Julian Quinting

Midlatitude winter storms cause substantial socioeconomic impact through strong winds and heavy precipitation. Challenges remain in understanding the multiscale dynamics of these systems and accurately predicting their impacts. New observational and modeling capabilities enable tackling these challenges and motivate the “North Atlantic Waveguide, Dry Intrusion, and Downstream Impact Campaign (NAWDIC)” scheduled for winter 2025/26. We present a case study of two high-impact winter cyclones, Eunice and Franklin, in February 2022 to illustrate the interacting processes across scales. First, Eunice brought damaging wind gusts, and subsequently Franklin caused heavy precipitation and flooding. Both systems and their impacts were linked in a Lagrangian sense through a dry intrusion (DI) airstream. The descending DI reaching the cold sector of Eunice produced gale force winds and strong ocean evaporation leading to a rapid moistening of the originally dry air mass. These moistened air masses were transported westward around the subtropical anticyclone before feeding an atmospheric river over the North Atlantic 2 days later. As the air masses ascended in the warm conveyor belt of cyclone Franklin, they caused heavy precipitation and flooding across the British Isles. Based on the case study, we highlight existing observational gaps potentially leading to forecast uncertainties. Finally, we emphasize the need for the internationally coordinated NAWDIC airborne campaign and present measurement strategies aimed at improving our understanding of the coupled multiscale processes leading to high-impact weather.

Extratropical cyclones frequently exhibit pronounced poleward propagation during their life cycle. This behavior is typically associated with the poleward advection of a low-level PV anomaly by an upper-level PV anomaly located to its west, which can be enhanced by diabatic production of positive low- to mid-level PV (LPV) through latent heat release. In CMIP6 models, the storm tracks tend to be too zonal, particularly in the North Atlantic, and the frequency and intensity of rapidly deepening cyclones are often underestimated. Such biases may partly arise from misrepresentation of the magnitude of diabatic processes and/or from the dynamical response of cyclone propagation to those processes.The aim of this study is to assess the contribution of latent heating to the poleward propagation of extratropical cyclones and to evaluate how both the magnitude of LPV and the associated dynamical response contribute to the storm track biases in CMIP6 models. Using ERA5 reanalysis and CMIP6 model data for the period 1979–2014, this study applies ensemble sensitivity analysis and cyclone composite methods to quantify the sensitivity of cyclone poleward propagation, measured by the cyclone meridional velocity at the time of maximum intensity, to LPV associated with latent heating. The analysis is conducted over the North Atlantic and North Pacific basins, considering both western and eastern sectors.In ERA5, preliminary results show that North Atlantic cyclones have larger LPV than North Pacific cyclones throughout the entire development phase. Within the North Atlantic, although latent heating is stronger in western cyclones than in eastern ones, the sensitivity of poleward propagation to LPV is largest for eastern cyclones. In contrast, in the North Pacific, cyclones in the eastern sector show slightly stronger latent heating than those in the western sector. However, the sensitivity of poleward propagation to LPV is largest for western cyclones.The CMIP6 models evaluated so far are able to capture the overall structure of LPV and the sensitivity of poleward motion to latent heating in extratropical cyclones across both oceans and sectors, as well as the differences between them. However, model resolution appears to impact the accuracy in representing the magnitude of these sensitivities, particularly for eastern North Atlantic cyclones. This may help explain the reduced storm track biases found in higher resolution CMIP6 models.These results suggest that the poleward motion of western North Pacific and eastern North Atlantic cyclones is more strongly responsive to diabatic forcing via latent heat release, even though the magnitude of latent heating is smaller in those sectors. In contrast, western North Atlantic and eastern North Pacific cyclones appear to be more directly controlled by dry baroclinic processes. Finally, improving the representation of moist processes and LPV generation in climate models is essential for reducing biases in storm track orientation, cyclone intensity, and associated uncertainties in future climate projections.

