Research

The above video shows a simulation of an extratropical cyclone with the ICON model at 2 km horizontal grid spacing and explicit deep convection, illustrating our use of atmosphere and climate models. Our research activities fall into five broad categories:

Circulation of the atmosphere

The dynamics of the atmosphere shape regional climate and impact local weather. We are interested in the atmospheric circulation response to different drivers, including climate change,  internal variability, and microphysical processes. We mainly focus on the northern hemisphere mid-latitudes and the tropics using a hierarchy of models ranging from idealized to fully coupled. In addition, we investigate how small-scale uncertainty (such as cloud processes) can feed up to the synoptic scale to influence the radiation budget and predictability. 

In the past we have shown how clouds shape climate (e.g., the tropical rain belts and the extratropical storm tracks) via their coupling to the atmospheric circulation. We also research how small-scale microphysical uncertainties, like the interaction between cloud droplets and radiation, can control the radiative budget. Further we are interested in the interplay between the climate and weather times scales.

Key publications

  • Keshtgar et al. (2023): Cloud-radiative impact on the dynamics and predictability of an idealized extratropical cyclone, Weather and Climate Dynamics https://doi.org/10.5194/wcd-4-115-2023
  • Voigt et al. (2021): Clouds, radiation, and atmospheric circulation in the present-day climate and under climate change. WIREs Clim Change https://doi.org/10.1002/wcc.694
A view of a cluster of tropical thunderstorms with extensive anvil cloud shields (Credit: NASA/ISS)

Clouds and radiation

Due to their interaction with radiation and their heating of the atmosphere, clouds are essential in shaping climate at both the global and regional scales. Thus, we believe that climate cannot be understood without also understanding clouds.

Our main interest lies on ice clouds since they play a large role in the radiative balance and through their heating in the upper atmosphere, they can influence the strength and shape the large-scale flow/circulation patterns. The way ice crystals form, grow, decay and interact with radiative fluxes is not well understood, which leads to uncertainties in representation of present day climate and in climate change projections. We try to address those uncertainties with the help of a hierarchy of climate models of increased complexity as better described under the climate computation tab.

Key publications

Snowball Earth and Exoplanets

While surface ice is limited to high-latitude and high-altitude regions in today’s climate, there is ample evidence that this was not always the case. During the Neoproterozoic era 1-0.5 billion years ago, Earth experienced pan-glaciations in which all or almost all of Earth’s surface was covered in ice. We study the dynamics of these Snowball Earth states by means of hierarchies and ensembles of global climate models. A question of particular relevance to us is the existence of near-Snowball-Earth states that escape full glaciation and maintain a narrow strip of tropical ocean. Such states may explain the survival of life during the Neoproterozoic pan-glaciations and provide further insight into the habitability of exoplanets.

In fact, many aspects of the climate of a Snowball Earth might be more similar to that of extrasolar planets than modern Earth. The same fundamental physics that governs the climate of Earth allows one to study the climate of extrasolar planets. More than 5,000 exoplanets have been discovered to date, and the James Webb Space Telescope is expected to discover Earth-like exoplanets whose atmosphere contains water as a condensable component. We are applying our tools and understanding developed from studying the climate of modern Earth to such Earth-like exoplanets.

Key publications

Surface albedo of sea ice (grayscale) and sea-surface temperatures (thermal) of a near-Snowball-Earth state in ICON-ESM. © Johannes Hörner

Climate Computation

The latest climate models provide an increasingly realistic representation of the Earth system based on increasing resolution, complexity and processes covered. The application and evaluation of these models raises many questions as well as requiring novel data processing techniques. This is why climate modeling and data analysis are core activities in our research team. We develop and run hierarchies and ensembles of climate models based on the ICON, CESM and MPI-ESM modeling frameworks. In addition to running our own simulations we also make use of community datasets such as CMIP, SMILEs or Dyamond.

To analyze these datasets we use various methods including statistics, budget calculations, and machine learning. Since the datasets we work with can exceed the terabyte scale we are also concerned with the development of tools and workflows to handle such data volumes.

For our climate computation activities we work on several high-performance computing systems, including the Vienna Scientific cluster and IMG’s departmental linux clusters.

Key publications

Evolution of an idealized baroclinic life cycle shown by 3-dimensional cloud cover and surface temperature simulated by ICON. © Behrooz Keshtgar

High-resolution modeling

Global climate models, with their coarse horizontal grid spacings of around 100 km, rely on empirical relationships to simulate convective systems and clouds, which limits our understanding of these key processes in the climate system. To address this limitation, high-resolution models with grid spacings of 5 km or finer have been developed. These models can more accurately resolve convective, cloud, and precipitation processes, offering a more physical and reliable representation of both global and regional climate. By using these advanced models, we aim to deepen the fundamental understanding of the role of clouds in the climate system and provide critical insights to inform decision-making in sectors such as infrastructure and energy systems.

Photovoltaics play a key role in decarbonizing the energy sector and therefore need to be resilient to climate change and extreme weather conditions. The main objective of “KlipPer” (Klimawandelprojektionen und Photovoltaik-Ertrag: climate change projections and photovoltaic yield) is to make the knowledge and data from climate modeling usable for the design and operation of a resilient photovoltaic infrastructure in Austria in a changing climate. We investigate future changes in photovoltaic yield over Austria and assess the added value of high temporal and spatial resolution climate data for PV yield modeling.

Moreover, within “HighResLearn“, we use machine learning to identify the added value of high-resolution simulations. In particular, we estimate climatological model biases and climate change signals for the greater Alpine domain. Beside using ML methods for evaluating high-resolution climate model output, we also develop data recipes that enable the climate community in Austria and beyond to easily access and process global high-resolution model output.

Key publications