oceans. The following graph by Dr. Spencer compares the Levitus observations of ocean warming trends during 1955-1999 to 15 IPCC 4AR climate model runs.

Note that the climate models exhibit wildly different trends, with the deep ocean cooling just as often as warming. The green curve is the Levitus actual observation to a depth of 700 m. Most of the models produce too much warming in the layer to 700 m. Many models produce unexpected ocean cooling below 100 m while the surface warms. None of the models even remotely match the observations. The weak ocean warming in the 700 m layer suggests low climate sensitivity, even if all the warming was due to CO2 emissions. See here.

The graphs below in this section prepared by Bob Tisdale compare temperature series to hindcasts of computer models used by the IPCC. Computer model hindcasts should be compare to the actual historical observations to determine how well the models matched the historical record. A model that fails to history match will not produce realistic projections.  

The animation below compares observed North Atlantic temperature anomalies to the modeled surface air temperatures for the 6 individual ensemble members and the ensemble mean of the National Center of Climate Research (NCCR) coupled climate model CCSM4. All data have been smoothed with a 121-month filter.

 Bob Tisdale writes "The NCAR CCSM4 coupled climate model appears to do a poor job of hindcasting the multi-decadal variability of North Atlantic temperature anomalies." See here.

The animation below compareas the sea surface temperature (SST) in the NINO 3 region to the climate model hindcasts. NINO 3 is a region in the Eastern Pacific tropics where El Nino events occur. It shows how poorly the models hindcast the frequency, magnitude, and trend of ENSO events. The model ensemble mean trend is 14 times greater than the trend of the observations.

Bob Tisdale writes "the frequency and magnitude of El Nino and La Nina events of the individual ensemble members do not come close to matching those observed in the instrument temperature record. Should they? Yes. During a given time period, it is the frequency and magnitude of ENSO events that determines how often and how much heat is released by the tropical Pacific into the atmosphere ..."

The graph below compares the linear trends for the observations and the model mean of the IPCC AR4 hindcasts/projections of SST for the period of January 1982 to February 2011 in 5-degree-latitude bands. The models predicted much greater warming trends in the tropics than what was observed. The actual warming in the northern regions is greater than modeled. Warming was predicted in the southern region but the SST trend was actually negative in much of the region. This shows that the models do an extremely poor job at simulating how topical heat is transported to the Polar Regions. See here.

The sea surface temperatures from -50 to -80 degrees latitude (south) and from 50 to 80 degrees latitude (north) are shown below. The IPCC claims that CO2 is the main driver of climate change, but the best fit linear temperature trends declined at 0.4 C/century in the southern region and increased at 2.0 C/century in the northern region despite the fact that the are CO2 concentrations in the two regions are the same.

The graph below compares the Easter Pacific SST to models by latitude. This includes the important El Nino region so a good history match here is critical. The Eastern Pacific tropical SST has declined at the equator at 0.14 C/decade, but the models show a strong warming of 0.19 C/decade. See here.

The graph below compares SST observations to climate model outputs for the period of 1910 to August 2011. The SST is from the HADISST dataset and the model hindcast is the IPCC model mean published in 2007. The models do not match the temperature variability in the period 1910 to 1975. They are made to match the warming trend from 1975 to 2002 by assuming most of the warming is due to CO2 and using high sensitivity to greenhouse gases. The projections diverge from observations after 2002 despite the continued increase in CO2 emissions. See here.

The graph below shows the northern hemisphere sea surface temperature measurements and the climate model hindcasts for the period 1910 to 1944. The actual temperature rise was 4.5 times greater than the modeled trend. The models cannot replicate the measurements because they do not include natural causes of climate change. The graph is from here.

The global surface temperature trend from HadCRUT for the early 20th century warming period 1917 to 1944 at 0.174 C/decade is similar to the late warming period 1976 to 2005 at 0.195 C/decade as shown below. But the net anthropogenice forcing in the climate models during the late warming period is 3.8 times higher than the forcing during the early warming period. The 3.8 fold increase in forcing had almost no effect on the temperature trends of the two warming periods, indicating that the theory of anthropogenic global warming is seriously flawed. The graph from here.

