Christian Kühnlein
Research Subjects
Research Interests
Christian Kühnlein has a total of 183 co-authors. Here are the top co-authors by number of shared publications:
- Heini Wernli
Institute for Atmospheric and Climate Science, ETH Zurich, Zurich, Switzerland
13 shared publications - Nicolai Krieger
Institute for Atmospheric and Climate Science, ETH Zurich, Zurich, Switzerland
10 shared publications James Kent
7 shared publicationsMichael Sprenger
Institute for Atmospheric and Climate Science, ETH Zurich, Zurich, Switzerland
7 shared publicationsCelal Konor
Department of Atmospheric Science, Colorado State University, Fort Collins, CO, USA
6 shared publications- Christiane Jablonowski
Department of Climate and Space Sciences and Engineering, University of Michigan, Ann Arbor MI, United States
6 shared publications
Latest Publications
- Open Access
- 28 September 2026
- Published by Elsevier in SSRN Electronic Journal
AdHImEx: Adaptively High-Order Implicit-Explicit Transport for Large Time Steps
Adaptively Implicit-Explicit (AdImEx) time stepping provides stability for large time steps for mass-conservative transport, making it attractive for the numerical representation of advection in weather and climate prediction. Existing AdImEx transport schemes become first-order accurate in the large-time-step, implicit regime and may require multiple sparse matrix solutions per time step. This work introduces AdHImEx, a new AdImEx scheme that addresses both limitations simultaneously. AdHImEx provides second-order accuracy for large Courant numbers while requiring only a single matrix solution per time step. It is a Runge-Kutta scheme that blends third-order accurate explicit time stepping and Crank-Nicolson implicit time stepping to provide numerically verified stability for Courant numbers up to 100. AdHImEx is designed for flows that are predominantly explicit, activating implicit time stepping only where large Courant numbers occur. It retains the full efficiency and third-order accuracy of the explicit time stepping in regions with small Courant numbers. In this work, it is combined with a fifth-order accurate finite-volume discretisation in space for mass conservation. A second contribution of this work is a novel stage-dependent treatment of the divergence operator that preserves constancy despite different implicit and explicit Runge-Kutta stage time steps and spatially varying implicitness, removing spurious divergence. This innovation is applicable to other AdImEx time stepping schemes. Convergence, transport, and efficiency tests demonstrate substantial improvements over previous first-order AdImEx schemes, little phase and amplitude error, accurate transport on locally refined meshes with large Courant numbers, and promising efficiency characteristics for atmospheric transport.
- 28 September 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessAdHImEx: Adaptively High-Order Implicit-Explicit Transport for Large Time Steps
Adaptively Implicit-Explicit (AdImEx) time stepping provides stability for large time steps for mass-conservative transport, making it attractive for the numerical representation of advection in weather and climate prediction. Existing AdImEx transport schemes become first-order accurate in the large-time-step, implicit regime and may require multiple sparse matrix solutions per time step. This work introduces AdHImEx, a new AdImEx scheme that addresses both limitations simultaneously. AdHImEx provides second-order accuracy for large Courant numbers while requiring only a single matrix solution per time step. It is a Runge-Kutta scheme that blends third-order accurate explicit time stepping and Crank-Nicolson implicit time stepping to provide numerically verified stability for Courant numbers up to 100. AdHImEx is designed for flows that are predominantly explicit, activating implicit time stepping only where large Courant numbers occur. It retains the full efficiency and third-order accuracy of the explicit time stepping in regions with small Courant numbers. In this work, it is combined with a fifth-order accurate finite-volume discretisation in space for mass conservation. A second contribution of this work is a novel stage-dependent treatment of the divergence operator that preserves constancy despite different implicit and explicit Runge-Kutta stage time steps and spatially varying implicitness, removing spurious divergence. This innovation is applicable to other AdImEx time stepping schemes. Convergence, transport, and efficiency tests demonstrate substantial improvements over previous first-order AdImEx schemes, little phase and amplitude error, accurate transport on locally refined meshes with large Courant numbers, and promising efficiency characteristics for atmospheric transport.
