Articles | Volume 3, issue 1
https://doi.org/10.5194/wcd-3-113-2022
© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/wcd-3-113-2022
© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Research article
| 
01 Feb 2022
Research article |
| 01 Feb 2022
| 01 Feb 2022
Automated detection and classification of synoptic-scale fronts from atmospheric data grids
Stefan Niebler
CORRESPONDING AUTHOR
Institut für Informatik, Johannes Gutenberg-Universität Mainz, Staudingerweg 9, 55128 Mainz, Germany
Annette Miltenberger
Institut für Physik der Atmosphäre, Johannes Gutenberg-Universität Mainz, Johann-Joachim-Becher-Weg 21 , 55128 Mainz, Germany
Bertil Schmidt
Institut für Informatik, Johannes Gutenberg-Universität Mainz, Staudingerweg 9, 55128 Mainz, Germany
Peter Spichtinger
Institut für Physik der Atmosphäre, Johannes Gutenberg-Universität Mainz, Johann-Joachim-Becher-Weg 21 , 55128 Mainz, Germany
Related authors
Helena Zoe Schuh, Philipp Reutter, Stefan Niebler, and Peter Spichtinger
EGUsphere, https://doi.org/10.5194/egusphere-2025-2498, https://doi.org/10.5194/egusphere-2025-2498, 2025
Short summary
Short summary
We studied ice-supersaturated regions in the upper troposphere and lower stratosphere where high humidity can lead to cloud and contrail formation. Using data from 2010 to 2020, we found these regions to have fractal characteristics by applying an area-perimeter method. The fractal dimension follows a seasonal cycle. Our results can help improve climate models and have possible implications on contrail mitigation.
Sarah Brüning, Stefan Niebler, and Holger Tost
Atmos. Meas. Tech., 17, 961–978, https://doi.org/10.5194/amt-17-961-2024, https://doi.org/10.5194/amt-17-961-2024, 2024
Short summary
Short summary
We apply the Res-UNet to derive a comprehensive 3D cloud tomography from 2D satellite data over heterogeneous landscapes. We combine observational data from passive and active remote sensing sensors by an automated matching algorithm. These data are fed into a neural network to predict cloud reflectivities on the whole satellite domain between 2.4 and 24 km height. With an average RMSE of 2.99 dBZ, we contribute to closing data gaps in the representation of clouds in observational data.
Alena Kosareva, Stamen Dolaptchiev, Peter Spichtinger, and Ulrich Achatz
Geosci. Model Dev., 18, 6117–6133, https://doi.org/10.5194/gmd-18-6117-2025, https://doi.org/10.5194/gmd-18-6117-2025, 2025
Short summary
Short summary
This study improves how we predict ice formation in clouds by accounting for variable ice sizes and different weather conditions. Using simulations, we developed a more accurate method that works efficiently, making it suitable for application in weather and climate prediction models. The new approach is numerically verified and provides precise predictions of ice formation events and reliable estimates of key parameters.
Tim Lüttmer, Annette Miltenberger, and Peter Spichtinger
Atmos. Chem. Phys., 25, 10245–10265, https://doi.org/10.5194/acp-25-10245-2025, https://doi.org/10.5194/acp-25-10245-2025, 2025
Short summary
Short summary
We investigate ice formation pathways in a warm conveyor belt case study. We employ a multi-phase microphysics scheme that distinguishes between ice from different nucleation processes. Ice crystals in the cirrus outflow mostly stem from in situ formation. Hence, they were formed directly from the vapor phase. Sedimentational redistribution modulates cirrus properties and leads to disagreement between cirrus origin classifications based on thermodynamic history and nucleation processes.
Helena Zoe Schuh, Philipp Reutter, Stefan Niebler, and Peter Spichtinger
EGUsphere, https://doi.org/10.5194/egusphere-2025-2498, https://doi.org/10.5194/egusphere-2025-2498, 2025
Short summary
Short summary
We studied ice-supersaturated regions in the upper troposphere and lower stratosphere where high humidity can lead to cloud and contrail formation. Using data from 2010 to 2020, we found these regions to have fractal characteristics by applying an area-perimeter method. The fractal dimension follows a seasonal cycle. Our results can help improve climate models and have possible implications on contrail mitigation.
