Articles | Volume 15, issue 3
https://doi.org/10.5194/amt-15-797-2022
© 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/amt-15-797-2022
© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Research article
| 
14 Feb 2022
Research article |
| 14 Feb 2022
| 14 Feb 2022
Applying self-supervised learning for semantic cloud segmentation of all-sky images
Institute of Solar Research, German Aerospace Center (DLR), 04001 Almeria, Spain
Bijan Nouri
CORRESPONDING AUTHOR
Institute of Solar Research, German Aerospace Center (DLR), 04001 Almeria, Spain
Stefan Wilbert
Institute of Solar Research, German Aerospace Center (DLR), 04001 Almeria, Spain
Niklas Blum
Institute of Solar Research, German Aerospace Center (DLR), 04001 Almeria, Spain
Rudolph Triebel
Institute of Robotics and Mechatronics, German Aerospace Center (DLR), 82234 Oberpfaffenhofen-Weßling, Germany
Chair of Computer Vision and Artificial Intelligence, Technical University of Munich, 85748 Garching, Germany
Marcel Hasenbalg
Pascal Kuhn
EnBW Energie Baden-Württemberg AG, 76131 Karlsruhe, Germany
Luis F. Zarzalejo
Renewable Energy Division, CIEMAT Energy Department, 28040 Madrid, Spain
Robert Pitz-Paal
Institute of Solar Research, German Aerospace Center (DLR), 51147 Cologne, Germany
Related authors
No articles found.
Niklas Benedikt Blum, Bijan Nouri, Stefan Wilbert, Thomas Schmidt, Ontje Lünsdorf, Jonas Stührenberg, Detlev Heinemann, Andreas Kazantzidis, and Robert Pitz-Paal
Atmos. Meas. Tech., 14, 5199–5224, https://doi.org/10.5194/amt-14-5199-2021, https://doi.org/10.5194/amt-14-5199-2021, 2021
Short summary
Short summary
Cloud base height (CBH) is important, e.g., to forecast solar irradiance and, with it, photovoltaic production. All-sky imagers (ASIs), cameras monitoring the sky above their point of installation, can provide such forecasts and also measure CBH. We present a network of ASIs to measure CBH. The network provides numerous readings of CBH simultaneously. We combine these with a statistical procedure. Validation attests to significantly higher accuracy of the combination compared to two ASIs alone.
Pascal Kuhn, Stefan Wilbert, Christoph Prahl, Dominik Garsche, David Schüler, Thomas Haase, Lourdes Ramirez, Luis Zarzalejo, Angela Meyer, Philippe Blanc, and Robert Pitz-Paal
Adv. Sci. Res., 15, 11–14, https://doi.org/10.5194/asr-15-11-2018, https://doi.org/10.5194/asr-15-11-2018, 2018
Short summary
Short summary
Downward-facing shadow cameras might play a major role in future energy meteorology. Shadow cameras image shadows directly on the ground from an elevated position. They are used to validate other systems (e.g. all-sky imager based nowcasting systems, cloud speed sensors or satellite forecasts) and can potentially provide short term forecasts for solar power plants. Such forecasts are needed for electricity grids with high penetrations of renewable energy and solar power plants.
Wilko Jessen, Stefan Wilbert, Bijan Nouri, Norbert Geuder, and Holger Fritz
Atmos. Meas. Tech., 9, 1601–1612, https://doi.org/10.5194/amt-9-1601-2016, https://doi.org/10.5194/amt-9-1601-2016, 2016
Short summary
Short summary
This paper covers procedures and requirements of two calibration methods and examines the necessary duration of acquisition of test measurements.
Site-specific seasonal changes of environmental conditions cause small but noticeable fluctuation of calibration results. Calibration results within certain periods show a higher likelihood of deviation. These effects can partially be attenuated by including more measurements from outside these periods.
N. Hanrieder, S. Wilbert, R. Pitz-Paal, C. Emde, J. Gasteiger, B. Mayer, and J. Polo
Atmos. Meas. Tech., 8, 3467–3480, https://doi.org/10.5194/amt-8-3467-2015, https://doi.org/10.5194/amt-8-3467-2015, 2015
B. Reinhardt, R. Buras, L. Bugliaro, S. Wilbert, and B. Mayer
Atmos. Meas. Tech., 7, 823–838, https://doi.org/10.5194/amt-7-823-2014, https://doi.org/10.5194/amt-7-823-2014, 2014
Related subject area
Subject: Clouds | Technique: Remote Sensing | Topic: Data Processing and Information Retrieval
ampycloud: an open-source algorithm to determine cloud base heights and sky coverage fractions from ceilometer data
Simulation and detection efficiency analysis for measurements of polar mesospheric clouds using a spaceborne wide-field-of-view ultraviolet imager
The Chalmers Cloud Ice Climatology: retrieval implementation and validation
The algorithm of microphysical-parameter profiles of aerosol and small cloud droplets based on the dual-wavelength lidar data
Bayesian cloud-top phase determination for Meteosat Second Generation
Lidar–radar synergistic method to retrieve ice, supercooled water and mixed-phase cloud properties
Deriving cloud droplet number concentration from surface-based remote sensors with an emphasis on lidar measurements
A random forest algorithm for the prediction of cloud liquid water content from combined CloudSat–CALIPSO observations
Identification of ice-over-water multilayer clouds using multispectral satellite data in an artificial neural network
A new approach to crystal habit retrieval from far-infrared spectral radiance measurements
Severe hail detection with C-band dual-polarisation radars using convolutional neural networks
Multiple-scattering effects on single-wavelength lidar sounding of multi-layered clouds
Contrail altitude estimation using GOES-16 ABI data and deep learning
The Ice Cloud Imager: retrieval of frozen water column properties
An advanced spatial co-registration of cloud properties for the atmospheric Sentinel missions: Application to TROPOMI
A cloud-by-cloud approach for studying aerosol–cloud interaction in satellite observations
Infrared Radiometric Image Classification and Segmentation of Cloud Structure Using Deep-learning Framework for Ground-based Infrared Thermal Camera Observations
Information Content of Brightness Temperature Differences of Spaceborne Imagers with respect to Cloud Phase
Geometrical and optical properties of cirrus clouds in Barcelona, Spain: analysis with the two-way transmittance method of 4 years of lidar measurements
Determination of the vertical distribution of in-cloud particle shape using SLDR-mode 35 GHz scanning cloud radar
Artificial intelligence (AI)-derived 3D cloud tomography from geostationary 2D satellite data
Marine cloud base height retrieval from MODIS cloud properties using machine learning
The EarthCARE mission: science data processing chain overview
Cloud optical and physical properties retrieval from EarthCARE multi-spectral imager: the M-COP products
Cloud top heights and aerosol columnar properties from combined EarthCARE lidar and imager observations: the AM-CTH and AM-ACD products
Raman lidar-derived optical and microphysical properties of ice crystals within thin Arctic clouds during PARCS campaign
Evaluation of four ground-based retrievals of cloud droplet number concentration in marine stratocumulus with aircraft in situ measurements
