Comparisons between long-term model simulations and empirical reconstructions can provide some potential insight into the sensitivity of the climate system to radiative forcing (e.g., Crowley and Kim, 1999). Preliminary climate system modeling, with an emphasis on the carbon cycle of the past millennium (Gerber et al., 2003), indicates that the larger-amplitude century-scale variability evident in some temperature reconstructions is inconsistent with constraints provided by comparison of modeled and observed pre-anthropogenic CO2 variations. This reinforces the evidence for relatively modest (less than 1°C) variations prior to the twentieth century and for a moderate equilibrium climate sensitivity of roughly 2-3°C for a doubling of CO2 concentrations (e.g., Cubasch et al., 2001).
Volcanic and solar radiative forcing of changes in the El Niño/Southern Oscillation (ENSO) in past centuries (Ruzmaikin, 1999; Adams et al., 2003; Mann et al., 2005) may explain preliminary empirical evidence for a prevalence of La Niña-like conditions during the eleventh to fourteenth centuries, and El Niño-like conditions during the seventeenth century (Cobb et al., 2003). Such findings, along with anthropogenic land-use change, further emphasize the potential spatial complexity of the climate in response to past changes in radiative forcing.
The last 25 years feature unprecedented data documenting simultaneous variations in radiative forcings, climate feedbacks, and climate itself. Many of the more recent datasets have been acquired using space-based instruments, which achieve essentially continuous, global coverage compared to ground-based observations. Space-based observations are available for solar irradiance; volcanic aerosols; concentrations of ozone, CO2, other greenhouse gases, and CFCs; cloud cover and cloud properties; water vapor; land features, including snow cover, ice, and albedo; temperature of the ocean, land surface, and atmosphere; and other quantities relevant to radiative forcing. Indices are routinely produced, including for ENSO and the North Atlantic Oscillation, based on the extensive datasets and new analysis procedures that extract variability modes. Many of the original databases have been reprocessed, recognizing the need for improved algorithms to remove instrumental drifts so as to better quantify actual change. Examples include the National Centers for Environmental Prediction reanalysis of atmospheric variables, the ISCCP cloud data, and the ground-based Dobson ozone network.
Furthermore, the epoch of the past 25 years is sufficiently long that a range of natural radiative forcing strengths and internal variability modes is sampled concurrently with known anthropogenic forcings. This period includes notable volcanic episodes (i.e., El Chichon, Mt. Pinatubo), two com-