The Madden–Julian Oscillation (MJO) and stratospheric polar vortex (SPV) are prominent sources of subseasonal predictability in the extratropics. It has been shown that the joint interaction of the MJO and the SPV can modulate the preferred phase of the North Atlantic Oscillation (NAO) and the occurrence of weather regimes. However, improving numerical weather prediction (NWP) at 3‐week lead times remain underexplored. This study investigates how MJO and SPV phases affect Greenland Blocking (GL) activity and integrates atmospheric state information into a neural network to enhance week 3 weather regime activity forecasts. We define a weather regime activity metric using European Centre for Medium‐Range Weather Forecasts (ECMWF) reanalysis and reforecasts. In reanalyses we find increased GL activity following MJO phases 7, 8, and 1, as well as weak SPV phases, indicating climatological windows of opportunity in line with previous studies. However, ECMWF forecast skill improves only in MJO phases 8 and 1 and weak SPV phases, identifying somewhat different model windows of opportunity. Next, we explore using these findings in postprocessing tools. Climatological forecasts based on MJO/SPV–NAO relationships provide a purely statistical approach to subseasonal GL activity forecasting, independent of NWP models. Notably, MJO‐conditioned climatological forecasts show clear signals when evaluated against observed GL activity. Statistical–dynamical models, using neural networks that combine historical atmospheric state data with NWP‐derived weather regime metrics, improve weather regime activity forecasts across all regimes considered, achieving an absolute accuracy increase of 5.8 percentage points in forecasting the dominant weather regime compared with ECMWF. This is particularly beneficial to blocking in the European domain, where NWP models often underperform. Atmospheric conditioned and neural network forecasts serve as valuable decision‐support tools alongside NWP models, enhancing the reliability of subseasonal predictions.

Extratropical cyclones are the primary driver of midlatitude weather variability, including extreme events with heavy precipitation and severe wind gusts, potentially leading to widespread damage. Both baroclinicity and diabatic heating contribute to cyclone intensification. Examples of severe European winter storms with rapid intensification enhanced through diabatic heating include Lothar (1999), Klaus (2009), and Xynthia (2010). Diabatic heating through cloud condensational processes in extratropical cyclones is associated with the warm conveyor belt (WCB) airstream, among others. Focusing on European winter storms, this study provides a comprehensive assessment of structural differences important for cyclone intensification between cyclones that intensify to a relatively large extent through diabatic processes and cyclones that intensify mostly through baroclinic processes. Considering 247 winter storms affecting western and central Europe from 1979 to 2023, we combine a WCB diagnostic with the pressure tendency equation to quantify the influence of diabatic heating on deepening processes. Our results indicate that diabatic processes have a mean contribution to cyclone intensification of 26.1% (median at 25.3%, 25th and 75th percentiles at 15.2% and 33.4%). Cyclones with a relatively large diabatic influence ( diabatic contribution) have similar minimum mean sea‐level pressure compared with cyclones with a small diabatic influence ( diabatic contribution) but exhibit steeper deepening rates, stronger northward displacement, increased precipitation, stronger wind gusts, and more WCB activity. Moreover, they are characterised by increased moisture and warmer temperatures in the WCB inflow, facilitating stronger deepening rates driven by diabatic heating. Our results suggest that WCB activity is a valuable first indicator of the diabatic contribution to cyclone intensification. These findings may be useful for predicting and estimating the risks associated with winter storms that have a relatively large influence of diabatic processes in numerical weather predictions and climate projections.

This study explores the use of relaxation experiments in machine learning-based weather prediction (MLWP) models to identify sources of subseasonal forecast errors and compares their effectiveness to a traditional numerical weather prediction (NWP) system. Relaxation involves nudging specific regions of a model toward reanalysis data to isolate their influence on forecast skill. We apply this technique to two MLWP models, Pangu-Weather (fully data-driven) and NeuralGCM (hybrid) and compare them to the Unified Forecast System (UFS) during Week 3–4 forecasts of two major precipitation events in western North America in winter 2022/2023, both linked to Madden-Julian Oscillation (MJO) activity. MLWP models outperform UFS at subseasonal lead times. While tropical relaxation improves all forecasts, gains are greater for UFS, reflecting MLWP models’ stronger baseline performance. These results highlight relaxation experiment as a low-cost, effective diagnostic for understanding and improving subseasonal forecasts, especially in emerging MLWP systems.

