Temporal analogues make use of climatic information from the past as an analogue for possible future climate (Webb and Wigley, 1985; Pittock, 1993). They are of two types: palaeo-climatic analogues and instrumentally based analogues.
Palaeoclimatic analogues are based on reconstructions of past climate from fossil evidence, such as plant or animal remains and sedimentary deposits. Two periods have received particular attention (Budyko, 1989; Shabalova and Können, 1995): the mid-Holocene (about 5 to 6 ky BP2) and the Last (Eemian) Interglacial (about 120 to 130 ky BP). During these periods, mean global temperatures were as warm as or warmer than today (see Chapter 2, Section 2.4.4), perhaps resembling temperatures anticipated during the 21st century. Palaeoclimatic analogues have been adopted extensively in the former Soviet Union (e.g., Frenzel et al., 1992; Velichko et al., 1995a,b; Anisimov and Nelson, 1996), as well as elsewhere (e.g., Kellogg and Schware, 1981; Pittock and Salinger, 1982). The major disadvantage of using palaeoclimatic analogues for climate scenarios is that the causes of past changes in climate (e.g., variations in the Earth's orbit about the Sun; continental configuration) are different from those posited for the enhanced greenhouse effect, and the resulting regional and seasonal patterns of climate change may be quite different (Crowley, 1990; Mitchell, 1990). There are also large uncertainties about the quality of many palaeoclimatic reconstructions (Covey, 1995). However, these scenarios remain useful for providing insights about the vulnerability of systems to abrupt climate change (e.g., Severinghaus et al., 1998) and to past El Niño-Southern Oscillation (ENSO) extremes (e.g., Fagan, 1999; Rodbell et al., 1999). They also can provide valuable information for testing the ability of climate models to reproduce past climate fluctuations (see Chapter 8).
Periods of observed global scale warmth during the historical period have also been used as analogues of a greenhouse gas induced warmer world (Wigley et al., 1980). Such scenarios are usually constructed by estimating the difference between the regional climate during the warm period and that of the long-term average or a similarly selected cold period (e.g., Lough et al., 1983). An alternative approach is to select the past period on the basis not only of the observed climatic conditions but also of the recorded impacts (e.g., Warrick, 1984; Williams et al., 1988; Rosenberg et al., 1993; Lapin et al., 1995). A further method employs observed atmospheric circulation patterns as analogues (e.g., Wilby et al., 1994). The advantage of the analogue approach is that the changes in climate were actually observed and so, by definition, are internally consistent and physically plausible. Moreover, the approach can yield useful insights into past sensitivity and adaptation to climatic variations (Magalhães and Glantz, 1992). The major objection to these analogues is that climate anomalies during the past century have been fairly minor compared to anticipated future changes, and in many cases the anomalies were probably associated with naturally occurring changes in atmospheric circulation rather than changes in greenhouse gas concentrations (e.g., Glantz, 1988; Pittock, 1989).
Other reports in this collection