Extended Data Fig. 5: 101-year maxima and minima.
From: No evidence for globally coherent warm and cold periods over the preindustrial Common Era

a–e, As for Fig. 3, but for 101-year instead of 51-year periods. Colours in maps indicate the century that has the largest ensemble-based probability of containing the warmest (a–c) and coldest (d, e) 51-year period within each climatic epoch (see Methods). f, As for Fig. 4, but for 101-year periods. The figure shows the fraction of Earth’s surface (y axis) that simultaneously experienced the warmest (top panels) or coldest (bottom panels) multidecadal period (51 years) during each of five different epochs (see Methods). Each solid circle represents an ensemble member, plotted according to the year in which the largest area experienced peak warm/cold conditions. Horizontal grey shading represents the distributions from the same analysis based on multivariate noise fields from an AR1 analysis, with darker colours indicating higher probabilities. Boxplots on the right show area fractions integrated over time. The centre line is the median; the ends of the boxes represent the interquartile range; and whiskers represent the 90% range. Bold text represents epochs with reconstructed area fractions significantly higher than those from the noise fields (Mann–Whitney U-test; α = 0.05). Recall that we searched for CWP maxima within the full 2,000-year reconstruction period. Unlike the 51-year maxima displayed in Fig. 4, some of the 101-year maxima within this 2,000-year range fall within the pre-1350 period, thus overlapping with the search windows of the RWP and MCA periods. Therefore, circles representing the CWP have a black border to distinguish them from other epochs.