- Open Access
- 22 June 2026
- Published by Copernicus GmbH
Investigating synoptic- and meso-scale severe weather phenomena using PMAP at sub-kilometer resolution
Understanding and forecasting severe weather events requires atmospheric models that not only reliably predict the circulation on the planetary and synoptic scales but also adequately represent meso- and micro-scale processes. Within extratropical cyclones, for instance, severe surface winds can occur along the bent-back extension of the warm front. In many cases, these are accompanied by a short-lived meso-scale vigorous jet, the so-called sting jet, whose underlying processes are an active subject of research. Much of the cyclone-induced heavy precipitation, in turn, is attributable to the frontal zones outside of the cyclone center. While a grid spacing of a few kilometers is typically sufficient to capture the occurrence of such events, models at higher resolution are needed to pinpoint the most disruptive wind gusts and precipitation peaks.The Portable Model for multi-scale Atmospheric Prediction (PMAP) is a flexible physics-based framework that supports horizontal grid spacings from a few kilometers down to tens of meters. This Python-based model is currently under active development at ECMWF, ETH Zürich, and CSCS. PMAP solves the non-hydrostatic and fully compressible equations using a bespoke finite-volume, 3D semi-implicit dynamical core coupled to state-of-the-art physical parametrizations. The key strengths of the model are its portability across diverse hardware architectures and its high computational performance, both enabled by the programming implementation with the GridTools for Python (GT4Py) domain-specific library.To evaluate the model representation of a real weather event, we simulate the passage of cyclone Goretti along the English Channel on 8 January 2026. This storm serves as a prime case study since Goretti induced strong wind gusts over the UK and France associated with a sting jet, as well as disruptive snow accumulation in Northern Germany. Forced by kilometer-scale DestinE forecasts at the lateral boundaries, PMAP is run over a regional domain at hectometer scale including nested large-eddy simulations. We benchmark PMAP using near-surface wind observations in the vicinity of the sting jet at the southern coast of the UK, and radar data covering Northern Germany. Beyond model evaluation, we leverage these simulations to investigate sting jet formation mechanisms and the structure of frontal precipitation at high spatial and temporal resolution.
- 22 June 2026
- Published by Copernicus GmbH
Open AccessInvestigating synoptic- and meso-scale severe weather phenomena using PMAP at sub-kilometer resolution
Understanding and forecasting severe weather events requires atmospheric models that not only reliably predict the circulation on the planetary and synoptic scales but also adequately represent meso- and micro-scale processes. Within extratropical cyclones, for instance, severe surface winds can occur along the bent-back extension of the warm front. In many cases, these are accompanied by a short-lived meso-scale vigorous jet, the so-called sting jet, whose underlying processes are an active subject of research. Much of the cyclone-induced heavy precipitation, in turn, is attributable to the frontal zones outside of the cyclone center. While a grid spacing of a few kilometers is typically sufficient to capture the occurrence of such events, models at higher resolution are needed to pinpoint the most disruptive wind gusts and precipitation peaks.The Portable Model for multi-scale Atmospheric Prediction (PMAP) is a flexible physics-based framework that supports horizontal grid spacings from a few kilometers down to tens of meters. This Python-based model is currently under active development at ECMWF, ETH Zürich, and CSCS. PMAP solves the non-hydrostatic and fully compressible equations using a bespoke finite-volume, 3D semi-implicit dynamical core coupled to state-of-the-art physical parametrizations. The key strengths of the model are its portability across diverse hardware architectures and its high computational performance, both enabled by the programming implementation with the GridTools for Python (GT4Py) domain-specific library.To evaluate the model representation of a real weather event, we simulate the passage of cyclone Goretti along the English Channel on 8 January 2026. This storm serves as a prime case study since Goretti induced strong wind gusts over the UK and France associated with a sting jet, as well as disruptive snow accumulation in Northern Germany. Forced by kilometer-scale DestinE forecasts at the lateral boundaries, PMAP is run over a regional domain at hectometer scale including nested large-eddy simulations. We benchmark PMAP using near-surface wind observations in the vicinity of the sting jet at the southern coast of the UK, and radar data covering Northern Germany. Beyond model evaluation, we leverage these simulations to investigate sting jet formation mechanisms and the structure of frontal precipitation at high spatial and temporal resolution.