In the past 5 years, data‐driven prediction models and Machine Learning (ML) techniques have revolutionized weather forecasting. Meteorological services around the world are now developing ML components to enhance (or even replace) their numerical weather prediction systems. This shift creates new challenges and opportunities for universities and research centers, calling for a much closer cooperation of meteorology with mathematics and computer sciences, updates of teaching curricula, and new research infrastructures and strategies. To address these challenges, an interdisciplinary team of scientists from the Karlsruhe Institute of Technology (KIT) and the German Meteorological Service (DWD) created the TEstbed for Exploring Machine LEarning in Atmospheric Prediction (TEEMLEAP). Implemented on KIT's supercomputer HoreKa, the TEEMLEAP testbed simulates the entire operational weather forecasting chain using ERA5 reanalysis data as pseudo‐observations and DWD's Basic Cycling environment for conducting assimilation‐prediction‐cycling experiments. Moreover, first steps are taken toward the integration of new data‐driven components like FourCastNet and ML‐based post‐processing methods. The TEEMLEAP testbed allows systematic investigation of a wide range of issues related to weather forecasting such as optimizing the observational system, uncertainty quantification, and developing hybrid systems that integrate ML with physics‐based models. This document outlines the testbed's setup, demonstrates its functionality with a pilot experiment, and discusses examples of potential applications. Future plans include creating educational modules and developing a higher‐resolution regional version of the testbed that could be used for assimilating field campaign observations.

Dry intrusions emerge as an important component of the extratropical circulation. Governing damaging winds directly, and triggering moisture uptake – thus influencing heavy precipitation indirectly, it is necessary to observe dry intrusions and accurately predict them. NAWDIC is a new initiative for an international field campaign aiming to advance our understanding and modelling of the synoptic- to micro-scale dynamical and physical processes associated with the triggering of severe wind gusts, heavy precipitation, and cold air outbreaks in the North Atlantic-European region. More specifically, NAWDIC will focus on the physical understanding and quantification of the interactions between mesoscale tropopause structure, dry intrusion airstream, and downstream dynamics near the surface cold front for the evolution of high-impact weather in winter extratropical cyclones. Initiated as a campaign with the German high-altitude, long-range aircraft HALO, NAWDIC has become a major international effort. It consists of multiple components, which are planned as stand-alone measurement campaigns by different groups in Europe and North America but will benefit from synergies when coordinated under the umbrella of NAWDIC. During the six-week core observation period in January and February 2026, the HALO aircraft will operate from Shannon (Ireland) and will be complemented by in-situ and remote sensing instruments onboard of additional mid- to long range aircrafts as well as dense ground-based measurements at the French Atlantic coast. Further, the observation activities will be accompanied by a modelling component in collaboration with weather services including the assimilation of NAWDIC observations. This presentation will outline the scientific idea behind NAWDIC and the observational strategies for addressing its goals.

Subseasonal weather predictions for the extratropics remain a significant challenge. Although sources of extratropical subseasonal predictability are linked to slowly evolving components of the atmospheric system, such as tropical modes of variability (e.g., the Madden-Julian Oscillation and tropical waves), these sources are not yet fully leveraged due to systematic errors in numerical weather prediction models. In particular, short-term errors in the tropics can degrade forecast skill in the extratropics on subseasonal timescales. However, the specific tropical regions where such errors most strongly influence extratropical forecast skill remain unclear. Relaxation experiments using NWP models provide a means to identify these key regions, though such experiments are computationally expensive, especially when run in ensemble mode. In this study, we utilize machine learning-based weather prediction models to perform relaxation experiments across various tropical regions on the subseasonal timescale. Probabilistic forecasts are generated using initial conditions from the Ensemble of Data Assimilations (EDA) system of the European Centre for Medium-Range Weather Forecasts (ECMWF). By evaluating reforecasts for a five year period (2020-2024), which lies outside the training period of the models, this study systematically investigates the role of nudging techniques and the influence of tropical variability, particularly the MJO, on subseasonal forecast skill. Additionally, we evaluate the experiments conditioned on the state of the MJO to assess the regional contributions of tropical predictability to forecast improvements in the extratropics. Our results underscore the potential of nudging as both a diagnostic tool and a means to enhance extended-range forecasting skill, especially when combined with emerging machine learning-based prediction approaches.