Deep convective cloud system size and structure across the global tropics and subtropics
A neural-network-based method for generating synthetic 1.6 µm near-infrared satellite images
Numerical model generation of test frames for pre-launch studies of EarthCARE's retrieval algorithms and data management system
Segmentation of polarimetric radar imagery using statistical texture
Retrieval of surface solar irradiance from satellite imagery using machine learning: pitfalls and perspectives
Retrieving 3D distributions of atmospheric particles using Atmospheric Tomography with 3D Radiative Transfer – Part 2: Local optimization
Particle inertial effects on radar Doppler spectra simulation
Detection of aerosol and cloud features for the EarthCARE atmospheric lidar (ATLID): the ATLID FeatureMask (A-FM) product
A unified synergistic retrieval of clouds, aerosols, and precipitation from EarthCARE: the ACM-CAP product
Supercooled liquid water cloud classification using lidar backscatter peak properties
Incorporating EarthCARE observations into a multi-lidar cloud climate record: the ATLID (Atmospheric Lidar) cloud climate product
Introduction to EarthCARE synthetic data using a global storm-resolving simulation
Validation of a camera-based intra-hour irradiance nowcasting model using synthetic cloud data
Liquid cloud optical property retrieval and associated uncertainties using multi-angular and bispectral measurements of the airborne radiometer OSIRIS
Global evaluation of Doppler velocity errors of EarthCARE cloud-profiling radar using a global storm-resolving simulation
Cloud and precipitation microphysical retrievals from the EarthCARE Cloud Profiling Radar: the C-CLD product
Cloud mask algorithm from the EarthCARE Multi-Spectral Imager: the M-CM products
Across-track extension of retrieved cloud and aerosol properties for the EarthCARE mission: the ACMB-3D product
Insights into 3D cloud radiative transfer effects for the Orbiting Carbon Observatory
Evaluation of polarimetric ice microphysical retrievals with OLYMPEX campaign data
Retrieving 3D distributions of atmospheric particles using Atmospheric Tomography with 3D Radiative Transfer – Part 1: Model description and Jacobian calculation
Simulation and sensitivity analysis for cloud and precipitation measurements via spaceborne millimeter-wave radar
The Virga-Sniffer – a new tool to identify precipitation evaporation using ground-based remote-sensing observations
Frédéric P. A. Vogt, Loris Foresti, Daniel Regenass, Sophie Réthoré, Néstor Tarin Burriel, Mervyn Bibby, Przemysław Juda, Simone Balmelli, Tobias Hanselmann, Pieter du Preez, and Dirk Furrer
Atmos. Meas. Tech., 17, 4891–4914, https://doi.org/10.5194/amt-17-4891-2024, https://doi.org/10.5194/amt-17-4891-2024, 2024
Short summary
Short summary
ampycloud is a new algorithm developed at MeteoSwiss to characterize the height and sky coverage fraction of cloud layers above aerodromes via ceilometer data. This algorithm was devised as part of a larger effort to fully automate the creation of meteorological aerodrome reports (METARs) at Swiss civil airports. The ampycloud algorithm is implemented as a Python package that is made publicly available to the community under the 3-Clause BSD license.
Ke Ren, Haiyang Gao, Shuqi Niu, Shaoyang Sun, Leilei Kou, Yanqing Xie, Liguo Zhang, and Lingbing Bu
Atmos. Meas. Tech., 17, 4825–4842, https://doi.org/10.5194/amt-17-4825-2024, https://doi.org/10.5194/amt-17-4825-2024, 2024
Short summary
Short summary
Ultraviolet imaging technology has significantly advanced the research and development of polar mesospheric clouds (PMCs). In this study, we proposed the wide-field-of-view ultraviolet imager (WFUI) and built a forward model to evaluate the detection capability and efficiency. The results demonstrate that the WFUI performs well in PMC detection and has high detection efficiency. The relationship between ice water content and detection efficiency follows an exponential function distribution.
Adrià Amell, Simon Pfreundschuh, and Patrick Eriksson
Atmos. Meas. Tech., 17, 4337–4368, https://doi.org/10.5194/amt-17-4337-2024, https://doi.org/10.5194/amt-17-4337-2024, 2024
Short summary
Short summary
The representation of clouds in numerical weather and climate models remains a major challenge that is difficult to address because of the limitations of currently available data records of cloud properties. In this work, we address this issue by using machine learning to extract novel information on ice clouds from a long record of satellite observations. Through extensive validation, we show that this novel approach provides surprisingly accurate estimates of clouds and their properties.
Huige Di, Xinhong Wang, Ning Chen, Jing Guo, Wenhui Xin, Shichun Li, Yan Guo, Qing Yan, Yufeng Wang, and Dengxin Hua
Atmos. Meas. Tech., 17, 4183–4196, https://doi.org/10.5194/amt-17-4183-2024, https://doi.org/10.5194/amt-17-4183-2024, 2024
Short summary
Short summary
This study proposes an inversion method for atmospheric-aerosol or cloud microphysical parameters based on dual-wavelength lidar data. It is suitable for the inversion of uniformly mixed and single-property aerosol layers or small cloud droplets. For aerosol particles, the inversion range that this algorithm can achieve is 0.3–1.7 μm. For cloud droplets, it is 1.0–10 μm. This algorithm can quickly obtain the microphysical parameters of atmospheric particles and has better robustness.
Johanna Mayer, Luca Bugliaro, Bernhard Mayer, Dennis Piontek, and Christiane Voigt
Atmos. Meas. Tech., 17, 4015–4039, https://doi.org/10.5194/amt-17-4015-2024, https://doi.org/10.5194/amt-17-4015-2024, 2024
Short summary
Short summary
ProPS (PRObabilistic cloud top Phase retrieval for SEVIRI) is a method to detect clouds and their thermodynamic phase with a geostationary satellite, distinguishing between clear sky and ice, mixed-phase, supercooled and warm liquid clouds. It uses a Bayesian approach based on the lidar–radar product DARDAR. The method allows studying cloud phases, especially mixed-phase and supercooled clouds, rarely observed from geostationary satellites. This can be used for comparison with climate models.
Clémantyne Aubry, Julien Delanoë, Silke Groß, Florian Ewald, Frédéric Tridon, Olivier Jourdan, and Guillaume Mioche
Atmos. Meas. Tech., 17, 3863–3881, https://doi.org/10.5194/amt-17-3863-2024, https://doi.org/10.5194/amt-17-3863-2024, 2024
Short summary
Short summary
Radar–lidar synergy is used to retrieve ice, supercooled water and mixed-phase cloud properties, making the most of the radar sensitivity to ice crystals and the lidar sensitivity to supercooled droplets. A first analysis of the output of the algorithm run on the satellite data is compared with in situ data during an airborne Arctic field campaign, giving a mean percent error of 49 % for liquid water content and 75 % for ice water content.