- Open Access
- 14 March 2026
- Published by Copernicus GmbH
The Portable Model for multi-scale Atmospheric Prediction (PMAP): Capabilities and development workflows
The Portable Model for multi-scale Atmospheric Prediction (PMAP) is a numerical model currently under development at ECMWF and ETH aimed at large-eddy simulation (LES) of atmospheric flows at the weather-system scale. It builds on a non-hydrostatic, locally conserving, finite-volume, 3D semi-implicit dynamical core coupled to state-of-the-art physics parametrizations. Written entirely in Python, it leverages the GT4Py domain-specific language to achieve high performance and portability – running straightforwardly on laptops and GPU-accelerated HPC systems alike. The systematic separation of concerns between domain science (physics, numerics) and performance engineering (parallelisation, hardware optimisation) provides new avenues for model development, setup, and refinement, which we present in two related contributions.In this first contribution, we highlight PMAPs strengths as a numerical model framework to flexibly develop and refine numerical algorithms, as well as to implement and extend diagnostics to address research questions in atmospheric sciences. This is illustrated here with an LES tracer transport experiment over complex terrain. We first demonstrate PMAPs capability to perform decametre-scale LES using a height-based terrain-following vertical coordinate in steep terrain with slopes exceeding 80°. This is possible thanks to the locally conservative advection and the 3D semi-implicit integration scheme, which ensures stability of the integration and regularization of the flow. Moreover, we exemplarily show how the Python-based model formulation facilitates evaluating and improving various aspects of flux-form semi-Lagrangian tracer transport schemes in terms of directional splitting approaches and monotonic limiters, and how these impact simulated power spectra of the tracer fields. Lastly, we present how available model diagnostics can easily be extended to perform targeted analyses of sensitivities to implementation details.
- 14 March 2026
- Published by Copernicus GmbH
Open AccessThe Portable Model for multi-scale Atmospheric Prediction (PMAP): Capabilities and development workflows
The Portable Model for multi-scale Atmospheric Prediction (PMAP) is a numerical model currently under development at ECMWF and ETH aimed at large-eddy simulation (LES) of atmospheric flows at the weather-system scale. It builds on a non-hydrostatic, locally conserving, finite-volume, 3D semi-implicit dynamical core coupled to state-of-the-art physics parametrizations. Written entirely in Python, it leverages the GT4Py domain-specific language to achieve high performance and portability – running straightforwardly on laptops and GPU-accelerated HPC systems alike. The systematic separation of concerns between domain science (physics, numerics) and performance engineering (parallelisation, hardware optimisation) provides new avenues for model development, setup, and refinement, which we present in two related contributions.In this first contribution, we highlight PMAPs strengths as a numerical model framework to flexibly develop and refine numerical algorithms, as well as to implement and extend diagnostics to address research questions in atmospheric sciences. This is illustrated here with an LES tracer transport experiment over complex terrain. We first demonstrate PMAPs capability to perform decametre-scale LES using a height-based terrain-following vertical coordinate in steep terrain with slopes exceeding 80°. This is possible thanks to the locally conservative advection and the 3D semi-implicit integration scheme, which ensures stability of the integration and regularization of the flow. Moreover, we exemplarily show how the Python-based model formulation facilitates evaluating and improving various aspects of flux-form semi-Lagrangian tracer transport schemes in terms of directional splitting approaches and monotonic limiters, and how these impact simulated power spectra of the tracer fields. Lastly, we present how available model diagnostics can easily be extended to perform targeted analyses of sensitivities to implementation details.
- Open Access
- 14 March 2026
- Published by Copernicus GmbH
The Portable Model for multi-scale Atmospheric Prediction (PMAP): Sub-kilometre simulations of recent extreme weather
The Portable Model for multi-scale Atmospheric Prediction (PMAP) is a numerical model currently under development at ECMWF and ETH aimed at large-eddy simulation (LES) of atmospheric flows at the weather-system scale. It builds on a non-hydrostatic, locally conserving, finite-volume, 3D semi-implicit dynamical core coupled to state-of-the-art physics parametrizations. Written entirely in Python, it leverages the GT4Py domain-specific language to achieve high performance and portability – running straightforwardly on laptops and GPU accelerated HPC systems alike. The systematic separation of concerns between domain science (physics, numerics) and performance engineering (parallelisation, hardware optimisation) provides new avenues for model development, setup, and refinement, which we present in two related contributions.In this second contribution, we present two examples that demonstrate the benefits of resolving small-scale processes for simulating real weather and showcase how PMAP can flexibly be used as a research tool for atmospheric dynamics. (i) First, we demonstrate the benefits of sub-kilometer spatial resolution for accurately simulating the intensification of Hurricane Melissa in late October 2025 as compared to an operational km-scale model. (ii) Second, we present results from a process study of storm Éowyn, which brought record-strong winds to the British Isles in January 2025. LES of the low-level jet along the storm’s bent-back front not only accurately predicts peak wind speeds, but also resolves individual wind gusts in close agreement with observations. We highlight a range of diagnostic tools implemented in PMAP that make such analyses straightforward. Moreover, we demonstrate how the model can be employed to shed light on the atmospheric dynamical processes leading to the storm’s rapid intensification. Specifically, we show the importance of latent heat release by performing modified latent heating experiments, which in the PMAP framework are straightforward to set up, and quantitatively corroborate the crucial impact of cloud microphysical processes for the rapid intensification of Éowyn.
