Initial studies of global climate change have sought to identify significant changes in global mean surface temperature and to attribute such changes to human influences. However, cause-and-effect are difficult to identify unequivocally in such a simple globally-averaged measure. More recent studies have focussed on fingerprint methods, which make use of the spatial patterns of temperature change to try to attribute the observed changes to one or more climate forcing factors. However, fingerprint methods also make use of more complex multi-variate statistics and the results may be harder to interpret or to communicate than those using global mean temperature.
In this study, we seek to follow the detection and attribution methodologies used in fingerprint detection studies but apply them using a small number of indices of global climate change. These indices have been selected based on earlier studies of climate change detection and on some of the key features identified in the climate change fingerprints that have been used commonly. They include the global mean temperature, the global mean temperature contrast between land and ocean, and the mean magnitude of the seasonal temperature cycle and of the diurnal temperature cycle on land. -- First, the observed trends in these global indices are compared with estimates of natural variability from control coupled climate model simulations to detect significant changes. Next, they are compared with forced climate model simulations to try to attribute the observed changes to one or more causes. The combination of these simple global indices has more power to attribute climate change than does the use of global mean temperature alone. While this approach may not be as statistically powerful as the fingerprint method, it is easier to explain