Anna Breuninger, Philipp Joppe, Jonas Wilsch, Cornelis Schwenk, Heiko Bozem, Nicolas Emig, Laurin Merkel, Rainer Rossberg, Timo Keber, Arthur Kutschka, Philipp Waleska, Stefan Hofmann, Sarah Richter, Florian Ungeheuer, Konstantin Dörholt, Thorsten Hoffmann, Annette Miltenberger, Johannes Schneider, Peter Hoor, and Alexander L. Vogel
EGUsphere, https://doi.org/10.5194/egusphere-2025-3129, https://doi.org/10.5194/egusphere-2025-3129, 2025
Short summary
Short summary
This study investigates molecular organic aerosol composition in the upper troposphere and lower stratosphere from an airborne campaign over Central Europe in summer 2024. Via ultra-high-performance liquid chromatography and high-resolution mass spectrometry of tropospheric and stratospheric filter samples, we identified various organic compounds. Our findings underscore the significant cross-tropopause transport of biogenic secondary organic aerosol and anthropogenic pollutants.
Patrick Konjari, Christian Rolf, Martina Krämer, Armin Afchine, Nicole Spelten, Irene Bartolome Garcia, Annette Miltenberger, Nicolar Emig, Philipp Joppe, Johannes Schneider, Yun Li, Andreas Petzold, Heiko Bozem, and Peter Hoor
EGUsphere, https://doi.org/10.5194/egusphere-2025-2847, https://doi.org/10.5194/egusphere-2025-2847, 2025
Short summary
Short summary
We investigated how a powerful storm over southern Sweden in June 2024 transported ice particles and moist air into the normally dry stratosphere. We observed unusually high water vapor and ice levels up to 1.5 kilometers above the tropopause. Although the extra water vapor lasted only a few days to weeks, it shows how such storms can temporarily alter the upper atmosphere’s composition.
Philipp Reutter and Peter Spichtinger
EGUsphere, https://doi.org/10.5194/egusphere-2025-2474, https://doi.org/10.5194/egusphere-2025-2474, 2025
Short summary
Short summary
We present a new technique to determine the tropopause based on the gradient of relative humidity over ice. This reflects the character of the tropopause as a transport barrier very well, both in individual vertical profiles and in the long-term average. The results of the investigations using radio sondes are also supported by theoretical considerations.
Cornelis Schwenk, Annette Miltenberger, and Annika Oertel
EGUsphere, https://doi.org/10.5194/egusphere-2025-1816, https://doi.org/10.5194/egusphere-2025-1816, 2025
Short summary
Short summary
We studied how different parameter choices concerning cloud processes affect the simulated transport of water and ice into the upper atmosphere (which affects the greenhouse effect) during a weather system called a warm conveyor belt. Using a set of model experiments, we found that some parameters have a strong effect on humidity and ice, especially during fast ascents. These findings could help improve weather and climate models and may also be relevant for future climate engineering studies.
Philipp Joppe, Johannes Schneider, Jonas Wilsch, Heiko Bozem, Anna Breuninger, Joachim Curtius, Martin Ebert, Nicolas Emig, Peter Hoor, Sadath Ismayil, Konrad Kandler, Daniel Kunkel, Isabel Kurth, Hans-Christoph Lachnitt, Yun Li, Annette Miltenberger, Sarah Richter, Christian Rolf, Lisa Schneider, Cornelis Schwenk, Nicole Spelten, Alexander L. Vogel, Yafang Cheng, and Stephan Borrmann
EGUsphere, https://doi.org/10.5194/egusphere-2025-1346, https://doi.org/10.5194/egusphere-2025-1346, 2025
Short summary
Short summary
We show measurements of a filament with biomass burning influence transported by a warm conveyor belt (WCB) into the tropopause region over Europe. The pollution originates from Canadian forest fires and is transported in the lower troposphere towards Europe. The WCB transport is followed by mixing with air masses of stratospheric chemical signatures. We hypothesize that this mixing leads to a change in the vertical gradient of the potential temperature.
Tim Lüttmer, Peter Spichtinger, and Axel Seifert
Atmos. Chem. Phys., 25, 4505–4529, https://doi.org/10.5194/acp-25-4505-2025, https://doi.org/10.5194/acp-25-4505-2025, 2025
Short summary
Short summary
We investigate ice formation pathways in idealized convective clouds using a novel microphysics scheme that distinguishes between five ice classes each with their own unique formation mechanism. Ice crystals from rime splintering form the lowermost layer of ice crystals around the updraft core. The majority of ice crystals in the anvil of the convective cloud stems from frozen droplets. Ice stemming from homogeneous and deposition nucleation was only relevant in the overshoot.