Gerald G. Mace
Atmos. Meas. Tech., 17, 3679–3695, https://doi.org/10.5194/amt-17-3679-2024, https://doi.org/10.5194/amt-17-3679-2024, 2024
Short summary
Short summary
The number of cloud droplets per unit volume, Nd, in a cloud is important for understanding aerosol–cloud interaction. In this study, we develop techniques to derive cloud droplet number concentration from lidar measurements combined with other remote sensing measurements such as cloud radar and microwave radiometers. We show that deriving Nd is very uncertain, although a synergistic algorithm seems to produce useful characterizations of Nd and effective particle size.
Richard M. Schulte, Matthew D. Lebsock, John M. Haynes, and Yongxiang Hu
Atmos. Meas. Tech., 17, 3583–3596, https://doi.org/10.5194/amt-17-3583-2024, https://doi.org/10.5194/amt-17-3583-2024, 2024
Short summary
Short summary
This paper describes a method to improve the detection of liquid clouds that are easily missed by the CloudSat satellite radar. To address this, we use machine learning techniques to estimate cloud properties (optical depth and droplet size) based on other satellite measurements. The results are compared with data from the MODIS instrument on the Aqua satellite, showing good correlations.
Sunny Sun-Mack, Patrick Minnis, Yan Chen, Gang Hong, and William L. Smith Jr.
Atmos. Meas. Tech., 17, 3323–3346, https://doi.org/10.5194/amt-17-3323-2024, https://doi.org/10.5194/amt-17-3323-2024, 2024
Short summary
Short summary
Multilayer clouds (MCs) affect the radiation budget differently than single-layer clouds (SCs) and need to be identified in satellite images. A neural network was trained to identify MCs by matching imagery with lidar/radar data. This method correctly identifies ~87 % SCs and MCs with a net accuracy gain of 7.5 % over snow-free surfaces. It is more accurate than most available methods and constitutes a first step in providing a reasonable 3-D characterization of the cloudy atmosphere.
Gianluca Di Natale, Marco Ridolfi, and Luca Palchetti
Atmos. Meas. Tech., 17, 3171–3186, https://doi.org/10.5194/amt-17-3171-2024, https://doi.org/10.5194/amt-17-3171-2024, 2024
Short summary
Short summary
This work aims to define a new approach to retrieve the distribution of the main ice crystal shapes occurring inside ice and cirrus clouds from infrared spectral measurements. The capability of retrieving these shapes of the ice crystals from satellites will allow us to extend the currently available climatologies to be used as physical constraints in general circulation models. This could could allow us to improve their accuracy and prediction performance.
Vincent Forcadell, Clotilde Augros, Olivier Caumont, Kévin Dedieu, Maxandre Ouradou, Cloe David, Jordi Figueras i Ventura, Olivier Laurantin, and Hassan Al-Sakka
EGUsphere, https://doi.org/10.5194/egusphere-2024-1336, https://doi.org/10.5194/egusphere-2024-1336, 2024
Short summary
Short summary
This study demonstrates the potential for enhancing severe hail detection through the application of convolutional neural networks (CNNs) to dual-polarization radar data. It is shown that current methods can be calibrated to significantly enhance their performance for severe hail detection. This study establishes the foundation for the solution of a more complex problem: the estimation of the maximum size of hailstones on the ground using deep learning applied to radar data.
Valery Shcherbakov, Frédéric Szczap, Guillaume Mioche, and Céline Cornet
Atmos. Meas. Tech., 17, 3011–3028, https://doi.org/10.5194/amt-17-3011-2024, https://doi.org/10.5194/amt-17-3011-2024, 2024
Short summary
Short summary
We performed Monte Carlo simulations of single-wavelength lidar signals from multi-layered clouds with special attention focused on the multiple-scattering (MS) effect in regions of the cloud-free molecular atmosphere. The MS effect on lidar signals always decreases with the increasing distance from the cloud far edge. The decrease is the direct consequence of the fact that the forward peak of particle phase functions is much larger than the receiver field of view.
Vincent R. Meijer, Sebastian D. Eastham, Ian A. Waitz, and Steven R.H. Barrett
EGUsphere, https://doi.org/10.5194/egusphere-2024-961, https://doi.org/10.5194/egusphere-2024-961, 2024
Short summary
Short summary
Aviation's climate impact is partly due to contrails: the clouds that form behind aircraft and which can linger for hours under certain atmospheric conditions. Accurately forecasting these conditions could allow aircraft to avoid forming these contrails and thus reduce their environmental footprint. Our research uses deep learning to identify three-dimensional contrail locations in two-dimensional satellite imagery, which can be used to assess and improve these forecasts.
Eleanor May, Bengt Rydberg, Inderpreet Kaur, Vinia Mattioli, Hanna Hallborn, and Patrick Eriksson
EGUsphere, https://doi.org/10.5194/egusphere-2024-829, https://doi.org/10.5194/egusphere-2024-829, 2024
Short summary
Short summary
The upcoming Ice Cloud Imager (ICI) mission is set to improve measurements of atmospheric ice through passive microwave and sub-millimetre wave observations. In this study, we perform detailed simulations of ICI observations. Machine learning is used to characterise the atmospheric ice present for a given simulated observation. This study acts as a final pre-launch assessment of ICI's capability to measure atmospheric ice, providing valuable information to climate and weather applications.
Athina Argyrouli, Diego Loyola, Fabian Romahn, Ronny Lutz, Víctor Molina García, Pascal Hedelt, Klaus-Peter Heue, and Richard Siddans
Atmos. Meas. Tech. Discuss., https://doi.org/10.5194/amt-2024-28, https://doi.org/10.5194/amt-2024-28, 2024
Revised manuscript accepted for AMT
Short summary
Short summary
This manuscript describes a new treatment of the spatial mis-registration of cloud properties for Sentinel-5 Precursor, when the footprints of different spectral bands are not perfectly aligned. The methodology exploits synergies between spectrometers and imagers, like TROPOMI and VIIRS. The largest improvements have been identified for heterogeneous scenes at cloud edges. This approach is generic and can also be applied to future Sentinel-4 and Sentinel-5 instruments.
Fani Alexandri, Felix Müller, Goutam Choudhury, Peggy Achtert, Torsten Seelig, and Matthias Tesche
Atmos. Meas. Tech., 17, 1739–1757, https://doi.org/10.5194/amt-17-1739-2024, https://doi.org/10.5194/amt-17-1739-2024, 2024
Short summary
Short summary
We present a novel method for studying aerosol–cloud interactions. It combines cloud-relevant aerosol concentrations from polar-orbiting lidar observations with the development of individual clouds from geostationary observations. Application to 1 year of data gives first results on the impact of aerosols on the concentration and size of cloud droplets and on cloud phase in the regime of heterogeneous ice formation. The method could enable the systematic investigation of warm and cold clouds.
