We present a continuous humidity history in arid Central Asia over the past millennium based on the ~1.8-year high-resolution multiproxy record from Lake Dalongchi. Our findings emphasize that the Gleissberg solar cycle and quasi-regular period of ENSO amplitude play critical roles in controlling the effective humidity at century and multidecadal timescales, respectively. Our analysis provides new insights for hydroclimate predictions and climate simulations in arid Central Asia in the future.

Our proxy-based observations of the glacial inception following MIS 11 advance our mechanistic understanding of (and elucidates antecedent conditions that can lead to) high-magnitude climate instability during low- and intermediate-ice boundary conditions. We find that irrespective of the magnitude of climate variability or boundary conditions, the reorganization between Polar Water and Atlantic Water at subpolar latitudes appears to influence deep-water flow in the Nordic Seas.

We present a melt event record from an ice core from central Greenland, which covers the past 10 000 years. Our record displays warm summer events, which can be used to enhance our understanding of the past climate. We compare our data to anomalies in tree ring width, which also represents summer temperatures, and find a good correlation. Furthermore, we investigate an outstandingly warm event in the year 986 AD or 991 AD, which has not been analyzed before.

Palaeoenvironmental records only provide a fragmentary picture of the lake and wetland extent in North Africa during the mid-Holocene. Therefore, we investigate the possible range of mid-Holocene precipitation changes caused by an estimated small and maximum lake extent and a maximum wetland extent. Results show a particularly strong monsoon precipitation response to lakes and wetlands over the Western Sahara and an increased monsoon precipitation when replacing lakes with vegetated wetlands.

Since the orbital parameters of the past are different from today, applying the modern calendar to the past climate can lead to an artificial bias in seasonal cycles. With the use of multiple model outputs, we found that such a bias is non-ignorable and should be corrected to ensure an accurate comparison between modeled results and observational records, as well as between simulated past and modern climates, especially for the Last Interglacial.

Descriptions of climate held in archives are a valuable source of past climate variability, but there is a large potential for error in assigning quantitative indices (e.g. −2, v. dry to +2, v. wet) to descriptive data. This is the first study to examine this uncertainty. We gave the same dataset to 71 postgraduate students and 6 professional scientists, findings that error can be minimized by taking an average of indices developed by eight postgraduates and only two professional climatologists.

The mid-17th century saw several volcanic eruptions, deteriorating climate, political instability, and famine in Europe, China, and Japan. We analyze impacts of the eruptions on climate but also study their socio-political context. We show that an unambiguous distinction of volcanic cooling or wetting from natural climate variability is not straightforward. It also shows that political instability, poor harvest, and famine cannot only be attributed to volcanic climatic impacts.