This study reconstructed vegetation and tree cover in the Northern Hemisphere from a harmonized dataset of pollen counts from sediment and peat cores for the past 14 000 years. A model was applied to correct for differences in pollen production between different plants, and modern remote-sensing forest cover was used to validate the reconstructed tree cover. Accurate data on past vegetation are invaluable for the investigation of vegetation–climate dynamics and the validation of vegetation models.

Recent studies showed that a large number of independent vertical profiles allow for inferring a common local climate signal from the stacked stable water isotope record. Through investigating instrumental limitation and the effect of percolation of such porous samples, this study assesses the continuous flow analysis (CFA) technique in order to analyze the significant number of snow surface profiles within a reasonable time and with high quality.

We use a series of spectral techniques to quantify the strength of high-frequency climate variability in northeastern Greenland to 50 000 ka before present. Importantly, we find that variability consistently decreases hundreds of years prior to Dansgaard–Oeschger warming events. Model simulations suggest a change in North Atlantic sea ice behavior contributed to this pattern, thus providing new information on the conditions which preceded abrupt climate change during the Last Glacial Period.

Earth's past temperature reconstructions are critical for understanding climate change. We test the ability of these reconstructions using climate simulations. Uncertainties, mainly from past temperature measurement methods and age determination, impact reconstructions over time. While more data enhance accuracy for long-term trends, high-quality data are more important for short-term precision. Our study lays the groundwork for better reconstructions and suggests avenues for improvement.

We present a method to optimize the number and placement of temperature sensors in the borehole for borehole thermometry. Based on heat transfer model simulations, a greedy algorithm chooses sensor locations that minimize sampling errors. Applications in Antarctic and Greenland boreholes show this method outperforms traditional linear and exponential spacing, reducing errors up to tenfold. This approach offers an efficient, cost-effective solution to improve subsurface temperature monitoring.