Kélian Sommer, Wassim Kabalan, and Romain Brunet
EGUsphere, https://doi.org/10.5194/egusphere-2024-101, https://doi.org/10.5194/egusphere-2024-101, 2024
Short summary
Short summary
Our research introduces a novel deep-learning approach for classifying and segmenting ground-based infrared thermal images, a crucial step in cloud monitoring. Tests on self-captured data showcase its excellent accuracy in distinguishing image types and in structure segmentation. With potential applications in astronomical observations, our work pioneers a robust solution for ground-based sky quality assessment, promising advancements in the photometric observations experiments.
Johanna Mayer, Bernhard Mayer, Luca Bugliaro, Ralf Meerkötter, and Christiane Voigt
EGUsphere, https://doi.org/10.5194/egusphere-2024-540, https://doi.org/10.5194/egusphere-2024-540, 2024
Short summary
Short summary
This study uses radiative transfer calculations to characterize the relation of two satellite channel combinations (namely infrared window brightness temperature differences (BTDs) of the SEVIRI imager) to the thermodynamic cloud phase. A sensitivity analysis reveals the complex interplay of cloud parameters and their contribution to the observed phase dependence of BTDs. This knowledge helps to design optimal cloud phase retrievals and to understand their potential and limitations.
Cristina Gil-Díaz, Michäel Sicard, Adolfo Comerón, Daniel Camilo Fortunato dos Santos Oliveira, Constantino Muñoz-Porcar, Alejandro Rodríguez-Gómez, Jasper R. Lewis, Ellsworth J. Welton, and Simone Lolli
Atmos. Meas. Tech., 17, 1197–1216, https://doi.org/10.5194/amt-17-1197-2024, https://doi.org/10.5194/amt-17-1197-2024, 2024
Short summary
Short summary
In this paper, a statistical study of cirrus geometrical and optical properties based on 4 years of continuous ground-based lidar measurements with the Barcelona (Spain) Micro Pulse Lidar (MPL) is analysed. The cloud optical depth, effective column lidar ratio and linear cloud depolarisation ratio have been calculated by a new approach to the two-way transmittance method, which is valid for both ground-based and spaceborne lidar systems. Their associated errors are also provided.
Audrey Teisseire, Patric Seifert, Alexander Myagkov, Johannes Bühl, and Martin Radenz
Atmos. Meas. Tech., 17, 999–1016, https://doi.org/10.5194/amt-17-999-2024, https://doi.org/10.5194/amt-17-999-2024, 2024
Short summary
Short summary
The vertical distribution of particle shape (VDPS) method, introduced in this study, aids in characterizing the density-weighted shape of cloud particles from scanning slanted linear depolarization ratio (SLDR)-mode cloud radar observations. The VDPS approach represents a new, versatile way to study microphysical processes by combining a spheroidal scattering model with real measurements of SLDR.
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.
Julien Lenhardt, Johannes Quaas, and Dino Sejdinovic
EGUsphere, https://doi.org/10.5194/egusphere-2024-327, https://doi.org/10.5194/egusphere-2024-327, 2024
Short summary
Short summary
Clouds play a key role in the regulation of the Earth's climate. Aspects like the height of their base are of essential interest, but remain difficult to derive from satellite data. In this study, we combine observations from the surface and satellite retrievals of cloud properties to build a robust and accurate method to retrieve the cloud base height.
Michael Eisinger, Fabien Marnas, Kotska Wallace, Takuji Kubota, Nobuhiro Tomiyama, Yuichi Ohno, Toshiyuki Tanaka, Eichi Tomita, Tobias Wehr, and Dirk Bernaerts
Atmos. Meas. Tech., 17, 839–862, https://doi.org/10.5194/amt-17-839-2024, https://doi.org/10.5194/amt-17-839-2024, 2024
Short summary
Short summary
The Earth Cloud Aerosol and Radiation Explorer (EarthCARE) is an ESA–JAXA satellite mission to be launched in 2024. We presented an overview of the EarthCARE processors' development, with processors developed by teams in Europe, Japan, and Canada. EarthCARE will allow scientists to evaluate the representation of cloud, aerosol, precipitation, and radiative flux in weather forecast and climate models, with the objective to better understand cloud processes and improve weather and climate models.
Anja Hünerbein, Sebastian Bley, Hartwig Deneke, Jan Fokke Meirink, Gerd-Jan van Zadelhoff, and Andi Walther
Atmos. Meas. Tech., 17, 261–276, https://doi.org/10.5194/amt-17-261-2024, https://doi.org/10.5194/amt-17-261-2024, 2024
Short summary
Short summary
The ESA cloud, aerosol and radiation mission EarthCARE will provide active profiling and passive imaging measurements from a single satellite platform. The passive multi-spectral imager (MSI) will add information in the across-track direction. We present the cloud optical and physical properties algorithm, which combines the visible to infrared MSI channels to determine the cloud top pressure, optical thickness, particle size and water path.
Moritz Haarig, Anja Hünerbein, Ulla Wandinger, Nicole Docter, Sebastian Bley, David Donovan, and Gerd-Jan van Zadelhoff
Atmos. Meas. Tech., 16, 5953–5975, https://doi.org/10.5194/amt-16-5953-2023, https://doi.org/10.5194/amt-16-5953-2023, 2023
Short summary
Short summary
The atmospheric lidar (ATLID) and Multi-Spectral Imager (MSI) will be carried by the EarthCARE satellite. The synergistic ATLID–MSI Column Products (AM-COL) algorithm described in the paper combines the strengths of ATLID in vertically resolved profiles of aerosol and clouds (e.g., cloud top height) with the strengths of MSI in observing the complete scene beside the satellite track and in extending the lidar information to the swath. The algorithm is validated against simulated test scenes.
Patrick Chazette and Jean-Christophe Raut
Atmos. Meas. Tech., 16, 5847–5861, https://doi.org/10.5194/amt-16-5847-2023, https://doi.org/10.5194/amt-16-5847-2023, 2023
Short summary
Short summary
The vertical profiles of the effective radii of ice crystals and ice water content in Arctic semi-transparent stratiform clouds were assessed using quantitative ground-based lidar measurements. The field campaign was part of the Pollution in the ARCtic System (PARCS) project which took place from 13 to 26 May 2016 in Hammerfest (70° 39′ 48″ N, 23° 41′ 00″ E). We show that under certain cloud conditions, lidar measurement combined with a dedicated algorithmic approach is an efficient tool.
Damao Zhang, Andrew M. Vogelmann, Fan Yang, Edward Luke, Pavlos Kollias, Zhien Wang, Peng Wu, William I. Gustafson Jr., Fan Mei, Susanne Glienke, Jason Tomlinson, and Neel Desai
Atmos. Meas. Tech., 16, 5827–5846, https://doi.org/10.5194/amt-16-5827-2023, https://doi.org/10.5194/amt-16-5827-2023, 2023
Short summary
Short summary
Cloud droplet number concentration can be retrieved from remote sensing measurements. Aircraft measurements are used to validate four ground-based retrievals of cloud droplet number concentration. We demonstrate that retrieved cloud droplet number concentrations align well with aircraft measurements for overcast clouds, but they may substantially differ for broken clouds. The ensemble of various retrievals can help quantify retrieval uncertainties and identify reliable retrieval scenarios.
