The recent availability of global networks of annual or seasonal resolution proxy data, combined with the few long instrumental and historical climate records available during the past few centuries, make it possible now to reconstruct annual and seasonal spatial patterns of temperature variation, as well as hemispheric, global-mean, and regional temperature trends, several centuries back in time.

Reconstructions of large-scale global or hemispheric trends during centuries past can place the instrumental assessments of climate during the 20th century in a longer term perspective, and provide more robust evidence regarding the roles of potential climate forcings over time. The reconstructed spatial patterns lead to important inferences regarding ENSO-scale variability, the spatial influences of climatic forcings, and the regional patterns that underlie large-scale climate variations. Here we expand upon the proxy-based annual global temperature pattern reconstructions described recently by Mann et al (1998). We present for the first time, seasonally-resolved versions of the proxy-reconstructed surface temperature patterns, and diagnose the seasonal differences between key climate indices and patterns of variations. We enable the reader to interactively examine spatial as well as temporal details (and their uncertainties) of yearly temperatures back in time for both annual-mean and seasonal windows. Annual and seasonal time histories of reconstructed Northern Hemisphere (NH), Southern Hemisphere (SH), and global (GLB) mean temperature are made available, as are time histories of the NINO3 index describing El Nino-related variations, time histories for particular regions of interest such as North America and Europe, and time series for temperature variations in different (e.g. tropical and extratropical) latitude bands. Time histories for specific gridpoints are available along with their estimated uncertainties. Time histories for the different eigenvectors (ie, the reconstructed Principal Components or "PC"s) are also available, along with the raw instrumental series, which underlie the temperature pattern reconstructions. For both the annual mean and seasonally-resolved temperature reconstructions, the reader can directly compare reconstructed patterns for different years, as well as the raw and reconstructed patterns during calibration and verification intervals, and view animated year-by-year sequences of reconstructed global temperature patterns. We also analyze in more detail the statistical relationships between climate forcings and temperature variations, taking into account potential lagged responses to climate forcings in empirical attribution analyses.

On to... The Introduction