Matteo Krüger, Tommaso Galeazzo, Ivan Eremets, Bertil Schmidt, Ulrich Pöschl, Manabu Shiraiwa, and Thomas Berkemeier
EGUsphere, https://doi.org/10.5194/egusphere-2025-1191, https://doi.org/10.5194/egusphere-2025-1191, 2025
Short summary
Short summary
This work uses machine learning to predict saturation vapor pressures of atmospherically-relevant organic compounds, crucial for partitioning of secondary organic aerosol (SOA). We introduce a new method using graph convolutional neural networks, in which molecular graphs enable the model to capture molecular connectivity better than with non-structural embeddings. The method shows strong agreement with experimentally determined vapor pressures, and outperforms existing estimation methods.
Nicolas Emig, Annette K. Miltenberger, Peter M. Hoor, and Andreas Petzold
EGUsphere, https://doi.org/10.5194/egusphere-2024-3919, https://doi.org/10.5194/egusphere-2024-3919, 2025
Short summary
Short summary
This study presents in situ observations of cirrus occurrence from aircraft measurements in the extra-tropical transition layer (ExTL) using simultaneous measurements from two platforms. Lagrangian diagnostics based on high-resolution ICON simulations show long residence times of the cirrus in stratospheric air allowing to separate different diabatic processes during transit. The findings suggest that radiative diabatic cloud processes significantly impact the tropopause thermodynamic structure.
Cornelis Schwenk and Annette Miltenberger
Atmos. Chem. Phys., 24, 14073–14099, https://doi.org/10.5194/acp-24-14073-2024, https://doi.org/10.5194/acp-24-14073-2024, 2024
Short summary
Short summary
Warm conveyor belts (WCBs) transport moisture into the upper atmosphere, where it acts as a greenhouse gas. This transport is not well understood, and the role of rapidly rising air is unclear. We simulate a WCB and look at fast- and slow-rising air to see how moisture is (differently) transported. We find that for fast-ascending air more ice particles reach higher into the atmosphere and that frozen cloud particles are removed differently than during slow ascent, which has more water vapour.
Daniel Köhler, Philipp Reutter, and Peter Spichtinger
Atmos. Chem. Phys., 24, 10055–10072, https://doi.org/10.5194/acp-24-10055-2024, https://doi.org/10.5194/acp-24-10055-2024, 2024
Short summary
Short summary
In this work, the influence of humidity on the properties of the tropopause is studied. The tropopause is the interface between the troposphere and the stratosphere and represents a barrier for the transport of air masses between the troposphere and the stratosphere. We consider not only the tropopause itself, but also a layer around it called the tropopause inversion layer (TIL). It is shown that the moister the underlying atmosphere is, the more this layer acts as a barrier.
Edward Groot, Patrick Kuntze, Annette Miltenberger, and Holger Tost
Weather Clim. Dynam., 5, 779–803, https://doi.org/10.5194/wcd-5-779-2024, https://doi.org/10.5194/wcd-5-779-2024, 2024
Short summary
Short summary
Deep convective clouds (thunderstorms), which may cause severe weather, tend to coherently organise into structured cloud systems. Accurate representation of these systems in models is difficult due to their complex dynamics and, in numerical simulations, the dependence of their dynamics on resolution. Here, the effect of convective organisation and geometry on their outflow winds (altitudes of 7–14 km) is investigated. Representation of their dynamics and outflows improves at higher resolution.
Sarah Brüning, Stefan Niebler, and Holger Tost
Atmos. Meas. Tech., 17, 961–978, https://doi.org/10.5194/amt-17-961-2024, https://doi.org/10.5194/amt-17-961-2024, 2024
Short summary
Short summary
We apply the Res-UNet to derive a comprehensive 3D cloud tomography from 2D satellite data over heterogeneous landscapes. We combine observational data from passive and active remote sensing sensors by an automated matching algorithm. These data are fed into a neural network to predict cloud reflectivities on the whole satellite domain between 2.4 and 24 km height. With an average RMSE of 2.99 dBZ, we contribute to closing data gaps in the representation of clouds in observational data.