Eric M. Wilcox, Tianle Yuan, and Hua Song
Atmos. Meas. Tech., 16, 5387–5401, https://doi.org/10.5194/amt-16-5387-2023, https://doi.org/10.5194/amt-16-5387-2023, 2023
Short summary
Short summary
A new database is constructed from over 20 years of satellite records that comprises millions of deep convective clouds and spans the global tropics and subtropics. The database is a collection of clouds ranging from isolated cells to giant cloud systems. The cloud database provides a means of empirically studying the factors that determine the spatial structure and coverage of convective cloud systems, which are strongly related to the overall radiative forcing by cloud systems.
Florian Baur, Leonhard Scheck, Christina Stumpf, Christina Köpken-Watts, and Roland Potthast
Atmos. Meas. Tech., 16, 5305–5326, https://doi.org/10.5194/amt-16-5305-2023, https://doi.org/10.5194/amt-16-5305-2023, 2023
Short summary
Short summary
Near-infrared satellite images have information on clouds that is complementary to what is available from the visible and infrared parts of the spectrum. Using this information for data assimilation and model evaluation requires a fast, accurate forward operator to compute synthetic images from numerical weather prediction model output. We discuss a novel, neural-network-based approach for the 1.6 µm near-infrared channel that is suitable for this purpose and also works for other solar channels.
Zhipeng Qu, David P. Donovan, Howard W. Barker, Jason N. S. Cole, Mark W. Shephard, and Vincent Huijnen
Atmos. Meas. Tech., 16, 4927–4946, https://doi.org/10.5194/amt-16-4927-2023, https://doi.org/10.5194/amt-16-4927-2023, 2023
Short summary
Short summary
The EarthCARE satellite mission Level 2 algorithm development requires realistic 3D cloud and aerosol scenes along the satellite orbits. One of the best ways to produce these scenes is to use a high-resolution numerical weather prediction model to simulate atmospheric conditions at 250 m horizontal resolution. This paper describes the production and validation of three EarthCARE test scenes.
Adrien Guyot, Jordan P. Brook, Alain Protat, Kathryn Turner, Joshua Soderholm, Nicholas F. McCarthy, and Hamish McGowan
Atmos. Meas. Tech., 16, 4571–4588, https://doi.org/10.5194/amt-16-4571-2023, https://doi.org/10.5194/amt-16-4571-2023, 2023
Short summary
Short summary
We propose a new method that should facilitate the use of weather radars to study wildfires. It is important to be able to identify the particles emitted by wildfires on radar, but it is difficult because there are many other echoes on radar like clear air, the ground, sea clutter, and precipitation. We came up with a two-step process to classify these echoes. Our method is accurate and can be used by fire departments in emergencies or by scientists for research.
Hadrien Verbois, Yves-Marie Saint-Drenan, Vadim Becquet, Benoit Gschwind, and Philippe Blanc
Atmos. Meas. Tech., 16, 4165–4181, https://doi.org/10.5194/amt-16-4165-2023, https://doi.org/10.5194/amt-16-4165-2023, 2023
Short summary
Short summary
Solar surface irradiance (SSI) estimations inferred from satellite images are essential to gain a comprehensive understanding of the solar resource, which is crucial in many fields. This study examines the recent data-driven methods for inferring SSI from satellite images and explores their strengths and weaknesses. The results suggest that while these methods show great promise, they sometimes dramatically underperform and should probably be used in conjunction with physical approaches.
Jesse Loveridge, Aviad Levis, Larry Di Girolamo, Vadim Holodovsky, Linda Forster, Anthony B. Davis, and Yoav Y. Schechner
Atmos. Meas. Tech., 16, 3931–3957, https://doi.org/10.5194/amt-16-3931-2023, https://doi.org/10.5194/amt-16-3931-2023, 2023
Short summary
Short summary
We test a new method for measuring the 3D spatial variations of water within clouds, using measurements of reflections of the Sun's light observed at multiple angles by satellites. This is a great improvement on older methods, which typically assume that clouds occur in a slab shape. Our study used computer modeling to show that our 3D method will work well in cumulus clouds, where older slab methods do not. Our method will inform us about these clouds and their role in our climate.
Zeen Zhu, Pavlos Kollias, and Fan Yang
Atmos. Meas. Tech., 16, 3727–3737, https://doi.org/10.5194/amt-16-3727-2023, https://doi.org/10.5194/amt-16-3727-2023, 2023
Short summary
Short summary
We show that large rain droplets, with large inertia, are unable to follow the rapid change of velocity field in a turbulent environment. A lack of consideration for this inertial effect leads to an artificial broadening of the Doppler spectrum from the conventional simulator. Based on the physics-based simulation, we propose a new approach to generate the radar Doppler spectra. This simulator provides a valuable tool to decode cloud microphysical and dynamical properties from radar observation.
Gerd-Jan van Zadelhoff, David P. Donovan, and Ping Wang
Atmos. Meas. Tech., 16, 3631–3651, https://doi.org/10.5194/amt-16-3631-2023, https://doi.org/10.5194/amt-16-3631-2023, 2023
Short summary
Short summary
The Earth Clouds, Aerosols and Radiation (EarthCARE) satellite mission features the UV lidar ATLID. The ATLID FeatureMask algorithm provides a high-resolution detection probability mask which is used to guide smoothing strategies within the ATLID profile retrieval algorithm, one step further in the EarthCARE level-2 processing chain, in which the microphysical retrievals and target classification are performed.
Shannon L. Mason, Robin J. Hogan, Alessio Bozzo, and Nicola L. Pounder
Atmos. Meas. Tech., 16, 3459–3486, https://doi.org/10.5194/amt-16-3459-2023, https://doi.org/10.5194/amt-16-3459-2023, 2023
Short summary
Short summary
We present a method for accurately estimating the contents and properties of clouds, snow, rain, and aerosols through the atmosphere, using the combined measurements of the radar, lidar, and radiometer instruments aboard the upcoming EarthCARE satellite, and evaluate the performance of the retrieval, using test scenes simulated from a numerical forecast model. When EarthCARE is in operation, these quantities and their estimated uncertainties will be distributed in a data product called ACM-CAP.
Luke Edgar Whitehead, Adrian James McDonald, and Adrien Guyot
EGUsphere, https://doi.org/10.5194/egusphere-2023-1085, https://doi.org/10.5194/egusphere-2023-1085, 2023
Short summary
Short summary
Supercooled liquid water cloud is important to represent in weather and climate models, particularly in the Southern Hemisphere. Previous work has developed a new machine learning method for measuring supercooled liquid water in Antarctic clouds using simple lidar observations. We evaluate this technique using a lidar dataset from Christchurch, New Zealand, and develop an updated algorithm for accurate supercooled liquid water detection at mid-latitudes.