- 14 March 2026
- Published by Copernicus GmbH
Open AccessThe Portable Model for multi-scale Atmospheric Prediction (PMAP): Sub-kilometre simulations of recent extreme weather
The Portable Model for multi-scale Atmospheric Prediction (PMAP) is a numerical model currently under development at ECMWF and ETH aimed at large-eddy simulation (LES) of atmospheric flows at the weather-system scale. It builds on a non-hydrostatic, locally conserving, finite-volume, 3D semi-implicit dynamical core coupled to state-of-the-art physics parametrizations. Written entirely in Python, it leverages the GT4Py domain-specific language to achieve high performance and portability – running straightforwardly on laptops and GPU accelerated HPC systems alike. The systematic separation of concerns between domain science (physics, numerics) and performance engineering (parallelisation, hardware optimisation) provides new avenues for model development, setup, and refinement, which we present in two related contributions.In this second contribution, we present two examples that demonstrate the benefits of resolving small-scale processes for simulating real weather and showcase how PMAP can flexibly be used as a research tool for atmospheric dynamics. (i) First, we demonstrate the benefits of sub-kilometer spatial resolution for accurately simulating the intensification of Hurricane Melissa in late October 2025 as compared to an operational km-scale model. (ii) Second, we present results from a process study of storm Éowyn, which brought record-strong winds to the British Isles in January 2025. LES of the low-level jet along the storm’s bent-back front not only accurately predicts peak wind speeds, but also resolves individual wind gusts in close agreement with observations. We highlight a range of diagnostic tools implemented in PMAP that make such analyses straightforward. Moreover, we demonstrate how the model can be employed to shed light on the atmospheric dynamical processes leading to the storm’s rapid intensification. Specifically, we show the importance of latent heat release by performing modified latent heating experiments, which in the PMAP framework are straightforward to set up, and quantitatively corroborate the crucial impact of cloud microphysical processes for the rapid intensification of Éowyn.
- Open Access
- 13 March 2026
- Published by Copernicus GmbH
AdHImEx: Adaptively High-order Implicit-Explicit Time Stepping for Atmospheric Transport with Long Time Steps
Weather and climate models require atmospheric transport to be numerically stable, accurate, efficient, and mass-conserving. Adaptively Implicit-Explicit (AdImEx) time stepping can provide mass conservation and stability while taking long time steps for efficiency. However, challenges persist: (1) strongly reduced accuracy for large Courant numbers, and (2) a lack of preservation of uniform fields with substage time step sizes varying in space. We introduce AdHImEx, a novel AdImEx approach that resolves these issues, delivering an unconditionally stable, conservative, efficient advection scheme that preserves unity and increases temporal accuracy for large Courant numbers from first- to second-order. Remarkably, it requires only a single implicit solve per time step, performed with an iterative matrix solver, while retaining the full efficiency and third-order accuracy of the explicit time stepping in regions with small Courant numbers. Two-dimensional results will be shown.