The graph below compares the 17-year (240 months) trends of the global SST to the IPCC model mean. Each point on curves represents the 17-year straight-line best-fit trend to that point in time. The IPCC models projected the global 17-year SST trend ending August 2011 at 0.15 C/decade, but the observed rise was only 0.02 C/decade. See here.

Bob Tisdale writes, "The coupled climate models used to hindcast past and project future climate in the IPCC 2007 report AR4 were not initialized so that they could reproduce the multi-decadal variations that exist in the global temperature record. This has been known for years." and  "The climate models used by the IPCC appear to be missing a number of components that produce the natural multi-decadal signal that exists in the instrument-based Sea Surface Temperature record."

The daily temperature range over land has been decreasing because the daily minimum temperatures (Tmin) has increased more than the daily maximum temperatures (Tmax) over the 20th century. The NOAA Global Historical Network database shows that 2/3 of the warming is due to the increase in the minimum temperatures. The trend of the difference between the maximum and minimim daily temperatures is called the Diurnal Temperature Range and it is a very important climate parameter. A paper by McNider et al (2012) shows that 6 climate models with published minimum and maximum temperatures replicate only 20% of the measured diurnal temperature trend as shown in the figure below. This is a five-fold climate model error.

Climate models are tuned to match only the 1970 to 2000 temperature rise of the average of the minimum and maximum temperatures (Tmean). If models are replicating Tmean but are not capturing the trend in Tmin, then this must mean that the model Tmax is warming faster than the actual Tmax. A computer analysis of the near surface boundary layer shows that an increase in greenhouse gases causes increased mixing of the boundary layer which brings warm nighttime air aloft down to the surface. Only 20% of the warming was due to longwave energy in the model simulation and 80% was due to increased turbulence. A layer only 20 to 50 m in thickness is warmed by this turbulence. The Tmax measured during the daytime represents a boundary layer 1 to 2 km deep. The climate models assume the Tmean represents an air thickness of 1 to 2 km, but it is actually only 20 to 50 m thick. The modeled Tmax is warming much faster and represents a much greater air thickness than the actual Tmax.  See here.

Most of the warming in climate models is due to increasing water vapour as temperatures rise. The climate models greatly overestimate the Tmax trend, which represents the deep atmosphere, so they greatly overestimate the increase in water vapour in the lower atmosphere as well.

About 50% of human-caused CO2 emissions are absorbed by natural sinks and 50% remains in the atmosphere. The fraction of emissions that are sequestered in sinks is called the sink efficiency. The graph below shows that as more CO2 is produced, the fraction of emissions that is sequestered in sinks has increased at 0.94%/decade.

Most of the models forecast the sink efficiency will decline so that the CO2 concentration in the atmosphere will rise by an additional 50 to 100 ppm by 2100 compared to a constant sink efficiency. But the actual sink efficiency change is in the opposite direction of the climate models so it is likely the CO2 content will rise slower that climate model predictions. A paper that discusses the climate model sink efficiency forecasts is here. The annual CO2 concentration data from Mauna Loa is here. The annual CO2 emissions data is here.

Many important inputs to climate models are very uncertain and real world observational evidence does not support them, so it is foolish to rely on their projections to make expensive policy decisions.

A scorecard listing the success of models is here.

Water Vapour Feedback

Relative humidity is the fraction of water vapour in a small parcel of air relative to the total amount of water vapour the air could contain at the given temperature and pressure. All the General Circulation Models, also known as Global Climate Models (GCM), just set various evaporation and precipitation parameters to achieve approximately the result:   Relative humidity = constant.

Box 8.1 of 4AR Chapter 8 page 632 states:

The radiative effect of absorption by water vapour is roughly proportional to the logarithm of its concentration, so it is the fractional change in water vapour concentration, not the absolute change, that governs its strength as a feedback mechanism. Calculations with GCMs suggest that water vapour remains at an approximately constant fraction of its saturated value (close to unchanged relative humidity (RH)) under global-scale warming (see Section 8.6.3.1). Under such a response, for uniform warming, the largest fractional change in water vapour, and thus the largest contribution to the feedback, occurs in the upper troposphere.