Artem G. Feofilov, Hélène Chepfer, Vincent Noël, and Frederic Szczap
Atmos. Meas. Tech., 16, 3363–3390, https://doi.org/10.5194/amt-16-3363-2023, https://doi.org/10.5194/amt-16-3363-2023, 2023
Short summary
Short summary
The response of clouds to human-induced climate warming remains the largest source of uncertainty in model predictions of climate. We consider cloud retrievals from spaceborne observations, the existing CALIOP lidar and future ATLID lidar; show how they compare for the same scenes; and discuss the advantage of adding a new lidar for detecting cloud changes in the long run. We show that ATLID's advanced technology should allow for better detecting thinner clouds during daytime than before.
Woosub Roh, Masaki Satoh, Tempei Hashino, Shuhei Matsugishi, Tomoe Nasuno, and Takuji Kubota
Atmos. Meas. Tech., 16, 3331–3344, https://doi.org/10.5194/amt-16-3331-2023, https://doi.org/10.5194/amt-16-3331-2023, 2023
Short summary
Short summary
JAXA EarthCARE synthetic data (JAXA L1 data) were compiled using the global storm-resolving model (GSRM) NICAM (Nonhydrostatic ICosahedral
Atmospheric Model) simulation with 3.5 km horizontal resolution and the Joint-Simulator. JAXA L1 data are intended to support the development of JAXA retrieval algorithms for the EarthCARE sensor before launch of the satellite. The expected orbit of EarthCARE and horizontal sampling of each sensor were used to simulate the signals.
Philipp Gregor, Tobias Zinner, Fabian Jakub, and Bernhard Mayer
Atmos. Meas. Tech., 16, 3257–3271, https://doi.org/10.5194/amt-16-3257-2023, https://doi.org/10.5194/amt-16-3257-2023, 2023
Short summary
Short summary
This work introduces MACIN, a model for short-term forecasting of direct irradiance for solar energy applications. MACIN exploits cloud images of multiple cameras to predict irradiance. The model is applied to artificial images of clouds from a weather model. The artificial cloud data allow for a more in-depth evaluation and attribution of errors compared with real data. Good performance of derived cloud information and significant forecast improvements over a baseline forecast were found.
Christian Matar, Céline Cornet, Frédéric Parol, Laurent C.-Labonnote, Frédérique Auriol, and Marc Nicolas
Atmos. Meas. Tech., 16, 3221–3243, https://doi.org/10.5194/amt-16-3221-2023, https://doi.org/10.5194/amt-16-3221-2023, 2023
Short summary
Short summary
The optimal estimation formalism is applied to OSIRIS airborne high-resolution multi-angular measurements to retrieve COT and Reff. The corresponding uncertainties related to measurement errors, which are up to 6 and 12 %, the non-retrieved parameters, which are less than 0.5 %, and the cloud model assumptions show that the heterogeneous vertical profiles and the 3D radiative transfer effects lead to average uncertainties of 5 and 4 % for COT and 13 and 9 % for Reff.
Yuichiro Hagihara, Yuichi Ohno, Hiroaki Horie, Woosub Roh, Masaki Satoh, and Takuji Kubota
Atmos. Meas. Tech., 16, 3211–3219, https://doi.org/10.5194/amt-16-3211-2023, https://doi.org/10.5194/amt-16-3211-2023, 2023
Short summary
Short summary
The CPR on the EarthCARE satellite is the first satellite-borne Doppler radar. We evaluated the effectiveness of horizontal integration and the unfolding method for the reduction of the Doppler error (the standard deviation of the random error) in the CPR_ECO product. The error was higher in the tropics than in the other latitudes due to frequent rain echo occurrence and limitation of its unfolding correction. If we use low-mode operation (high PRF), the errors become small enough.
Kamil Mroz, Bernat Puidgomènech Treserras, Alessandro Battaglia, Pavlos Kollias, Aleksandra Tatarevic, and Frederic Tridon
Atmos. Meas. Tech., 16, 2865–2888, https://doi.org/10.5194/amt-16-2865-2023, https://doi.org/10.5194/amt-16-2865-2023, 2023
Short summary
Short summary
We present the theoretical basis of the algorithm that estimates the amount of water and size of particles in clouds and precipitation. The algorithm uses data collected by the Cloud Profiling Radar that was developed for the upcoming Earth Clouds, Aerosols and Radiation Explorer (EarthCARE) satellite mission. After the satellite launch, the vertical distribution of cloud and precipitation properties will be delivered as the C-CLD product.
Anja Hünerbein, Sebastian Bley, Stefan Horn, Hartwig Deneke, and Andi Walther
Atmos. Meas. Tech., 16, 2821–2836, https://doi.org/10.5194/amt-16-2821-2023, https://doi.org/10.5194/amt-16-2821-2023, 2023
Short summary
Short summary
The Multi-Spectral Imager (MSI) on board the EarthCARE satellite will provide the information needed for describing the cloud and aerosol properties in the cross-track direction, complementing the measurements from the Cloud Profiling Radar, Atmospheric Lidar and Broad-Band Radiometer. The accurate discrimination between clear and cloudy pixels is an essential first step. Therefore, the cloud mask algorithm provides a cloud flag, cloud phase and cloud type product for the MSI observations.
Zhipeng Qu, Howard W. Barker, Jason N. S. Cole, and Mark W. Shephard
Atmos. Meas. Tech., 16, 2319–2331, https://doi.org/10.5194/amt-16-2319-2023, https://doi.org/10.5194/amt-16-2319-2023, 2023
Short summary
Short summary
This paper describes EarthCARE’s L2 product ACM-3D. It includes the scene construction algorithm (SCA) used to produce the indexes for reconstructing 3D atmospheric scene based on satellite nadir retrievals. It also provides the information about the buffer zone sizes of 3D assessment domains and the ranking scores for selecting the best 3D assessment domains. These output variables are needed to run 3D radiative transfer models for the radiative closure assessment of EarthCARE’s L2 retrievals.
Steven T. Massie, Heather Cronk, Aronne Merrelli, Sebastian Schmidt, and Steffen Mauceri
Atmos. Meas. Tech., 16, 2145–2166, https://doi.org/10.5194/amt-16-2145-2023, https://doi.org/10.5194/amt-16-2145-2023, 2023
Short summary
Short summary
This paper provides insights into the effects of clouds on Orbiting Carbon Observatory (OCO-2) measurements of CO2. Calculations are carried out that indicate the extent to which this satellite experiment underestimates CO2, due to these cloud effects, as a function of the distance between the surface observation footprint and the nearest cloud. The paper discusses how to lessen the influence of these cloud effects.