- 13 March 2026
- Published by Copernicus GmbH
Open AccessAdHImEx: Adaptively High-order Implicit-Explicit Time Stepping for Atmospheric Transport with Long Time Steps
Weather and climate models require atmospheric transport to be numerically stable, accurate, efficient, and mass-conserving. Adaptively Implicit-Explicit (AdImEx) time stepping can provide mass conservation and stability while taking long time steps for efficiency. However, challenges persist: (1) strongly reduced accuracy for large Courant numbers, and (2) a lack of preservation of uniform fields with substage time step sizes varying in space. We introduce AdHImEx, a novel AdImEx approach that resolves these issues, delivering an unconditionally stable, conservative, efficient advection scheme that preserves unity and increases temporal accuracy for large Courant numbers from first- to second-order. Remarkably, it requires only a single implicit solve per time step, performed with an iterative matrix solver, while retaining the full efficiency and third-order accuracy of the explicit time stepping in regions with small Courant numbers. Two-dimensional results will be shown.
- Open Access
- 17 April 2025
- Published by Copernicus GmbH in Weather and Climate Dynamics
- Vol. 6 (2) , 447-469
Revealing the dynamics of a local Alpine windstorm using large-eddy simulations
The local atmospheric flow in mountainous terrain can be highly complex and deviate considerably from the ambient conditions. One example is a notorious local windstorm in a narrow and deep valley in northeastern Switzerland, known as the Laseyer, that had previously even caused a train derailment. This windstorm is characterized by strong southeasterly winds blowing perpendicular to the valley axis during strong northwesterly ambient flow conditions. We investigate the mechanism of this local windstorm and its sensitivity to changes in the prescribed ambient wind using large-eddy simulation (LES). The LESs are performed at a horizontal grid spacing of 30 m and by applying a terrain-following vertical coordinate with steep slopes of the real topography reaching nearly 80°. The simulations, driven by strong northwesterly ambient winds, successfully capture the flow reversal in the valley with quasi-periodically occurring short episodes of wind bursts regularly exceeding 20 m s−1 and in exceptional cases exceeding 35 m s−1. The flow reversal is explained by an amplifying interplay of (1) a recirculation region formed by flow separation in the lee of the upstream ridge and (2) a vortex caused by a positive pressure anomaly formed by the northwesterly winds impinging on the downstream mountain. This formation mechanism is supported by a simulation in which the height of the downstream mountain is reduced, resulting in a decrease in the strength of the reversed in-valley flow. In agreement with previous observational studies, a series of simulations with modified ambient wind conditions reveal that the intense gusts ( ) only occur in a narrow window of ambient wind directions and if its speed is at least 16 m s−1. Smoothing the topography in the LESs reduces the maximum wind speeds in the target region by 10 %–30 %. Overall, our semi-idealized LESs in complex and steep terrain reveal the three-dimensional structure and the mechanism of the local windstorm. Moreover, they point to the importance of the local topography and its complex interplay with the three-dimensional and transient flow leading to the in-valley flow reversal and strong winds that characterize the Laseyer. The study further highlights the importance of the topographic details for the quantitatively correct simulation of atmospheric flows in complex terrain.- 17 April 2025
- Published by Copernicus GmbH in Weather and Climate Dynamics
- Vol. 6 (2) , 447-469
Open AccessRevealing the dynamics of a local Alpine windstorm using large-eddy simulations
The local atmospheric flow in mountainous terrain can be highly complex and deviate considerably from the ambient conditions. One example is a notorious local windstorm in a narrow and deep valley in northeastern Switzerland, known as the Laseyer, that had previously even caused a train derailment. This windstorm is characterized by strong southeasterly winds blowing perpendicular to the valley axis during strong northwesterly ambient flow conditions. We investigate the mechanism of this local windstorm and its sensitivity to changes in the prescribed ambient wind using large-eddy simulation (LES). The LESs are performed at a horizontal grid spacing of 30 m and by applying a terrain-following vertical coordinate with steep slopes of the real topography reaching nearly 80°. The simulations, driven by strong northwesterly ambient winds, successfully capture the flow reversal in the valley with quasi-periodically occurring short episodes of wind bursts regularly exceeding 20 m s−1 and in exceptional cases exceeding 35 m s−1. The flow reversal is explained by an amplifying interplay of (1) a recirculation region formed by flow separation in the lee of the upstream ridge and (2) a vortex caused by a positive pressure anomaly formed by the northwesterly winds impinging on the downstream mountain. This formation mechanism is supported by a simulation in which the height of the downstream mountain is reduced, resulting in a decrease in the strength of the reversed in-valley flow. In agreement with previous observational studies, a series of simulations with modified ambient wind conditions reveal that the intense gusts ( ) only occur in a narrow window of ambient wind directions and if its speed is at least 16 m s−1. Smoothing the topography in the LESs reduces the maximum wind speeds in the target region by 10 %–30 %. Overall, our semi-idealized LESs in complex and steep terrain reveal the three-dimensional structure and the mechanism of the local windstorm. Moreover, they point to the importance of the local topography and its complex interplay with the three-dimensional and transient flow leading to the in-valley flow reversal and strong winds that characterize the Laseyer. The study further highlights the importance of the topographic details for the quantitatively correct simulation of atmospheric flows in complex terrain. - Open Access