Annika Oertel, Annette K. Miltenberger, Christian M. Grams, and Corinna Hoose
Atmos. Chem. Phys., 23, 8553–8581, https://doi.org/10.5194/acp-23-8553-2023, https://doi.org/10.5194/acp-23-8553-2023, 2023
Short summary
Short summary
Warm conveyor belts (WCBs) are cloud- and precipitation-producing airstreams in extratropical cyclones that are important for the large-scale flow and cloud radiative forcing. We analyze cloud formation processes during WCB ascent in a two-moment microphysics scheme. Quantification of individual diabatic heating rates shows the importance of condensation, vapor deposition, rain evaporation, melting, and cloud-top radiative cooling for total heating and WCB-related potential vorticity structure.
Manuel Baumgartner, Christian Rolf, Jens-Uwe Grooß, Julia Schneider, Tobias Schorr, Ottmar Möhler, Peter Spichtinger, and Martina Krämer
Atmos. Chem. Phys., 22, 65–91, https://doi.org/10.5194/acp-22-65-2022, https://doi.org/10.5194/acp-22-65-2022, 2022
Short summary
Short summary
An important mechanism for the appearance of ice particles in the upper troposphere at low temperatures is homogeneous nucleation. This process is commonly described by the
Koop line, predicting the humidity at freezing. However, laboratory measurements suggest that the freezing humidities are above the Koop line, motivating the present study to investigate the influence of different physical parameterizations on the homogeneous freezing with the help of a detailed numerical model.
Rachel E. Hawker, Annette K. Miltenberger, Jill S. Johnson, Jonathan M. Wilkinson, Adrian A. Hill, Ben J. Shipway, Paul R. Field, Benjamin J. Murray, and Ken S. Carslaw
Atmos. Chem. Phys., 21, 17315–17343, https://doi.org/10.5194/acp-21-17315-2021, https://doi.org/10.5194/acp-21-17315-2021, 2021
Short summary
Short summary
We find that ice-nucleating particles (INPs), aerosols that can initiate the freezing of cloud droplets, cause substantial changes to the properties of radiatively important convectively generated anvil cirrus. The number concentration of INPs had a large effect on ice crystal number concentration while the INP temperature dependence controlled ice crystal size and cloud fraction. The results indicate information on INP number and source is necessary for the representation of cloud glaciation.
Ralf Weigel, Christoph Mahnke, Manuel Baumgartner, Martina Krämer, Peter Spichtinger, Nicole Spelten, Armin Afchine, Christian Rolf, Silvia Viciani, Francesco D'Amato, Holger Tost, and Stephan Borrmann
Atmos. Chem. Phys., 21, 13455–13481, https://doi.org/10.5194/acp-21-13455-2021, https://doi.org/10.5194/acp-21-13455-2021, 2021
Short summary
Short summary
In July and August 2017, the StratoClim mission took place in Nepal with eight flights of the M-55 Geophysica at up to 20 km in the Asian monsoon anticyclone. New particle formation (NPF) next to cloud ice was detected in situ by abundant nucleation-mode aerosols (> 6 nm) along with ice particles (> 3 µm). NPF was observed mainly below the tropopause, down to 15 % being non-volatile residues. Observed intra-cloud NPF indicates its importance for the composition in the tropical tropopause layer.
Rachel E. Hawker, Annette K. Miltenberger, Jonathan M. Wilkinson, Adrian A. Hill, Ben J. Shipway, Zhiqiang Cui, Richard J. Cotton, Ken S. Carslaw, Paul R. Field, and Benjamin J. Murray
Atmos. Chem. Phys., 21, 5439–5461, https://doi.org/10.5194/acp-21-5439-2021, https://doi.org/10.5194/acp-21-5439-2021, 2021
Short summary
Short summary
The impact of aerosols on clouds is a large source of uncertainty for future climate projections. Our results show that the radiative properties of a complex convective cloud field in the Saharan outflow region are sensitive to the temperature dependence of ice-nucleating particle concentrations. This means that differences in the aerosol source or composition, for the same aerosol size distribution, can cause differences in the outgoing radiation from regions dominated by tropical convection.
Annette K. Miltenberger and Paul R. Field
Atmos. Chem. Phys., 21, 3627–3642, https://doi.org/10.5194/acp-21-3627-2021, https://doi.org/10.5194/acp-21-3627-2021, 2021
Short summary
Short summary
The formation of ice in clouds is an important processes in mixed-phase and ice-phase clouds. However, the representation of ice formation in numerical models is highly uncertain. In the last decade, several new parameterizations for heterogeneous freezing have been proposed. Here, we investigate the impact of the parameterization choice on the representation of the convective cloud field and compare the impact to that of initial condition uncertainty.
