In order to gauge just how unusual climate trends during the 20th century truly are, and what the likely causative agents influencing these trends may be, we must rely on indirect lines of evidence to provide a broader context of past climate changes. For assessing some of the broad long-term trends, the history of mountain glacier fluctuations (Grove and Switsur, 1994) and geothermal, borehole-based estimates of past ground temperature (Pollack et al, 1998), can provide important information on climate changes during past centuries. For assessing a year-by-year or even decade-by-decade chronology of such climate changes, however, we must rely upon high-resolution "proxy" climate indicators--natural archives that record seasonal or annual climate conditions such as ice cores, tree-ring measurements, laminated sediments, and corals--combined with the scant available historical documentary or instrumental evidence available in prior centuries.

Increasingly, studies based on the assimilation and analysis of such global "multiproxy" networks of high-resolution proxy climate data have proven useful for assessing global or hemispheric patterns of climate in past centuries (e.g., Bradley and Jones, 1993; Hughes and Diaz, 1994; Lean et al, 1995; Mann et al, 1995; Jones et al, 1998, Crowley and Lowery, 2000), and reconstructing climate trends in particularly sensitive, high-latitude regions (Overpeck et al, 1997;, Briffa et al, 1998). Such studies have been spurred by hopes to better constrain the influences of natural and anthropogenic factors on long-term climate variability and change (Lean et al, 1995; Overpeck et al, 1997; Mann et al, 1998), to estimate climate sensitivity to external radiative forcing (Crowley and Kim, 1999; Waple et al, 2000) and to validate the behavior of climate models on multidecadal and longer timescales (Barnett et al, 1996; Jones et al, 1998, Delworth and Mann, 2000).

Most recently, global multiproxy climate data have been used to calibrate global-scale patterns of temperature on a yearly basis, several centuries in time by Mann et al (1998) (henceforth "MBH98"). The reliability of these reconstructions was demonstrated by cross-validation with independent data, and uncertainties back in time were assessed. These reconstructions have since been extended to estimate Northern Hemisphere temperature variations over the past millennium (Mann et al, 1999), to examine ENSO-scale patterns of climate variability during past centuries (Mann et al, 2000), to compare observed patterns of variability in the Atlantic with natural coupled ocean-atmosphere modes evident in long climate model integrations (Delworth and Mann, 2000) and to assess the relationship between global patterns of climate variation and particular regional patterns such as the North Atlantic Oscillation or "NAO" (Mann, 2000; Cullen et al, 2000) Here we further expand on the results of the MBH98 study. We make available for the first time, seasonally-resolved versions global surface temperature patterns based on the data and methods described in MBH98. These reconstructions, as well as the original annual-mean reconstructions of MBH98, are made available in a "user-friendly" interactive forum, allowing readers to select the particular spatial regions and time periods of particular interest. We also examine in greater detail than before issues related to the sensitivity of the climate reconstructions to varying networks of proxy data, the regional and latitudinal details of past climate variability, and the detection of natural and anthropogenic influences on past temperature changes. We encourage readers to investigate the details of the temperature reconstructions themselves through an interactive graphical user interface ("GUI"). It is our intent here to provide information about past climate changes for consumption by a broad, multidisciplinary readership. Details regarding regional variations, verification or "cross-validation", and uncertainties inherent in the reconstructions are stressed, so as to encourage a prudent interpretation of proxy-based estimates of past climate changes.

On to... Review of Data and Methods