A number of important conclusions are evident from the global-scale temperature reconstructions presented here. The sequences of annual and seasonal spatial temperature patterns presented in this study provide considerably more insight into the large-scale trends discussed in earlier work. The combination of multiple factors including El Nino/La Nina influences, interannual, decadal, and multidecadal patterns of extratropical variability, and putative responses to external forcings (such as volcanic aerosol loading of the upper atmosphere), leads to rich year-to-year spatial and temporal behavior which is readily documented in the annual temperature patterns. Considerable physical and dynamical insight is thus obtained into empirical climate variability over several centuries from the details of the patterns of annual and seasonal surface temperature variation. Some important new insights into the largest-scale climate trends are also available. As documented previously, not only is the global-scale warmth of the most recent decade observed to be quite unusual in the context of at least the past six centuries (and evidently, at least the past millennium), but 1998--the warmest year in the instrumental record--is seen to be truly exceptional in a long-term context. There is, however, a distinct latitudinal, seasonal, and spatial dependence evident in surface temperature trends during the past few centuries. Certain recent El Nino events (i.e., the 1997-98 and 1982-83) also appear somewhat anomalous in the context of the past few centuries, though the recent trends in ENSO indices are not nearly as dramatic as those in as the recent hemispheric warmth. Indeed, revised statistical attribution analyses comparing the hemispheric temperature series to candidate external forcings shows greater evidence for a likely anthropogenic influence than that presented previously, when the potential for a lagged response of the climate system to radiative forcing (owing to oceanic thermal inertia), is taken into account.
It is clear that the primary limitations of large-scale proxy-based reconstruction in past centuries, both temporally and spatially, reside in the increasingly sparse nature of available proxy networks available to provide reliable climate information back in time. Only through the arduous efforts of large numbers of paleoclimate researchers, can such networks be extended in space and time to the point where significant improvements will be possible in proxy-based reconstruction of the global climate. Such improvements will lead to further advances in our empirical understanding of climate variations during the past millennium, and will allow for more meaningful comparisons with the results obtained from model simulations of past climate variation and empirical climate variability.