The assumption of constant relative humidity is not correct. Here is a graph of global average annual relative humidity at various elevations in the atmosphere expressed in millibars (mb) from 300 mb to 700 mb for the period 1948 to 2012. [Standard atmospheric pressure = 1013 mb. 1 mb = 1 hectopascal (hPa)] The data is from the NOAA Earth System Research Laboratory here.

 This graph shows that the relative humidity has been dropping, especially at higher elevations allowing more heat to escape to space. The curve labelled 300 mb is at about 9 km altitude, which is in the middle of the predicted (but missing) tropical troposphere hot-spot. This is the critical elevation as this is where radiation can start to escape without being recaptured. The average relative humidity at this altitude has declined by 21% (or 9.8 percentiles) from 1948 to 2012!

This is no logical reason to expect relative humidity to remain constant with increasing CO2 above the cloud layer. Relative humidity in a cloud is exactly 100% because the water droplets that make up the cloud are in equilibrium with the air. Likewise, relative humidity immediately above the oceans is 100%. The relative humidity in air parcels moving up over mountains will increase up to 100% causing rainfall. This saturation limit controls the average humidity in the atmosphere up to the top of the cloud layer. But the relative humidity at 400 mbars averages only 35% globally, or 30% in the tropics, and rarely gets anywhere near the saturation limit except in high thunderstorm clouds. The saturation limit therefore plays little role in determining the water vapour content of the upper atmosphere.

Doubling the amount of CO2 would increase temperatures by only about 1 degree Celsius if nothing else changed according to the IPCC. But the amount of water vapour will change in response to a CO2 induced temperature increase. Warmer air can hold more water vapour, so if relative humidity remains constant, the amount of water vapour increases with increasing temperatures. More water vapour, being a powerful greenhouse gas, would cause a further temperature increase, which is called a positive feedback. Most of the IPCCs projected warming is due to this water vapour feedback.

But the above graph shows falling relative humidity where the IPCC says changing water vapour content is most important. If relative humidity declines with increasing CO2 concentrations, the amount of water vapour in the upper troposphere may not increase, but might decline instead, resulting in a negative water vapour feedback.

Here is a graph of specific humidity, or the actual water vapour content, in grams of water vapour per kilogram of air, at the 400 mb level (about 8 km altitude).

This shows that the actual water vapour content in the upper troposphere has declined by 13.7% (best fit line) from 1948 to 2012 at the 400 mb pressure level. The climate models predict that humidity will increase in the upper troposphere, but the data shows a large decrease; just where water vapour changes have the greatest effect on global temperatures.

The NASA water vapour project (NVAP) uses multiple satellite sensors to create a standard climate dataset to measure long-term variability of global water vapour. The Heritage NVAP merges data from several satellites and radiosonde water vapour products for the years 1988 to 2001. The graph below left was presents at the GEWEX/ESA Due GlobVapour workshop March 8, 2011 here. Water vapour content of an atmospheric layer is represented by the height in millimeters (mm) that would result from precipitating all the water vapour in a vertical column to liquid water.

The graph shows a significant decline in global water vapour in the atmosphere layer from 500 to 300 hPa, about 6 to 9 km altitude.  The animation above right shows the amount of water vapor over the earth in the 500 to 300 mbar pressure layer. The Heritage NVAP global water vapour data (1988 to 2001) by layer is available from a NASA website here.  

The global annual average precipitable water vapour by atmospheric layer and by hemisphere is shown in the follow graph. The data in Excel format is here.

The graph is presented on a logarithmic scale so the vertical change of the curves approximately represents the forcing effect of the change. The water content of the L1 layer, surface to 700 mb, is about 20 times greater than in the L3 layer, 500 to 300 mb, whereas the forcing effect of a change in the L3 is approximately 14.5 times the same change in the L1. From 1990 to 2001, the water vapour changed by: L3   -0.55 mm, L2   -0.57 mm, L1  +1.73 mm.  The decrease in L3 is equivalent to an 8 mm reduction in L1. The water vapor decline in the L2 and L3 layers overwhelms the forcing effect of the water vapor increase in the L1 layer, so the water vapor feedback is negative. The upper atmosphere (L2 and L3) water vapor content of the southern hemisphere is less than, and has declined more than the water vapor content of the northern hemisphere.

The above graphs show the precipitable water vapour by layer versus latitude by one degree bands.