Manuel Baumgartner, Ralf Weigel, Allan H. Harvey, Felix Plöger, Ulrich Achatz, and Peter Spichtinger
Atmos. Chem. Phys., 20, 15585–15616, https://doi.org/10.5194/acp-20-15585-2020, https://doi.org/10.5194/acp-20-15585-2020, 2020
Short summary
Short summary
The potential temperature is routinely used in atmospheric science. We review its derivation and suggest a new potential temperature, based on a temperature-dependent parameterization of the dry air's specific heat capacity. Moreover, we compare the new potential temperature to the common one and discuss the differences which become more important at higher altitudes. Finally, we indicate some consequences of using the new potential temperature in typical applications.
Martina Krämer, Christian Rolf, Nicole Spelten, Armin Afchine, David Fahey, Eric Jensen, Sergey Khaykin, Thomas Kuhn, Paul Lawson, Alexey Lykov, Laura L. Pan, Martin Riese, Andrew Rollins, Fred Stroh, Troy Thornberry, Veronika Wolf, Sarah Woods, Peter Spichtinger, Johannes Quaas, and Odran Sourdeval
Atmos. Chem. Phys., 20, 12569–12608, https://doi.org/10.5194/acp-20-12569-2020, https://doi.org/10.5194/acp-20-12569-2020, 2020
Short summary
Short summary
To improve the representations of cirrus clouds in climate predictions, extended knowledge of their properties and geographical distribution is required. This study presents extensive airborne in situ and satellite remote sensing climatologies of cirrus and humidity, which serve as a guide to cirrus clouds. Further, exemplary radiative characteristics of cirrus types and also in situ observations of tropical tropopause layer cirrus and humidity in the Asian monsoon anticyclone are shown.
Cited articles
Acuna, D., Kar, A., and Fidler, S.: Devil is in the Edges: Learning Semantic Boundaries from Noisy Annotations, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 11075–11083, 2019. a
Berry, G., Reeder, M. J., and Jakob, C.: A global climatology of atmospheric fronts, Geophys. Res. Lett., 38, L04809, https://doi.org/10.1029/2010GL046451, 2011. a, b
Bitsa, E., Flocas, H., Kouroutzoglou, J., Hatzaki, M., Rudeva, I., and Simmonds, I.: Development of a Front Identification Scheme for Compiling a Cold Front Climatology of the Mediterranean, Climate, 7, 130, https://doi.org/10.3390/cli7110130, 2019. a
Bochenek, B., Ustrnul, Z., Wypych, A., and Kubacka, D.: Machine Learning-Based Front Detection in Central Europe, Atmosphere, 12, 1312, https://doi.org/10.3390/atmos12101312, 2021. a
Brooks, H. E.: Tornado-Warning Performance in the Past and Future: A Perspective from Signal Detection Theory, B. Am. Meteorol. Soc., 85, 837–844, https://doi.org/10.1175/BAMS-85-6-837, 2004. a
Catto, J. and Dowdy, A.: Understanding compound hazards from a weather system perspective, Weather and Climate Extremes, 32, 100 313, https://doi.org/10.1016/j.wace.2021.100313, 2021. a
Catto, J., Madonna, E., Joos, H., Rudeva, I., and Simmonds, I.: Global Relationship between Fronts and Warm Conveyor Belts and the Impact on Extreme Precipitation, J. Climate, 28, 8411–8429, https://doi.org/10.1175/JCLI-D-15-0171.1, 2015. a
ECMWF: L137 model level definitions, available at: https://www.ecmwf.int/en/forecasts/documentation-and-support/137-model-levels, last access: 18 May 2021. a
Foss, M., Chou, S. C., and Seluchi, M. E.: Interaction of cold fronts with the Brazilian Plateau: a climatological analysis, Int. J. Climatol., 37, 3644–3659, https://doi.org/10.1002/joc.4945, 2017. a
