Changes in daily variability of temperature and rainfall are most obviously manifest in changes in extreme events and much of the work in this area will be discussed in the extreme events section (Section 9.3.6). However, changes in short time-scale variability do not necessarily only imply changes in extreme weather. More subtle changes in daily variability, when integrated over time, could still have important socio-economic impacts. Hennessey et al. (1997) found that the simulated number of wet days (days where the rainfall is non-zero) in two mixed-layer models went down in mid-latitudes and up in high latitudes when CO2 was doubled, whilst the mean precipitation increased in both areas. The global mean precipitation also increased, by around 10% in both models, typical of the changes in many mixed-layer models on doubling CO2. An analysis of changes in daily precipitation variability in a coupled model (Durman et al., 2001) suggests a similar reduction in wet days over Europe where the increase in precipitation efficiency exceeds the increase in mean precipitation.
Kattenberg et al. (1996) reported research on changes in inter-monthly temperatures and precipitation variability from two coupled models (Meehl et al, 1994; Parey, 1994). More recently, there have been several studies looking at changes in intra-seasonal circulation patterns using higher resolution atmosphere-only models with projected SSTs taken from coupled models at given time periods in the future (e.g., Beersma et al., 1997; Schubert et al., 1998). The effects of changes in extra-tropical storms on extreme wind and precipitation events are described in Section 9.3.6, but there has also been work on changes in lower-frequency variability such as persistent or "blocking" anti-cyclones. As discussed in the SAR, there still seems to be little consensus on the methodology for looking at changes in storms and blocks and it is likely that this is partly the reason for the lack of consistency in results. In new studies, Lupo et al. (1997) looked at the effect of doubled CO2 on several of the characteristics of blocking. They found an increase in the number of continental blocks and a general increase in the persistence of blocks, but with weakened amplitude. In contrast, Carnell and Senior (1998) found the largest change was a decrease in blocking in the North Pacific Ocean in winter in their model. Earlier studies have pointed to the possible model dependency of results (Bates and Meehl, 1986) and Carnell and Senior (2000) suggest that the changes in blocking found in their earlier study (Carnell and Senior, 1998) may depend on the meridional gradient of temperature change in the model, which may in turn depend on the simulation of cloud feedback in their model. Zhang and Wang (1997) found a decrease in the total number of Northern Hemisphere winter anticyclones under increased greenhouse gases, although they did not specifically look at blocking anticyclones.
Fyfe (1999) has looked at changes in African easterly waves due to a doubling of CO2 in one model. Significant low-level warming and increases in atmospheric humidity over the Northern Sahara lead to an increase in the easterly wave activity. Again, these results must be considered speculative given the relatively low resolution of the model (T32, about 3.5° res-olution), which leads to substantial systematic biases in the present day simulation of the low-level storm track in the region.
Other reports in this collection