The tables below provide a comparison of model predictions with actual observations and provide a yes-no-undetermined score of whether the models are successful or not. Later in listings there are pro and con discussions of various topics and these are not scored.

Type of prediction	1900-2000 surface temperature trend
Model prediction	1.1 to 3.3 C warming if all greenhouse gases are included (IPCC 2001)
Actual measurements	Surface temperature warming of 0.6 C
Comments	Predicted warming is 2 to 5 times greater than observed warming. Lindzen says it is 4 times too large. Alternative and additional sources of warming include the sun, UHI and land use changes, soot on snow, and other reasons. More on land use changes here. More on the warm bias in surface observations here.
Score	0-1-0 Scoring is won-lost-tie system. A win means models and observations reasonably agree. A loss means significant disagreement. A tie means the models or observations give contradictory results.

Type of prediction	1979-2005 mid-tropospheric warming
Model prediction	About 0.15 to 0.58 C warming per decade (IPCC 2001)
Actual measurements	Between 0.1 and 0.14 C/decade. “In all cases these trends are positive. The increase in the UAH time series is 0.12 C/decade (0.22 F/decade), 0.14 C/decade (0.24 F/decade) for the RSS analysis and 0.10 C/decade (0.17 F/decade) for the University of Washington. Trends in UAH, RSS and UW data are less than the trend in global surface temperatures, which increased at a rate near 0.18 C/decade (0.32 F/decade) during the same 27 year period.”
Comments	The predicted warming is less than the model warming.
Score	0-2-0

Type of prediction	Surface and mid-tropospheric warming, 1979-2005
Model prediction	Mid-tropospheric warming should be 50-100% larger than surface warming.
Actual measurements	Surface warming is 0.18 C/decade compared to mid-tropospheric warming of 0.12 C/decade, opposite of what theory predicts.
Comments	A discussion can be found here. And here. And here.
Score	0-3-0

Type of prediction	Arctic warming
Model prediction	1.0 to 3.0 C/decade warming (IPCC 1995)
Actual measurements	Temperatures now are nearly the same as they were in 1940, consistent with large oscillations rather than a trend.
Comments	The arctic is probably warming due to ocean currents rather than greenhouse gases. A trend outside normal variations has not yet happened.
Score	0-4-0

Type of prediction	Animals and plants are migrating towards the poles (Parmesan and Yohe, 2003)
Model prediction	Study claims it provides evidence that climate models are correct.
Actual measurements	Actually the migration rates are consistent with a warming of 0.025 C/decade which is much smaller than models predict.
Comments	These results actually undermine the model predictions and may be an indication that the surface temperature record is overestimating the warming.
Score	0-5-0

Type of prediction	Medieval Warm Period (ca. 1000-1200 AD)
Model prediction	The state of the art GFDL climate model claims the Medieval Warm Period is physically impossible (Stouffer et al., 1994)
Actual measurements	The MWP exists according to borehole temperature measurements at 6000 locations (Huang et al., 1997).
Comments	More discussion can be at http://www.climateaudit.org/index.php?cat=8
Score	0-6-0

Type of prediction	Diurnal temperature range
Model prediction	Originally no change was predicted in the models, but some later models may have a change in DTR.
Actual measurements	Decreasing.
Comments	More discussion here and here.
Score	0-6-1

Type of prediction	Annual cycle of temperature.
Model prediction	0.5 to 1.1 C decrease predicted to have occurred in 20^th century.
Actual measurements	0.1 C decrease observed (Mann and Park, 1996).
Comments	Model prediction is 5 to 10 times too large.
Score	0-7-1

Type of prediction	Phase of annual cycle
Model prediction	Predicted change of -1.7 days in 20^th century.
Actual measurements	+0.8 days increase (Mann and Park, 1997).
Comments	Model predictions have wrong sign.
Score	0-8-1

Type of prediction	Stratospheric cooling.
Model prediction	Several degrees per decade predicted (IPCC 1995)
Actual measurements	Some cooling to 1995 and no trend since then.
Comments	Model predictions are too large.
Score	0-9-1

Type of prediction	Temperature lapse rate in tropics
Model prediction	Decreasing.
Actual measurements	Appears to have increased.
Comments	There is probably a problem with the convection parameterizations in the model. More discussion here.
Score	0-10-1

Type of prediction	Temperature lapse rate in the arctic.
Model prediction	Increasing.
Actual measurements	No change observed.
Comments	There is probably a problem with the convection parameterizations in the model. More discussion here.
Score	0-11-1

Type of prediction	Hurricane frequency.
Model prediction	More (Houghton et al., 1988 and the popular press)
Actual measurements	No clear trend. Possible small decrease since 1940 (Landsea et al., 1996)
Comments	Climate modelers tend to say there is an increase. Hurricane experts say the numbers oscillate of many years and there is no evidence for a trend.
Score	0-11-2

Type of prediction	Hurricane intensity
Model prediction	Greater (Houghton et al., 1988 and the popular press).
Actual measurements	No trend in 20^th century.
Comments	Models not confirmed so far.
Score	0-12-2

Type of prediction	Sea levels
Model prediction	Rising at 0.5 cm/yr (IPCC 1995)
Actual measurements	Rising at 0.18 Cm/yr (1993-present).
Comments	Models not confirmed. More discussion here, and here.
Score	0-13-2

Type of prediction	Extreme weather events.
Model prediction	More
Actual measurements	Only one scientific study found an increase and that was for extreme precipitation events in the US (Karl et al., 1993). All other studies show no trend or decreases.
Comments	Discussion here and here. If you sift through enough climate parameters at enough locations, one of them is bound to show a significant trend (data mining).
Score	0-14-2

Type of prediction	Northern Hemisphere snow cover.
Model prediction	Decreasing.
Actual measurements	No clear trend.
Comments	More discussion here and here