The latest global temperature measurements are available for both satellite data  and for the Hadley CRU temperature data , so I thought it would be interesting to compare these with the predictions made in 1990 by the first IPCC report. There is now sufficient data to test whether the GCM modeling of greenhouse gases used by the IPCC really matches up to reality. The result is shown below.
This comparison is based on the analysis described below.
Predictions from the IPCC Report 1990 
“Based on the IPCC Business as Usual scenarios, the energy-balance upwelling diffusion model with best judgement parameters yields estimates of global warming from pre-industrial times (taken to be 1765) to the year 2030 between 1.3°C and 2.8″C, with a best estimate of 2 0°C This corresponds to a predicted rise from 1990 of 0.7-1.5°C with a best estimate of 1.1C. “
Prediction: 1990 to 2030 –> 0.7 – 1.5 degrees C
T = T(1990) + 0.0275*deltaY
Assuming a linear extrapolating to May 2011:
T(2011) = T(1990) + 0.58 (maximum of 0.79 and minimum of 0.37)
The data I have used are both the Hadley/CRU data  which is an IPCC reference set based on global surface temperature measurements and UAH data  using a NOAA satellite-based microwave measurement of the lower atmosphere temperature. The data for both sets is available as global averages. For the UAH data I calculated yearly averages for each year to allow a direct comparison with HadCru. The 2011 values are the averages as of May 2011. Both datasets actually publish “Temperature Anomalies” rather than the absolute temperature. These are merely the offset from a long-term average temperature. They use different intervals for the anomaly so this causes an offset. The actual data compared to the IPCC predictions are shown below in Figure 2. The IPCC curves are based on a linear increase using the 1990 temperature value of HadCrut. The curves through both datasets are least square smoothing fits.
Both datasets agree rather well in shape, ignoring the normalisation offset and the long term trends are remarkably similar. The two datasets are independent of each other; HadCrut is based on worldwide meteorological data, and the satellite IR data is calibrated without using any surface temperature data. This gives us confidence that they represent an accurate record of global temperatures over the last 21 years. The comparison with the IPCC predictions is made by normalising the UAH trend data to that of HadCru and then normalising the IPCC predictions to the 1990 HadCru trend value. The result is shown above in Figure 1.
Following a gradual rise of about 0.2 degrees from 1990 to 2000, global temperatures have stopped increasing and have actually fallen slightly. The only IPCC prediction which remains consistent with the current data is the lower prediction of a 0.7 degree rise from 1990 to 2030. The “Best” IPCC estimate and the higher 1.5 degree rise are ruled out by the data.
CO2 levels in the atmosphere have continued to rise over the last 10 years (see overlay to temperature comparison below in Figure 3) but temperatures have not risen since 2000. This implies that CO2 is not the main driver of global temperatures on these time periods and that other natural mechanisms are at least as important. No evidence of any positive temperature feedback with increasing CO2 levels is found.