- 15 March 2025
- Published by Copernicus GmbH
Developing ECMWF’s portable next-generation atmospheric dynamical core
Numerical weather prediction directly benefits from advancements in the accuracy, efficiency, and scalability of the atmospheric model. We present the development of a performance-portable high-level Python-based framework for the next-generation ECMWF global atmospheric dynamical core, enabling simulations at unprecedented numerical resolutions. This new model framework, called the Portable Model for Multi-Scale Atmospheric Prediction (PMAP), is an advancement of the Finite-Volume Module (FVM) originally developed in Fortran at ECMWF.A key feature of the global PMAP is its implementation with the latest version of the GridTools for Python (GT4Py) domain-specific library, named gt4py.next. This library is tailored to the efficient implementation of conservative finite-volume discretization methods that support, among others, ECMWF’s operational octahedral grid. Co-developed with various Swiss partners, the gt4py.next library itself is under continuous extension, optimization, and refinement alongside the PMAP. The model runs distributed across multiple nodes, enabling large-scale simulations on modern supercomputers with accelerators.We present recent model validation results and provide an analysis of its performance, portability, and scalability on latest European supercomputing platforms.
- 15 March 2025
- Published by Copernicus GmbH
Open AccessDeveloping ECMWF’s portable next-generation atmospheric dynamical core
Numerical weather prediction directly benefits from advancements in the accuracy, efficiency, and scalability of the atmospheric model. We present the development of a performance-portable high-level Python-based framework for the next-generation ECMWF global atmospheric dynamical core, enabling simulations at unprecedented numerical resolutions. This new model framework, called the Portable Model for Multi-Scale Atmospheric Prediction (PMAP), is an advancement of the Finite-Volume Module (FVM) originally developed in Fortran at ECMWF.A key feature of the global PMAP is its implementation with the latest version of the GridTools for Python (GT4Py) domain-specific library, named gt4py.next. This library is tailored to the efficient implementation of conservative finite-volume discretization methods that support, among others, ECMWF’s operational octahedral grid. Co-developed with various Swiss partners, the gt4py.next library itself is under continuous extension, optimization, and refinement alongside the PMAP. The model runs distributed across multiple nodes, enabling large-scale simulations on modern supercomputers with accelerators.We present recent model validation results and provide an analysis of its performance, portability, and scalability on latest European supercomputing platforms.
- Open Access
- 30 January 2025
- Published by Copernicus GmbH in Geoscientific Model Development
- Vol. 18 (2) , 529-546
Exploring a high-level programming model for the NWP domain using ECMWF microphysics schemes
We explore the domain-specific Python library GT4Py (GridTools for Python) for implementing a representative physical parametrization scheme and the related tangent-linear and adjoint algorithms from the Integrated Forecasting System (IFS) of ECMWF. GT4Py encodes stencil operators in an abstract and hardware-agnostic fashion, thus enabling more concise, readable, and maintainable scientific applications. The library achieves high performance by translating the application into targeted low-level coding implementations. Here, the main goal is to study the correctness and performance portability of the Python rewrites with GT4Py against the reference Fortran code and a number of automatically and manually ported variants created by ECMWF. The present work is part of a larger cross-institutional effort to port weather and climate models to Python with GT4Py. The focus of the current work is the IFS prognostic cloud microphysics scheme, a core physical parametrization represented by a comprehensive code that takes a significant share of the total forecast model execution time. In order to verify GT4Py for numerical weather prediction (NWP) systems, we put additional emphasis on the implementation and validation of the tangent-linear and adjoint model versions which are employed in data assimilation. We benchmark all prototype codes on three European supercomputers characterized by diverse graphics processing unit (GPU) and central processing unit (CPU) hardware, node designs, software stacks, and compiler suites. Once the application is ported to Python with GT4Py, we find excellent portability, competitive GPU performance, and robust execution in all tested scenarios including with single precision.