Synoptic- and large-scale features of atmospheric flow such as extratropical cyclones, Rossby wave packets, and atmospheric blocking modulate mid-latitude weather and climate. However, several studies have shown strong biases in the frequency and location of these features in state-of-the-art global climate models. One notable and persistent bias is an underestimation of the atmospheric blocking frequency in the Euro-Atlantic region. In this study, we validate the representation of synoptic- and large-scale features of the North Atlantic flow in eight climate models of the Coupled Model Intercomparison Project Phase 6 (CMIP6), taking the ERA5 reanalysis as a reference. Validation includes atmospheric blocking, storm tracks, eddy heat and moisture fluxes, and warm conveyor belts (WCBs). The selected CMIP6 models underestimate the atmospheric blocking frequency over the eastern North Atlantic and Europe in winter (December to February) by up to 80 %. The frequency biases result from combined biases at different spatial and temporal scales described in the following. First, we define the background flow as the most frequent value of the daily time series of the meridional gradient of geopotential height at 500 hPa. In the CMIP6 models, the strongest latitudinal geopotential gradients are shifted equatorward in the North Atlantic basin. This shift favours more zonal and stronger winds to the south of the climatological jet. The differences in the background flow affect Rossby wave breaking and blocking onset and persistence, as illustrated by analysing the eddies in the Euro-Atlantic region. We find an equatorward shift in the eddies in CMIP6 that accelerates the mean flow in the exit region of the Atlantic jet, as indicated by a reduction in the divergence of E vectors. The shift in the eddies leads to a less diffluent flow in the east Atlantic and, thus, a less favourable flow for blocking formation. Second, we find a negative bias in WCB outflow frequency in the CMIP6 models in the North Atlantic. Reduced WCB outflow indicates weaker transport of low potential vorticity (PV) from the lower to the upper troposphere by moist diabatic processes and consequently weaker downstream ridge amplification and, therefore, a less diffluent flow over the eastern North Atlantic and weaker diabatic contributions to blocking. The negative WCB outflow bias can be linked to an underestimation of the meridional moisture transport at low levels in the climatological WCB inflow area in the western North Atlantic. Thus, the misrepresentation of moist processes contributes to the negative blocking biases. Accordingly, an improved representation of the moist processes in the next generation of climate models could improve the blocking representation.

In the austral summer, parts of southeastern Australia are prone to heavy rainfall that causes major riverine flooding and fatalities. Easterly flow associated with an anticyclone in the Tasman Sea, large moisture transports from the Coral Sea, and upper tropospheric cyclonic disturbances all contribute to these heavy rainfall episodes. However, questions regarding their synoptic dynamics remain, including which of these ingredients are the most critical. These questions are addressed by comparing composite pressure and moisture fields of heavy rainfall days over selected regions with non‐heavy rainfall days that have a similar synoptic pattern. A synoptic climatology is constructed for this purpose by ‐means cluster analysis of 500 hPa geopotential height anomalies from the European Centre for Medium‐range Weather Forecasts Reanalysis v5, for all December to March days over a period of 40 years. Heavy rainfall days in the wettest clusters have negative 500 hPa geopotential height anomalies immediately west of the affected region that are stronger on average than those of non‐heavy rainfall days. Their accompanying distributions of surface pressure, precipitable water, and vertical motion are consistent with cyclonic baroclinic development and are preceded by anticyclonic Rossby wave breaking. Heavy rainfall days also show an increased frequency of blocking near E; however, this peaks 1–2 days after the onset of heavy rain. Regional rainfall in these clusters shows strong sensitivity to lower pressure immediately westward but little sensitivity to high pressure in the Tasman Sea until after the commencement of rain. A companion study using the same cluster analysis illustrated the link between anticyclonic Rossby wave breaking and heatwaves in southeastern Australia. These latest results highlight the upper cyclonic anomalies that often form on the equatorward flank of anticyclonic Rossby wave breaking as the key ingredient separating days with a favourable synoptic‐scale pattern of surface high pressure into those that rain heavily and those that do not.