Analysis of global climate model performance in reproducing observed regional climate variability has given widely varying results depending on model and region. Interannual variability in temperature was assessed regionally, as well as globally, in a long control simulation with the HadCM2 model (Tett et al., 1997). Many aspects of model variability compared well against observations, although there was a tendency for temperature variability to be too high over land. In the multi-regional study of Giorgi and Francisco (2000a), both regional temperature and precipitation interannual variability of HadCM2 were found to be generally overestimated. Similar results were obtained in the European study of Machenhauer et al. (1998) using the ECHAM/OPYC3 model. However, in a 200-year control simulation with the CGCM1 model (see Table 9.1), Flato et al. (2000) noted that simulated interannual variability in seasonal temperature and precipitation compared well with observations both globally and in five selected study regions (Sahel, North America, Australia, southern Europe and Southeast Asia).
Comparison against observations of daily precipitation variability as simulated at grid boxes in GCMs is problematic because the corresponding variability in the real world operates at much finer spatial scale (see Hennessy et al., 1997). A significant development in this area has been the work of Osborn and Hulme (1997) who devised a method of calculating grid box average observed daily precipitation that corrects for biases commonly introduced by insufficient station density. Using this correction, agreement between observations and the results of the CSIRO GCM were significantly improved. In an analysis of different AGCMs over Europe, Osborn and Hulme (1998) found that the models commonly simulated precipitation in winter to be more frequent and less intense than observed. Daily temperature variability over Europe was found to be too high in winter in the Hadley Centre model (Gregory and Mitchell, 1995) and in winter and spring in the ECHAM3/LSG model (Buishand and Beersma, 1996).
Synoptic circulation variability at daily and longer time-scales operates at a spatial scale which GCMs can simulate directly and work has focused on GCM performance in this area (e.g., Katzfey and McInnes, 1996; Huth, 1997; Schubert, 1998; Wilby et al., 1998a; Osborn et al., 1999; Fyfe, 1999). Regions studied include North America, Europe, southern Africa, Australia and East Asia. Although in many respects model performance is good, some studies have noted synoptic variability to be less than in the observations and the more extreme deviations from the mean flow to be less intense or less frequent than observed (e.g., Osborn et al., 1999).
Simulated climatic variability has also been examined as part of assessing model representation of the link between atmospheric circulation and local climate. Results have shown considerable regional differences. Osborn et al. (1999) examined the relationship between the circulation anomalies and grid-box average temperature and precipitation anomalies and found this to be well represented by the HadCM2 model. However, Wilby and Wigley (2000) found HadCM2 less satisfactory in reproducing the observed correlations between daily precipitation over six regions in the United States and a variety of different atmospheric predictor variables. In a similar investigation over Europe, Busuioc et al. (1999) found the performance of the ECHAM3 AGCM to be good in some seasons.
Widmann and Bretherton (2000) examined precipitation variability from atmospheric reanalyses as an alternative method of validating GCMs under historic flow conditions. By virtue of this approach, the atmospheric circulation is constrained to be unbiased, but the precipitation is calculated according to model physics and parametrizations. Results based on the GCM used in the reanalysis were found to be in good agreement with observations over Oregon and Washington.
Other reports in this collection