- 30 January 2025
- Published by Copernicus GmbH in Geoscientific Model Development
- Vol. 18 (2) , 529-546
Open AccessExploring a high-level programming model for the NWP domain using ECMWF microphysics schemes
We explore the domain-specific Python library GT4Py (GridTools for Python) for implementing a representative physical parametrization scheme and the related tangent-linear and adjoint algorithms from the Integrated Forecasting System (IFS) of ECMWF. GT4Py encodes stencil operators in an abstract and hardware-agnostic fashion, thus enabling more concise, readable, and maintainable scientific applications. The library achieves high performance by translating the application into targeted low-level coding implementations. Here, the main goal is to study the correctness and performance portability of the Python rewrites with GT4Py against the reference Fortran code and a number of automatically and manually ported variants created by ECMWF. The present work is part of a larger cross-institutional effort to port weather and climate models to Python with GT4Py. The focus of the current work is the IFS prognostic cloud microphysics scheme, a core physical parametrization represented by a comprehensive code that takes a significant share of the total forecast model execution time. In order to verify GT4Py for numerical weather prediction (NWP) systems, we put additional emphasis on the implementation and validation of the tangent-linear and adjoint model versions which are employed in data assimilation. We benchmark all prototype codes on three European supercomputers characterized by diverse graphics processing unit (GPU) and central processing unit (CPU) hardware, node designs, software stacks, and compiler suites. Once the application is ported to Python with GT4Py, we find excellent portability, competitive GPU performance, and robust execution in all tested scenarios including with single precision.
- Open Access
- 23 January 2025
- Published by Copernicus GmbH
- Vol. 2024, 1-36
Final author comments on egusphere-2024-3461
The local atmospheric flow in mountainous terrain can be highly complex and deviate considerably from the ambient conditions. One example is a notorious local windstorm in a narrow and deep valley in north-eastern Switzerland, known as the Laseyer, that had previously even caused a train derailment. This windstorm is characterized by strong south-easterly winds blowing perpendicular to the valley axis during strong north-westerly ambient flow conditions. We investigate the mechanism of this local windstorm and its sensitivity to changes in the prescribed ambient wind using large-eddy simulation (LES). The LES are performed using the Portable Model for Multi-Scale Atmospheric Prediction (PMAP) at a horizontal grid spacing of 30 m and applying a terrain-following vertical coordinate with steep slopes of the real topography reaching nearly 80°. The simulations, driven by strong north-westerly ambient winds, successfully capture the flow reversal in the valley with quasi-periodically occurring short episodes of wind bursts regularly exceeding 20 m s-1 and in exceptional cases exceeding 35 m s-1. The flow reversal is explained by an amplifying interplay of (1) a recirculation region formed by flow separation in the lee of the upstream ridge, and (2) a vortex caused by a positive pressure anomaly formed by the north-westerly winds impinging on the downstream mountain. This formation mechanism is supported by a simulation in which the height of the downstream mountain is reduced, resulting in a decrease in the strength of the reversed in-valley flow. In agreement with previous observational studies, a series of simulations with modified ambient wind conditions reveal that the intense gusts (> 20 m s-1) only occur in a narrow window of ambient wind directions and if its speed is at least 16 m s-1. Smoothing the topography in the LES reduces the maximum wind speeds in the target region by 10–30 %. Overall, our semi-idealized LES in complex and steep terrain reveal the three-dimensional structure and the mechanism of the local windstorm. Moreover, they point to the importance of the local topography and its complex interplay with the three-dimensional and transient flow leading to the in-valley flow reversal and strong winds that characterize the Laseyer. The study further highlights the importance of the topographic details for the quantitatively correct simulation of atmospheric flows in complex terrain.