Hersbach, H., Bell, B., Berrisford, P., Hirahara, S., Horányi, A., Muñoz-Sabater, J., Nicolas, J., Peubey, C., Radu, R., Schepers, D., Simmons, A., Soci, C., Abdalla, S., Abellan, X., Balsamo, G., Bechtold, P., Biavati, G., Bidlot, J., Bonavita, M., De Chiara, G., Dahlgren, P., Dee, D., Diamantakis, M., Dragani, R., Flemming, J., Forbes, R., Fuentes, M., Geer, A., Haimberger, L., Healy, S., Hogan, R. J., Hólm, E., Janisková, M., Keeley, S., Laloyaux, P., Lopez, P., Lupu, C., Radnoti, G., de Rosnay, P., Rozum, I., Vamborg, F., Villaume, S., and Thépaut, J.-N.: The ERA5 global reanalysis, Q. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803, 2020. a, b
Hewson, T. D.: Objective fronts, Meteorol. Appl., 5, 37–65, https://doi.org/10.1017/S1350482798000553, 1998. a, b
Hewson, T. D. and Titley, H. A.: Objective identification, typing and tracking of the complete life-cycles of cyclonic features at high spatial resolution, Meteorol. Appl., 17, 355–381, https://doi.org/10.1002/met.204, 2010. a
Hope, P., Keay, K., Pook, M., Catto, J., Simmonds, I., Mills, G., McIntosh, P., Risbey, J., and Berry, G.: A Comparison of Automated Methods of Front Recognition for Climate Studies: A Case Study in Southwest Western Australia, Mon. Weather Rev., 142, 343–363, https://doi.org/10.1175/MWR-D-12-00252.1, 2014. a
Hu, Y., Deng, Y., Lin, Y., Zhou, Z., Cui, C., and Dong, X.: Dynamics of the spatiotemporal morphology of Mei-yu fronts: an initial survey, Clim. Dynam., 56, 2715–2728, https://doi.org/10.1007/s00382-020-05619-2, 2021. a
Jakob, W., Rhinelander, J., and Moldovan, D.: pybind11 – Seamless operability between C++11 and Python, GitHub [code], https://github.com/pybind/pybind11 (last access: 17 January 2022), 2017. a
Jenkner, J., Sprenger, M., Schwenk, I., Schwierz, C., Dierer, S., and Leuenberger, D.: Detection and climatology of fronts in a high-resolution model reanalysis over the Alps, Meteorol. Appl., 17, 1–18, https://doi.org/10.1002/met.142, 2010. a, b, c, d
Martius, O., Pfahl, S., and Chavalier, C.: A global quantification of compound precipitation and wind extremes, Geophys. Res. Lett., 43, 7709–7717, https://doi.org/10.1002/2016GL070017, 2016. a
Matsuoka, D., Sugimoto, S., Nakagawa, Y., Kawahara, S., Araki, F., Onoue, Y., Iiyama, M., and Koyamada, K.: Automatic Detection of Stationary Fronts around Japan Using a Deep Convolutional Neural Network, SOLA, 15, 154–159, https://doi.org/10.2151/sola.2019-028, 2019. a, b, c, d
May, R. M., Arms, S. C., Marsh, P., Bruning, E., Leeman, J. R., Goebbert, K., Thielen, J. E., Bruick, Z. S., and Camron, M. D.: MetPy: A Python Package for Meteorological Data, UCAR [code], https://doi.org/10.5065/D6WW7G29, 2021. a
Mesinger, F., DiMego, G., Kalnay, E., Mitchell, K., Shafran, P. C., Ebisuzaki, W., Jović, D., Woollen, J., Rogers, E., Berbery, E. H., Ek, M. B., Fan, Y., Grumbine, R., Higgins, W., Li, H., Lin, Y., Manikin, G., Parrish, D., and Shi, W.: North American Regional Reanalysis, B. Am. Meteorol. Soc., 87, 343–360, https://doi.org/10.1175/BAMS-87-3-343, 2006. a
National Weather Service: National Weather Service Coded Surface Bulletins, 2003-, Zenodo [data set], https://doi.org/10.5281/zenodo.2642801, 2019. a, b
Niebler, S.: Front polylines extracted from DWD Maps, Zenodo [data set], https://doi.org/10.5281/zenodo.5785816, 2021a. a
Niebler, S.: FrontDetection, Zenodo [code], https://doi.org/10.5281/zenodo.5783934, 2021b. a
Niebler, S.: Detected Fronts January 2016, TIB-AV Portal, https://doi.org/10.5446/54716, 2021c. a, b, c
Parfitt, R., Czaja, A., and Seo, H.: A simple diagnostic for the detection of atmospheric fronts, Geophys. Res. Lett., 44, 4351–4358, https://doi.org/10.1002/2017GL073662, 2017. a
Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., Killeen, T., Lin, Z., Gimelshein, N., Antiga, L., Desmaison, A., Kopf, A., Yang, E., DeVito, Z., Raison, M., Tejani, A., Chilamkurthy, S., Steiner, B., Fang, L., Bai, J., and Chintala, S.: PyTorch: An Imperative Style, High-Performance Deep Learning Library, in: Advances in Neural Information Processing Systems 32, edited by Wallach, H., Larochelle, H., Beygelzimer, A., d'Alché-Buc, F., Fox, E., and Garnett, R., 8024–8035, Curran Associates, Inc. [code], http://papers.neurips.cc/paper/9015-pytorch-an-imperative-style-high-performance-deep-learning-library.pdf (last access: 17 January 2022), 2019. a
Pfahl, S. and Wernli, H.: Quantifying the Relevance of Cyclones for Precipitation Extremes, J. Climate, 25, 6770–6780, https://doi.org/10.1175/JCLI-D-11-00705.1, 2012. a
Renard, R. J. and Clarke, L. C.: Experiments In Numerical Objective Frontal Analysis, Mon. Weather Rev., 93, 541–556, 1965. a
Ribeiro, B. Z., Seluchi, M. E., and Chou, S. C.: Synoptic climatology of warm fronts in Southeastern South America, Int. J. Climatol., 36, 644–655, https://doi.org/10.1002/joc.4373, 2016. a
Sanders, F.: A proposed method of surface map analysis, Mon. Weather Rev., 127, 945–955, https://doi.org/10.1175/1520-0493(1999)127<0945:APMOSM>2.0.CO;2, 1999. a
Schemm, S., Sprenger, M., and Wernli, H.: When During Their Life Cycle Are Extratropical Cyclones Attended By Fronts?, B. Am. Meteorol. Soc., 99, 149–166, https://doi.org/10.1175/BAMS-D-16-0261.1, 2018. a, b
Schulzweida, U.: CDO User Guide, Zenodo [code], https://doi.org/10.5281/zenodo.3539275, 2019. a
Seabold, S. and Perktold, J.: statsmodels: Econometric and statistical modeling with python, in: 9th Python in Science Conference, Austin, TX, 61 pp., 2010. a
Shakina, N. P.: Identification of zones of atmospheric fronts as a problem of postprocessing the results of numerical prediction, Russ. Meteorol. Hydro+, 39, 1–10, https://doi.org/10.3103/S1068373914010014, 2014. a, b
Shelhamer, E., Long, J., and Darrell, T.: Fully Convolutional Networks for Semantic Segmentation, IEEE T. Pattern Anal., 39, 640–651, https://doi.org/10.1109/TPAMI.2016.2572683, 2017. a
Simmonds, I., Keay, K., and Bye, J. A. T.: Identification and Climatology of Southern Hemisphere Mobile Fronts in a Modern Reanalysis, J. Climate, 25, 1945–1962, https://doi.org/10.1175/JCLI-D-11-00100.1, 2012. a
Thomas, C. M. and Schultz, D. M.: Global Climatologies of Fronts, Airmass Boundaries, and Airstream Boundaries: Why the Definition of “Front” Matters, Mon. Weather Rev., 147, 691–717, https://doi.org/10.1175/MWR-D-18-0289.1, 2019a. a
Thomas, C. M. and Schultz, D. M.: What are the Best Thermodynamic Quantity and Function to Define a Front in Gridded Model Output?, B. Am. Meteorol. Soc., 100, 873–896, https://doi.org/10.1175/BAMS-D-18-0137.1, 2019b. a
Uccellini, L., Corfidi, S., Junker, N., Kocin, P., and Olson, D.: Report On The Surface-Analysis Workshop Held At The National-Meteorological-Center – 25–28 March 1991, B. Am. Meteorol. Soc., 73, 459–472, 1992. a
Short summary
We use machine learning to create a network that detects and classifies four types of synoptic-scale weather fronts from ERA5 atmospheric reanalysis data. We present an application of our method, showing its use case in a scientific context. Additionally, our results show that multiple sources of training data are necessary to perform well on different regions, implying differences within those regions. Qualitative evaluation shows that the results are physically plausible.
We use machine learning to create a network that detects and classifies four types of...