Armin Blanke, Andrew J. Heymsfield, Manuel Moser, and Silke Trömel
Atmos. Meas. Tech., 16, 2089–2106, https://doi.org/10.5194/amt-16-2089-2023, https://doi.org/10.5194/amt-16-2089-2023, 2023
Short summary
Short summary
We present an evaluation of current retrieval techniques in the ice phase applied to polarimetric radar measurements with collocated in situ observations of aircraft conducted over the Olympic Mountains, Washington State, during winter 2015. Radar estimates of ice properties agreed most with aircraft observations in regions with pronounced radar signatures, but uncertainties were identified that indicate issues of some retrievals, particularly in warmer temperature regimes.
Jesse Loveridge, Aviad Levis, Larry Di Girolamo, Vadim Holodovsky, Linda Forster, Anthony B. Davis, and Yoav Y. Schechner
Atmos. Meas. Tech., 16, 1803–1847, https://doi.org/10.5194/amt-16-1803-2023, https://doi.org/10.5194/amt-16-1803-2023, 2023
Short summary
Short summary
We describe a new method for measuring the 3D spatial variations in water within clouds using the reflected light of the Sun viewed at multiple different angles by satellites. This is a great improvement over older methods, which typically assume that clouds occur in a slab shape. Our study used computer modeling to show that our 3D method will work well in cumulus clouds, where older slab methods do not. Our method will inform us about these clouds and their role in our climate.
Leilei Kou, Zhengjian Lin, Haiyang Gao, Shujun Liao, and Piman Ding
Atmos. Meas. Tech., 16, 1723–1744, https://doi.org/10.5194/amt-16-1723-2023, https://doi.org/10.5194/amt-16-1723-2023, 2023
Short summary
Short summary
Forward modeling of spaceborne millimeter-wave radar composed of eight submodules is presented. We quantify the uncertainties in radar reflectivity that may be caused by the physical model parameters via a sensitivity analysis. The simulations with improved and conventional settings are compared with CloudSat data, and the simulation results are evaluated and analyzed. The results are instructive to the optimization of forward modeling and microphysical parameter retrieval.
Heike Kalesse-Los, Anton Kötsche, Andreas Foth, Johannes Röttenbacher, Teresa Vogl, and Jonas Witthuhn
Atmos. Meas. Tech., 16, 1683–1704, https://doi.org/10.5194/amt-16-1683-2023, https://doi.org/10.5194/amt-16-1683-2023, 2023
Short summary
Short summary
The Virga-Sniffer, a new modular open-source Python package tool to characterize full precipitation evaporation (so-called virga) from ceilometer cloud base height and vertically pointing cloud radar reflectivity time–height fields, is described. Results of its first application to RV Meteor observations during the EUREC4A field experiment in January–February 2020 are shown. About half of all detected clouds with bases below the trade inversion height were found to produce virga.
Cited articles
Aitken, A., Ledig, C., Theis, L., Caballero, J., Wang, Z., and Shi, W.:
Checkerboard artifact free sub-pixel convolution: A note on sub-pixel
convolution, resize convolution and convolution resize, arXiv [preprint], arXiv:1707.02937, 2017. a
Blanc, P., Massip, P., Kazantzidis, A., Tzoumanikas, P., Kuhn, P., Wilbert, S.,
Schüler, D., and Prahl, C.: Short-term forecasting of high resolution
local DNI maps with multiple fish-eye cameras in stereoscopic mode, AIP
Conference Proceedings, 1850, 140004, https://doi.org/10.1063/1.4984512, 2017. a
Calbó, J., Long, C. N., González, J.-A., Augustine, J., and McComiskey,
A.: The thin border between cloud and aerosol: Sensitivity of several ground
based observation techniques, Atmos. Res., 196, 248–260,
https://doi.org/10.1016/j.atmosres.2017.06.010, 2017. a
Caron, M., Bojanowski, P., Joulin, A., and Douze, M.: Deep clustering for
unsupervised learning of visual features, in: Proceedings of the European
Conference on Computer Vision (ECCV), pp. 132–149,
https://doi.org/10.1007/978-3-030-01264-9_9, 2018. a, b, c, d
Chauvin, R., Nou, J., Thil, S., Traore, A., and Grieu, S.: Cloud detection
methodology based on a sky-imaging system, Energy Proced., 69, 1970–1980,
https://doi.org/10.1016/j.egypro.2015.03.198, 2015. a
Cheng, H.-Y. and Lin, C.-L.: Cloud detection in all-sky images via multi-scale neighborhood features and multiple supervised learning techniques, Atmos. Meas. Tech., 10, 199–208, https://doi.org/10.5194/amt-10-199-2017, 2017. a, b, c
Chow, C. W., Urquhart, B., Lave, M., Dominguez, A., Kleissl, J., Shields, J.,
and Washom, B.: Intra-hour forecasting with a total sky imager at the UC San
Diego solar energy testbed, Sol. Energy, 85, 2881–2893,
https://doi.org/10.1016/j.solener.2011.08.025, 2011. a
Dev, S., Lee, Y. H., and Winkler, S.: Systematic study of color spaces and
components for the segmentation of sky/cloud images, in: 2014 IEEE
International Conference on Image Processing (ICIP), pp. 5102–5106, IEEE,
https://doi.org/10.1109/ICIP.2014.7026033, 2014. a
Dev, S., Lee, Y. H., and Winkler, S.: Multi-level semantic labeling of
sky/cloud images, in: 2015 IEEE International Conference on Image Processing
(ICIP), pp. 636–640, IEEE, https://doi.org/10.1109/ICIP.2015.7350876, 2015. a
Dev, S., Wen, B., Lee, Y. H., and Winkler, S.: Machine learning techniques and
applications for ground-based image analysis, arXiv [preprint], arXiv:1606.02811, 2016. a
Dev, S., Manandhar, S., Lee, Y. H., and Winkler, S.: Multi-label cloud
segmentation using a deep network, in: 2019 USNC-URSI Radio Science Meeting
(Joint with AP-S Symposium), pp. 113–114, IEEE,
https://doi.org/10.1109/USNC-URSI.2019.8861850, 2019. a, b
Doersch, C., Gupta, A., and Efros, A. A.: Unsupervised visual representation
learning by context prediction, in: Proceedings of the IEEE International
Conference on Computer Vision, pp. 1422–1430, https://doi.org/10.1109/ICCV.2015.167,
2015. a
Ghonima, M. S., Urquhart, B., Chow, C. W., Shields, J. E., Cazorla, A., and Kleissl, J.: A method for cloud detection and opacity classification based on ground based sky imagery, Atmos. Meas. Tech., 5, 2881–2892, https://doi.org/10.5194/amt-5-2881-2012, 2012. a
Hasenbalg, M., Kuhn, P., Wilbert, S., Nouri, B., and Kazantzidis, A.:
Benchmarking of six cloud segmentation algorithms for ground-based all-sky
imagers, Sol. Energy, 201, 596–614, https://doi.org/10.1016/j.solener.2020.02.042,
2020. a, b, c, d
He, K., Zhang, X., Ren, S., and Sun, J.: Delving deep into rectifiers:
Surpassing human-level performance on imagenet classification, in:
Proceedings of the IEEE international conference on computer vision, pp.
