Even the most sophisticated global climate models are known to have significant biases in the way they simulate the climate system. Correcting model biases is therefore essential for creating realistic reconstructions of past climate that can be used, for example, to study long-term ecological dynamics. Here, we evaluated three widely used bias correction methods by means of a global dataset of empirical temperature and precipitation records from the last 125 000 years.

This paper introduces a method to estimate the uncertainty of climate reconstructions from single sediment proxy records. The method can compute uncertainties as a function of averaging timescale, thereby accounting for the fact that some components of the uncertainty are autocorrelated in time. This is achieved by treating the problem in the spectral domain. Fully analytic expressions are derived. A companion paper (Part 2) complements this with application-oriented examples of the method.

A record of past sea temperature in the Indian sector of the Southern Ocean, spanning the last 14 200 years, has been developed by analysis of fossil diatoms in marine sediment. During the late deglaciation the reconstructed temperature changes were highly similar to those over Antarctica, most likely due to a reorganisation of global ocean and atmospheric circulation. During the last 11 600 years temperatures gradually cooled and became increasingly variable.

Computer simulations of the geological past are an important tool to improve our understanding of climate change. We present results from two simulations using the latest version of the UK's climate model, the mid-Holocene (6000 years ago) and Last Interglacial (127 000 years ago). The simulations reproduce temperatures consistent with the pattern of incoming radiation. Model–data comparisons indicate that some regions (and some seasons) produce better matches to the data than others.

This paper uses two different ways to perform model–data comparisons for the borehole temperature in Antarctica. The results suggest most models generally reproduce the long-term cooling in West Antarctica from 1000 to 1600 CE and the recent 50 years of warming in West Antarctica and Antarctic Peninsula. However, The 19th-century cooling in the Antarctic Peninsula (−0.94 °C) is not reproduced by any of the models, which tend to show warming instead.

During the globally warm early Eocene 56 million years ago, lush forests extended up to the high Arctic. Fossil plants from the Canadian High Arctic and Pacific Northwest of North America are a window into this past greenhouse world. We used an improved method for plant fossil climate reconstruction that provides a consensus reconstruction from all available proxies. Results show that the early Eocene climate in northern North America was similar across a broad range of latitudes.

An ice core drilled on the Renland ice cap (RECAP) in east-central Greenland contains a continuous climate record dating through the last glacial period. Here we present the water isotope record for the Holocene, in which high-resolution climate information is retained for the last 8 kyr. We find that the RECAP water isotope record exhibits seasonal and decadal variability which may reflect sea surface conditions and regional climate variability.

We present a novel post Last Glacial Maximum sediment record from Moossee (Swiss Plateau, southern central Europe). For the first time, five major reorganizations of vegetation could be definitely linked to paramount postglacial temperature and/or moisture changes. Present-day beech-dominated forests have been resilient to long-term climate change and human land use. They may prevail in future if climate warming does not exceed the amplitude of Mid Holocene temperature and moisture variability.

Data assimilation is becoming more and more important for past climate reconstructions. The assimilation of monthly resolved precipitation information has not been explored much so far. In this study we analyze the impact of assimilating monthly precipitation amounts and the number of wet days within an existing paleoclimate data assimilation framework. We find increased skill in the reconstruction, suggesting that monthly precipitation can constitute valuable input for future reconstructions.

This work analyses the behavior of atmospheric dynamics and hydroclimate in climate simulations of the last millennium. In particular, how external forcing factors, like solar and volcanic activity and greenhouse gas emissions, impact variables like temperature, pressure, wind, precipitation, and soil moisture is assessed. The results of these analyses show that changes in the forcing could alter the zonal circulation and the intensity and distribution of monsoons and convergence zones.

The early Eocene (~ 55 Ma) was an extreme warm period accompanied by a high atmospheric CO₂ level. We explore the relationships between ocean dynamics and this warm climate with the aid of the IPSL climate model. Our results show that the Eocene was characterized by a strong overturning circulation associated with deepwater formation in the Southern Ocean, which is analogous to the present-day North Atlantic. Consequently, poleward ocean heat transport was strongly enhanced.