- 23 January 2025
- Published by Copernicus GmbH
- Vol. 2024, 1-36
Open AccessFinal author comments on egusphere-2024-3461
The local atmospheric flow in mountainous terrain can be highly complex and deviate considerably from the ambient conditions. One example is a notorious local windstorm in a narrow and deep valley in north-eastern Switzerland, known as the Laseyer, that had previously even caused a train derailment. This windstorm is characterized by strong south-easterly winds blowing perpendicular to the valley axis during strong north-westerly ambient flow conditions. We investigate the mechanism of this local windstorm and its sensitivity to changes in the prescribed ambient wind using large-eddy simulation (LES). The LES are performed using the Portable Model for Multi-Scale Atmospheric Prediction (PMAP) at a horizontal grid spacing of 30 m and applying a terrain-following vertical coordinate with steep slopes of the real topography reaching nearly 80°. The simulations, driven by strong north-westerly ambient winds, successfully capture the flow reversal in the valley with quasi-periodically occurring short episodes of wind bursts regularly exceeding 20 m s-1 and in exceptional cases exceeding 35 m s-1. The flow reversal is explained by an amplifying interplay of (1) a recirculation region formed by flow separation in the lee of the upstream ridge, and (2) a vortex caused by a positive pressure anomaly formed by the north-westerly winds impinging on the downstream mountain. This formation mechanism is supported by a simulation in which the height of the downstream mountain is reduced, resulting in a decrease in the strength of the reversed in-valley flow. In agreement with previous observational studies, a series of simulations with modified ambient wind conditions reveal that the intense gusts (> 20 m s-1) only occur in a narrow window of ambient wind directions and if its speed is at least 16 m s-1. Smoothing the topography in the LES reduces the maximum wind speeds in the target region by 10–30 %. Overall, our semi-idealized LES in complex and steep terrain reveal the three-dimensional structure and the mechanism of the local windstorm. Moreover, they point to the importance of the local topography and its complex interplay with the three-dimensional and transient flow leading to the in-valley flow reversal and strong winds that characterize the Laseyer. The study further highlights the importance of the topographic details for the quantitatively correct simulation of atmospheric flows in complex terrain.
- Open Access
- 20 January 2025
- Published by Copernicus GmbH
Investigating a local windstorm using measurements and large-eddy simulations
During a winter storm in January 2007, a train derailed due to strong winds in a narrow valley in northeastern Switzerland. The accident was attributed to the Laseyer, a local windstorm characterized by flow reversal that manifests as easterly to southeasterly winds at the valley floor during strong prevailing northwesterly winds above. We analyze case studies of the local windstorm using sonic anemometer and Doppler lidar measurements. The data reveal a highly turbulent flow in the narrow valley and extreme wind speeds exceeding 45 m/s during Laseyer conditions.Additionally, we use a newly developed large-eddy simulation (LES) atmospheric model to improve our understanding of the local windstorm. The model is implemented in a Python environment with the GT4Py (GridTools for Python) domain-specific library to enable performance portability. Robust and efficient solution of the nonhydrostatic compressible equations is achieved using a finite-volume semi-implicit discretization following ECMWF’s IFS-FVM. LESs are performed above the highly complex terrain of northeastern Switzerland, which leads to extremely steep slopes exceeding 70°. With these LESs, we identify the mechanism behind the local windstorm, study its sensitivity to ambient flow conditions, and characterize the flow conditions in the narrow valley.
- 20 January 2025
- Published by Copernicus GmbH
Open AccessInvestigating a local windstorm using measurements and large-eddy simulations
During a winter storm in January 2007, a train derailed due to strong winds in a narrow valley in northeastern Switzerland. The accident was attributed to the Laseyer, a local windstorm characterized by flow reversal that manifests as easterly to southeasterly winds at the valley floor during strong prevailing northwesterly winds above. We analyze case studies of the local windstorm using sonic anemometer and Doppler lidar measurements. The data reveal a highly turbulent flow in the narrow valley and extreme wind speeds exceeding 45 m/s during Laseyer conditions.Additionally, we use a newly developed large-eddy simulation (LES) atmospheric model to improve our understanding of the local windstorm. The model is implemented in a Python environment with the GT4Py (GridTools for Python) domain-specific library to enable performance portability. Robust and efficient solution of the nonhydrostatic compressible equations is achieved using a finite-volume semi-implicit discretization following ECMWF’s IFS-FVM. LESs are performed above the highly complex terrain of northeastern Switzerland, which leads to extremely steep slopes exceeding 70°. With these LESs, we identify the mechanism behind the local windstorm, study its sensitivity to ambient flow conditions, and characterize the flow conditions in the narrow valley.