Fig. 2: Long-term trend of δ18Oice and temperature for different seasons during the Holocene.
From: Climate simulations and ice core data highlight the Holocene conundrum over tropical mountains
The leading reconstruction component (long-term trend) of the MSSA (see Methods) for the time series of annual (a1-a11) proxy δ18Oice (‰; black: left axis), model δ18Oice (‰; blue: left axis), model temperature (°C; dashed orange lines are for arithmetic annual mean \({T}_{s}^{{uw}}\); solid orange lines are for annual mean weighted by monthly precipitation \({T}_{s}^{{pw}}\): right axis). b1-b11 show the leading reconstruction component of the MSSA for model δ18Op during JJA (orange) and DJF (blue). Model annual model δ18Oice is included for reference (dashed black line, identical to the blue line in a1-a11). The eigenvalues, or variance contributions, of the MSSA leading mode are 50% for proxy δ18Oice, 66% for model annual δ18Oice, 71% for model JJA δ18Oice, and 71% model DJF δ18Oice. The geographic location of each ice cores is shown in Fig. 1f.
