General Circulation Models (GCM) used to forecast the future evolution of the atmosphere do not properly cover many of the important features of the last fifty years. This raises serious questions about their ability to predict future climate development with a precision that will be of use to policy makers.
The following are a simple and a sophisticated test of modelling the atmosphere.
First an analysis of regions with enhanced sensitivity to changes in CO2 concentration. These are found by comparing regions of varying concentrations of water vapour but constant CO2 concentration where changes in green house gases vary the amount of radiation directed downwards from the atmosphere to the surface. This should show changes in surface temperature as the global CO2 concentration increases with time.
This is followed by a test of the latest GCM models against measurements.
The amount of CO2 in the atmosphere has risen substantially in the last fifty years but it has been difficult to isolate its contribution to the world temperature rise that has occurred in that time.
A continuous stream of high quality measurements show the concentration of CO2 in the atmosphere has risen from 316 ppmv in 1959 to 382 ppmv in 2006. This is an increase of 21% in some fifty years. At the same time the global temperature has increased by 0.5 to 1.0 0C.
There are large variations in water vapour in the atmosphere with a maximum at the equator and minima at the poles. In addition there are bands of latitude where the seasonal temperature variations are small as the ocean interacts with the atmosphere. These should be regions where the effects of global changes in CO2 concentration are more obvious in year on year variations as other climate variations are reduced.
The damping of seasonal temperature change can be seen from Figure 1 showing the maximum variations of temperature as a function of latitude averaged over the years 1948 to 2006.
Figure 1
Two latitude bands were selected for analysis, the band 4 to 9 N in the tropics and the band 51 to 56 S in the Southern Ocean.
The tropical band surface is 75% ocean while the Southern Ocean band is 99% ocean. As a comparison, the band 51 to 56 N is only 40% ocean and has substantial seasonal variations.
The mean monthly temperatures are shown below averaged over the years 1948 to 2006 in Figures 2 and 3
Figure 2:
Figure 3
The variations are 0.8 0C in the Tropics and 4.0 0C in the Southern Ocean
The variations in the mean annual humidity are shown below in Figure 4:
There is a 70% reduction in water vapour in moving from the Tropics to the Southern Ocean with a consequent enhancement of the contribution of CO2 to the downward directed radiation from the atmosphere.
The seasonal variations show that humidity remains relatively stable in the two latitude bands chosen for analysis. This is not the case for the equivalent Northern latitude band where there is a factor of five seasonal change in humidity.
Figure 4
The mean annual temperatures for the Tropical band are shown below in Figure 5.
Figure 5
A simple straight line has been fitted to the temperatures although there is clearly some detailed short term structure present such as ENSO. The Southern Ocean temperatures have also been treated in the same way.
The linear gradients from the least squares fits are given in the Table 1
Table 1
The Southern Ocean temperatures are better described by a temperature increase for 1948 to 1976 and then a constant temperature. However as with the Tropical temperatures, there is clearly some short term structure seen in Figure 6.
Figure 6
The analysis shows the tropical temperature increase is substantially larger than the Southern Ocean increase.
The MODTRAN computer programme has been used to give a simple indication of the changes in downward radiation from the atmosphere to the surface. Relative humidity is held constant and temperatures and the water vapour scale adjusted to the measured values.
The temperature increases have been calculated using MODTRAN and assuming a latent energy contribution at the surface. The latent energy term is a function of surface temperature and reduces the temperature rise by a factor of two.
The results are shown in Table.2
Table 2
The calculations show that for increased CO2 there is a larger increase in downward radiation in the Antarctic region compared to the Tropics. This is also the case for a doubling of CO2 concentration in the atmosphere. Feedback effects have been included in the final results shown below in Table 3.
Table 3
Thus a simple calculation gives a larger temperature increase in the Antarctic region over the Tropics. However the atmosphere has energy transfer processes that may explain the apparent contradiction.
General Circulation Models (GCM) take into account many energy transfer processes and are used to forecast climate temperature changes. Many of these models are calibrated against past measurements of a number of atmospheric variables. Two models that offer access to their results have been examined with data taken from the GISS and GFDL websites. Both are members of the IPCC group listed at the LLNL website.
GCM surface temperature profiles for the fifty years from 1950 to 2000 were downloaded and the map longitude-latitude grid point temperatures averaged around latitude circles.
Surface temperature measurements were taken from the NCAR website and for comparison 5 year means have been used with temperatures averaged in latitude bands.
The results of the comparisons are shown below in Figure 7 for the Goddard Institute for Space Studies GCM with a coupled atmosphere-ocean. Data was downloaded for all forcings combined for the 1880-2003 Climate Forcings.
Figure 7
A similar comparison has been made for the Geophysical Fluid Dynamics Laboratory (Princeton) CM2.X Coupled Climate Model. The results are also shown below in Figure 8 and an overall summary is given in Table 4.
Figure 8
Table 4 – Global Temperature Changes
While the global values are consistent with the measurements, in detail the calculations are not supported by measurements. There are different responses at the North and South Poles and a complicated response in the latitudes 60 S to 60 N.
The GCM’s are closer to the measurements than the simple MODTRAN calculation. This demonstrates the importance of many processes other than the CO2 forcing. However the comparisons show that these processes do not seem to have been adequately modelled to date.
The contribution of increasing CO2 concentrations is not detectable with this analysis. This is not to doubt that it has an effect but that there are other processes also at work in the atmosphere ocean system that tend to dominate.
However the confidence with which the future predictions are presented coupled with the obvious mismatches with the past are an echo of the Soviet era Polish saying: “The future is certain only the past is unpredictable”.
Tom Quirk
Melbourne
Jennifer Marohasy BSc PhD has worked in industry and government. She is currently researching a novel technique for long-range weather forecasting funded by the B. Macfie Family Foundation.