Skillful forecasts of extreme weather events have a major socio-economic relevance. Here, we compare two approaches to diagnose the predictability of Eastern Mediterranean heat waves: one based on recent developments in dynamical systems theory and one leveraging numerical ensemble weather forecasts. We conclude that the former can be a useful and cost-efficient complement to conventional numerical forecasts for understanding the dynamics of Eastern Mediterranean heat waves.

This work covers multiple 3D simulations of the hydrodynamics of Lake Tanganyika, covering the inter-seasonal variations and the evolution linked to climate change. The research was done with COSMO-CLM² data, which was used to run the SLIM 3D Lake Tanganyika model. The main results explain how this stratified lake can still maintain a certain mixing between the different layers, but how this would come to an end due to climate change.

A collapse of the Atlantic Meridional Overturning Circulation can be described by six parameters and Langevin dynamics. These parameters can be found by a machine learning technique when applied to sample set of collapses seen in climate models of intermediate complexity. With this parametrisation, it might be possible to estimate how much freshwater is needed to observe a collapse in more complicated models and reality.

Variability of sea surface temperatures (SST) from 1200 to 2000 CE is quite well-known, but the history of deep ocean temperatures is not. Forcing an ocean model with these SSTs, we simulate temperatures in the ocean interior. The circulation changes alter amplitude and timing of deep ocean temperature fluctuations below 2 km depth, e.g., delay the atmospheric signal by ~ 200 yr in the deep Atlantic. Therefore, ocean circulation changes are shown to be as important as SST changes at these depths.

Many projections of future land use/cover exist. We evaluate a number of these and determine the variability they cause in ecosystems and their services. We found that projections differ a lot in regional patterns with some of them being at least questionable in a historical context. Across ecosystem service indicators resulting variability until 2040 was highest in crop production. Results emphasize that such variabilities should be acknowledged in assessments of future ecosystem provisions.

In the Paris Climate Agreement, countries agreed to keep global warming since pre-industrial times below 2 °C and ideally below 1.5 °C, requiring major reductions in greenhouse gas emissions. We investigate the regional effectiveness of lowering methane emissions from human activities, (re)afforestation, and growing bioenergy crops. Methane reduction always helps, especially for major methane-emitting regions. However, the preferred land-management strategy and its effectiveness is more variable.

The atmospheric and surface conditions that exist during the occurrence of daytime low-level clouds (LLCs) and their influence on solar radiation were investigated in West Africa. These LLCs are linked to horizontal transport of cold moist air from the ocean, surface heating and evaporation during the rainy season. In the dry season they are mostly linked to converging air at the surface and subsequent upward movement. Their occurrence lead to a strong reduction of the incoming solar radiation.

In 2016, northern France experienced an unprecedented wheat crop loss. This crop loss was likely due to an extremely warm December 2015 and abnormally high precipitation during the following spring season. Using stochastic weather generators we investigate how severe the metrological conditions leading to the crop loss could be in current climate conditions. We find that December temperatures were close to the plausible maximum but that considerably wetter springs would be possible.

Compound extremes such as heavy precipitation and extremes winds can lead to large damages. To date it is unclear how well climate models represent such compound extremes. Here we present a new measure to assess differences in the dependence structure of bivariate extremes. This measure is applied to assess differences in the dependence of compound precipitation and wind extremes between three model simulations and one reanalysis dataset in a domain in Central Europe.

To illustrate the fact that future economic damages from climate change are often highly underestimated, we applied two different statistically-based damage functions available in the literature to a global cooling of 4 °C. We show that the GDP projections obtained are qualitatively at odds with the state of our planet during the last ice age. Therefore, we conclude that such functions do not provide relevant information on potential damages of a large climate change, be it a cooling or a warming.

Two long-lasting high pressure systems in summer 2018 led to heat waves over Scandinavia and an extended summer period with devastating impacts on, e.g., agriculture and human life. Using 5 climate model ensembles, the unique 263-year Stockholm temperature time series along with a composite 150-year time series for whole Sweden, we found that anthropogenic climate change has strongly increased the probability of a warm summer such as the one observed 2018 to occur in Sweden.

We investigate North Atlantic weather seasonality over 1979–2100 by classifying year-round fields of 500 hPa geopotential height from one reanalysis dataset and 12 climate models. Generally, models have seasonal structures similar to the reanalyses. Historical winter (summer) conditions decrease (increase), due to uniform Z500 increase (i.e. uniform warming). However, relative to the increasing Z500 seasonal cycle, future seasonality (spatial patterns, seasonal cycle) appears almost stationary.