reconstructions decrease dramatically from century-to-century moving backward in time (see, e.g., Figure O-2). At present, fewer than 30 annually resolved proxy time series extend further back than A.D. 1000; relatively few of these are from the Southern Hemisphere and even fewer are from the tropics. Although fewer sites are required for defining long-term (e.g., century-to-century) variations in hemispheric mean temperature than for short-term (e.g., year-to-year) variations (see Chapter 2), the coarse spatial sampling limits our confidence in hemispheric mean or global mean temperature estimates prior to A.D. 1600 and makes it very difficult to generate meaningful quantitative estimates of global temperature variations prior to about A.D. 900. Moreover, the instrumental record is shorter than some of the features of interest in the preindustrial period (i.e., the extended period of sporadic warmth from A.D. 800 to 1300 and the subsequent Little Ice Age), so there are very few statistically independent pieces of information in the instrumental record for calibrating and validating long-term temperature reconstructions.
The first systematic, statistically based synthesis of multiple climate proxies was carried out in 1998 by Michael Mann, Raymond Bradley, and Malcolm Hughes (Mann et al. 1998); their study focused on temperature for the last 600 years in the Northern Hemisphere. The analysis was later extended to cover the last 1,000 years (Mann et al. 1999), and the results were incorporated into the 2001 report of the Intergovernmental Panel on Climate Change (IPCC 2001). Later, Mann and Jones (2003b) extended the multiproxy reconstruction further back to cover the last 1,800 years (see Figure 10-4). On the basis of these reconstructions, it was concluded that temperatures gradually dropped from a relative maximum at about A.D. 1000 to a minimum at about 1850 and then increased sharply through the 20th century. The graph illustrating the trend, often called the hockey stick curve (reproduced in Figure O-4), received wide attention because it was interpreted by some people as definitive evidence of human-induced global warming. The ensuing debate in the scientific literature continues even as this report goes to press (von Storch et al. 2006, Wahl et al. 2006).
The Mann et al. large-scale surface temperature reconstructions were the first to include explicit statistical error bars, which provide an indication of the confidence that can be placed in the results. In the Mann et al. work, the error bars were relatively small back to about A.D. 1600, but much larger for A.D. 1000–1600. The lower precision during earlier times is caused primarily by the limited availability of annually resolved paleoclimate data: That is, the farther back in time, the harder it is to find evidence that provides reliable annual information. For the period before about A.D. 900, annual data series are very few in number, and the non-annually resolved data used in reconstructions introduce additional uncertainties.
Since the late 1990s, a number of alternative reconstructions have been generated using different statistical methods and proxy datasets (Esper et al. 2002a,b, 2003; Cook et al. 2004; Moberg et al. 2005b; Rutherford et al. 2005; Hegerl et al. 2006; D’Arrigo et al. 2006; Wahl and Amman in press). Figure 11-1 shows the results of several of these efforts, some of which are described in additional detail in the next section.