Such an approach in the detection and attribution of anthropogenic or natural influences on climate (e.g. Allen and Tett '99) involves quantifying the level of agreement between one or more model-predicted forced-response patterns and observed changes from radiosondes in the recent climate record.
In a latitude-height fingerprint of the atmospheric response of the HadCM2 GCM to decadal solar forcing, a strong dipole component in the Southern Hemisphere is revealed (15 to 30 degrees South). Although the pattern of the model-simulated response appears realistic, it's amplitude is less than the observed amplitude by a factor of 2 to 3, suggesting the absence of a key feedback mechanism in the GCM.
Further analysis is carried out on seasonal vertical temperature data since one possible mechanism (Haigh '96) suggests a seasonal cycle may be present whereby, at solar maximum, warming of the stratosphere occurs, strengthening tropospheric easterlies in summer.
This model response to the solar signal lends credibility to previous observationally based reports. Quantifying the solar signal is critical for correct interpretation of long term trends and decadal climate prediction.