This graph also shows the highly fudged HadCRUT4 surface temperature "measurements" adjusted in a desperate attempt to minimize its disagreement with the many climate model predictions of a steady and rapid global surface temperature increase. The UAH Lower Troposphere measurements are also shown. These satellite measurements are more reliable measurements, though they are not actually surface temperature measurements. Almost all of the CMIP5 computer models of the climate are predicting substantially higher temperatures than are measured. This is an additional indictment of the idea that the climate models represent settled science.
Among the inputs into these models are the man-made or anthropogenic forcings shown below:
Now examine the error bars closely on these anthropogenic effects. The error bars on the stratospheric ozone effect are several times bigger than the expected effect. The error bar on the tropospheric ozone effect is as big as the expected effect. The error bar on the stratospheric water vapor effect is as big as the expected effect. The same is true of the land use and black carbon on snow effects. The error bar on the direct aerosol effect is as big as the expected effect, while that on the cloud albedo effect is larger than the expected effect. The error on the linear contrails is larger than the expected effect, as is that for what is here called the solar irradiance effect. According to the analysis of errors used for this graph, the net anthropogenic component is anywhere between 0.60 and 2.40 W/m2 . The settled science is so settled that the effect that is so widely claimed to be of a catastrophic scale may be a mere 0.60 W/m2 or it may four times greater.
Note that these inputs are only those in which man has played a role. The uncertainties relating to many natural effects are no smaller. Indeed, many natural effects are not included in the models at all because they are too poorly understood to be included!
It sure is hard to produce a plausible settled science from a great number of effects which are clearly not well-understood individually and which are not necessarily stochastically independent. Indeed, it takes some knowledge of an effect's science and magnitude to actually make an informed estimate of error. With errors this large, it is clear that there is insufficient knowledge to even estimate the errors accurately because they imply a massive lack of understanding.
Some time ago, we learned that the global climate models do not have sufficient spatial resolution to properly model convection, which is a critical effect I have long noted was underestimated in the catastrophic man-made global warming hypothesis. See http://hockeyschtick.blogspot.com/2013/09/new-paper-finds-ipcc-climate-models.html Dr. Noor van Andel explained that the human fingerprint so long predicted in the climate models was actually an artifact of the models not properly handling convection.
It is also known that the computer models have up to 30% errors relative to the observed absorption of solar insolation by the atmosphere. This is a huge error which at least in part is due to inadequate knowledge of the absorption of near ultraviolet radiation by water vapor. See http://hockeyschtick.blogspot.com/2013/09/new-paper-finds-up-to-30-discrepancy.html
I have long claimed that the catastrophic man-made global warming hypothesis was based on a number of violations of energy conservation. It turns out that the actual computer models find still more ways to violate energy conservation. Some of them concentrate energy in regions in a manner to violate both the Second and the First Laws of Thermodynamics, making nonsense of the alarmist claim that the computer models simply incorporate well-known physics. It was found that the water vapor fluxes of some models are wrong by up to 10% due to the model causing decreases in entropy, thereby violating the Second Law of Thermodynamics. See http://hockeyschtick.blogspot.com/2014/05/new-paper-finds-climate-models-violate.html
A new paper has shown that the "settled science" computer models have variations in the top of the atmosphere incident solar radiation of up to 30 W/m2, which vary with the longitude, not the latitude. This is nonsense and is entirely due to very poor thought in the programming. Compared to the claimed 1.68 W/m2 value for the effect of doubling CO2, this variation in the top of the atmosphere solar radiation is a very large error. See http://wattsupwiththat.com/2015/03/10/whoops-study-shows-huge-basic-errors-found-in-cmip5-climate-models/
The climate computer models do not attempt to model the actual interface of the surface with the atmosphere. They do not properly handle water evaporation or evapotranspiration. They also do not handle advection (wind) effects well, particularly as it interacts directly with the surface in molecular collisions and in inducing water evaporation.
The climate computer models are wrong for many reasons:
- The physics behind the infra-red gas warming effects is wrong, as I pointed out in Why Greenhouse Gas Theory is Wrong -- An Examination of the Theoretical Basis
- Still further errors in the physics are input into the computer models as missing information or wrong information.
- The computer models are poorly programmed so that they often do not preserve well-known scientific laws.
- The computer models do not have sufficient spatial resolution to calculate some of the effects they are tasked to calculate.