A spatio-temporal view of variability in pollen records during the last Glacial

Climate variability influences the probability of extreme events and is therefore of great importance for risk management. Nevertheless, changes in climate variability over time are far less studied than changes in the mean state of the climate system. Proxy records can be used to estimate the dependency of climate variability on the state and timescale, but their climate signal is perturbed by non-climatic processes and dating uncertainties. Analyzing ice cores and marine sediments, it was shown that temperature variability during the Last Glacial Maximum was larger than in the Holocene and that the magnitude of variability change depends on latitude.

We estimate millennial and orbital scale variability in pollen records during the last Glacial. We draw on a global network of published pollen records, which are influenced by local temperature and moisture availability, and compare these estimates with temperature, precipitation, and vegetation variability in climate simulations of the last Glacial cycle. We discuss the regional consistency of timescale dependent estimates. Differences between Marine Isotope Stages 2, 3, and 4 are examined by comparing spatial patterns during those three periods. Then, we use spectral methods to study the scaling behavior of the pollen records. This provides additional information on the continuum of variability from centennial to orbital scales. Finally, we quantify the co-occurrence of millennial and orbital scale fluctuations across different pollen records with paleoclimate network techniques.

Our work extends previous estimates to the terrestrial realm and to longer timescales. The results provide new insight on the climate variability differences between glacial and interglacial states, and on the mismatch between climate simulations and proxy data.