Two sediment cores retrieved from the SE Brazilian continental margin were studied using multiple organic (alkenones) and inorganic (oxygen isotopes in carbonate shells and water) proxies to reconstruct the sea surface temperature (SST) over the last 50 000 years. The findings indicate the formation of strong thermal gradients in the region during the last climate transition, a feature that may become more frequent in the future scenario of global water circulation changes.

Climatic reconstructions based on tree-ring isotopic series convey key information on past conditions prevailing in forested regions. However, in some cases, the relations between isotopes and climate appear unstable over time, generating isotopic divergences. Former reviews have thoroughly discussed the divergence concept for tree-ring width but not for isotopes. Here we present a synopsis of the isotopic divergence problem and suggest collaborative work for improving climatic reconstructions.

This paper presents 24 new sea-level index points in the Spermonde Archipelago, Indonesia, and the reconstruction of the local Holocene relative sea-level history in combination with glacial isostasic adjustment models. We further show the importance of surveying the height of living coral microatolls as modern analogs to the fossil ones. Other interesting aspects are the potential subsidence of one of the densely populated islands, and we present eight samples that are dated to the Common Era.

A new study on droughts in Germany since 1500 reveals the long-term trend of single extreme events, as well as drier periods. Extreme droughts appeared in 1540, 1590, 1615, 1706, 1834, 1893, 1921, 1949 and 2018. Like today, droughts had manifold impacts such as harvest failures, water deficits, low water levels and forest fires. This had different societal consequences ranging from famine, disease, rising prices, migration and riots leading to subsidies and discussions on climate change.

Despite intense focus on the ~ 190 ‰ drop in Δ¹⁴C across the deglacial mystery interval, the specific mechanisms responsible for the apparent Δ¹⁴C excess in the glacial atmosphere have received considerably less attention. Sensitivity experiments with the computationally efficient Bern3D Earth system model suggest that our inability to reproduce the elevated Δ¹⁴C levels during the last glacial may reflect an underestimation of ¹⁴C production and/or a biased-high reconstruction of Δ¹⁴C.

The Labrador Sea is important for the modern global thermohaline circulation system through the formation of Labrador Sea Water. However, the role of the southward flowing Labrador Current in Labrador Sea convection is still debated. In order to better assess its role in deep-water formation and climate variability, we present high-resolution mid- to late Holocene records of sea surface and bottom water temperatures, freshening, and sea ice cover on the Labrador Shelf during the last 6000 years.

Due to chronology uncertainties within paleoclimate archives, it is unclear how climate oscillations from different records relate to one another. By using radionuclides to synchronize Greenland ice cores and a German lake record over 11 000 years, we show that two oscillations observed in these records were not synchronous but terminated and began with the onset of a grand solar minimum. Both this and changes in ocean circulation could have played a role in the two climate oscillations.

The manuscript identifies the main volcanic phases in the Laguna del Maule volcanic field and their impact in the lake basin through the late glacial and Holocene. We show that the bio-productivity and geochemical variabilities in the lake are related with climatic dynamics type ENSO, SPA and SWW and that the main phases are synchronous with the major regional climate changes on millennial timescales.

In this study, we present RainNet, a deep convolutional neural network for radar-based precipitation nowcasting, which was trained to predict continuous precipitation intensities at a lead time of 5 min. RainNet significantly outperformed the benchmark models at all lead times up to 60 min. Yet an undesirable property of RainNet predictions is the level of spatial smoothing. Obviously, RainNet learned an optimal level of smoothing to produce a nowcast at 5 min lead time.

This study explores the influence of the input data choice on spatial climate reconstructions. We compare three tree-ring-based data sets which range from small sample size, small spatial coverage and strict screening for temperature sensitivity to the opposite. We achieve the best spatial reconstruction quality by combining all available input data but rejecting records with little and uncertain climatic information and considering moisture availability as an additional growth limitation.

Tree rings are the main data source for climate reconstructions over the last millennium. Statistical tree-growth models have limitations that process-based models could overcome. Here, we investigate the possibility of using a process-based ecophysiological model (MAIDEN) as a complex proxy system model for palaeoclimate applications. We show its ability to simulate tree-growth index time series that can fit robustly tree-ring width observations under certain conditions.