1026–1034, https://doi.org/10.1109/ICCV.2015.123, 2015. a
He, K., Zhang, X., Ren, S., and Sun, J.: Deep residual learning for image
recognition, in: Proceedings of the IEEE conference on computer vision and
pattern recognition, pp. 770–778, https://doi.org/10.1109/CVPR.2016.90, 2016. a
Heinle, A., Macke, A., and Srivastav, A.: Automatic cloud classification of whole sky images, Atmos. Meas. Tech., 3, 557–567, https://doi.org/10.5194/amt-3-557-2010, 2010. a, b
Howard, J. and Ruder, S.: Universal language model fine-tuning for text
classification, arXiv [preprint], arXiv:1801.06146,
2018. a
Howard, J. et al.: fastai, GitHub [code], https://github.com/fastai/fastai1 (last access: 10 February 2022), 2018. a
Ineichen, P. and Perez, R.: A new airmass independent formulation for the Linke
turbidity coefficient, Sol. Energy, 73, 151–157,
https://doi.org/10.1016/S0038-092X(02)00045-2, 2002. a
Ioffe, S. and Szegedy, C.: Batch normalization: Accelerating deep network
training by reducing internal covariate shift, arXiv [preprint], arXiv:1502.03167,
2015. a
Jayadevan, V. T., Rodriguez, J. J., and Cronin, A. D.: A new contrast-enhancing
feature for cloud detection in ground-based sky images, J.
Atmos. Ocean. Tech., 32, 209–219,
https://doi.org/10.1175/JTECH-D-14-00053.1, 2015. a
Jing, L. and Tian, Y.: Self-supervised visual feature learning with deep neural
networks: A survey, IEEE Transactions on Pattern Analysis and Machine
Intelligence, 43, 4037–4058, https://doi.org/10.1109/TPAMI.2020.2992393, 2020. a
Johnson, J., Alahi, A., and Fei-Fei, L.: Perceptual losses for real-time style
transfer and super-resolution, in: European conference on computer vision,
pp. 694–711, Springer, https://doi.org/10.1007/978-3-319-46475-6_43, 2016. a, b
Kazantzidis, A., Tzoumanikas, P., Bais, A. F., Fotopoulos, S., and Economou,
G.: Cloud detection and classification with the use of whole-sky ground-based
images, Atmos. Res., 113, 80–88,
https://doi.org/10.1016/j.atmosres.2012.05.005, 2012. a
Kingma, D. P. and Ba, J.: Adam: A method for stochastic optimization, arXiv [preprint], arXiv:1412.6980
2014. a
Koren, I., Remer, L. A., Kaufman, Y. J., Rudich, Y., and Martins, J. V.: On the
twilight zone between clouds and aerosols, Geophys. Res. Lett., 34, L08805,
https://doi.org/10.1029/2007GL029253, 2007. a
Kuhn, P., Nouri, B., Wilbert, S., Prahl, C., Kozonek, N., Schmidt, T., Yasser,
Z., Ramirez, L., Zarzalejo, L., Meyer, A., et al.: Validation of an all-sky
imager–based nowcasting system for industrial PV plants,
Progress in Photovoltaics: Research and Applications, 26, 608–621,
https://doi.org/10.1002/pip.2968, 2018. a, b, c
Lee, H.-Y., Huang, J.-B., Singh, M., and Yang, M.-H.: Unsupervised
representation learning by sorting sequences, in: Proceedings of the IEEE
International Conference on Computer Vision, pp. 667–676,
https://doi.org/10.1109/ICCV.2017.79, 2017. a
Li, Q., Lu, W., and Yang, J.: A hybrid thresholding algorithm for cloud
detection on ground-based color images, J. Atmos. Ocean.
Tech., 28, 1286–1296, https://doi.org/10.1175/JTECH-D-11-00009.1, 2011. a, b
Liu, S., Zhang, L., Zhang, Z., Wang, C., and Xiao, B.: Automatic cloud
detection for all-sky images using superpixel segmentation, IEEE Geosci.
Remote, 12, 354–358, https://doi.org/10.1109/LGRS.2014.2341291,
2014. a
Liu, S., Zhang, Z., Xiao, B., and Cao, X.: Ground-based cloud detection using
automatic graph cut, IEEE Geosci. Remote, 12,
1342–1346, https://doi.org/10.1109/LGRS.2015.2399857, 2015. a
Long, C., Slater, D., and Tooman, T. P.: Total sky imager model 880 status and
testing results, Pacific Northwest National Laboratory Richland, Wash, USA,
https://doi.org/10.2172/1020735, 2001. a
Long, C. N., Sabburg, J. M., Calbó, J., and Pages, D.: Retrieving cloud
characteristics from ground-based daytime color all-sky images, J.
Atmos. Ocean. Tech., 23, 633–652, https://doi.org/10.1175/JTECH1875.1,
2006. a
Noroozi, M. and Favaro, P.: Unsupervised learning of visual representations by
solving jigsaw puzzles, in: European Conference on Computer Vision, pp.
69–84, Springer, https://doi.org/10.1007/978-3-319-46466-4_5, 2016. a
Nouri, B., Wilbert, S., Kuhn, P., Hanrieder, N., Schroedter-Homscheidt, M.,
Kazantzidis, A., Zarzalejo, L., Blanc, P., Kumar, S., Goswami, N., Shankar, R., Affolter, R., and Pitz-Paal, R.:
Real-time uncertainty specification of all sky imager derived irradiance
nowcasts, Remote Sens., 11, 1059, https://doi.org/10.3390/rs11091059,
2019a. a
Nouri, B., Wilbert, S., Segura, L., Kuhn, P., Hanrieder, N., Kazantzidis, A.,
Schmidt, T., Zarzalejo, L., Blanc, P., and Pitz-Paal, R.: Determination of
cloud transmittance for all sky imager based solar nowcasting, Sol. Energy,
181, 251–263, https://doi.org/10.1016/j.solener.2019.02.004, 2019b. 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., pp. 8024–8035, Curran Associates, Inc., http://papers.neurips.cc/paper/9015-pytorch-an-imperative-style-high-performance-deep-learning-library.pdf (last access: 10 February 2022), 2019. a
Pathak, D., Krahenbuhl, P., Donahue, J., Darrell, T., and Efros, A. A.: Context
encoders: Feature learning by inpainting, in: Proceedings of the IEEE
conference on computer vision and pattern recognition, pp. 2536–2544,
https://doi.org/10.1109/CVPR.2016.278, 2016. a, b
Perez, R., David, M., Hoff, T. E., Jamaly, M., Kivalov, S., Kleissl, J.,
Lauret, P., and Perez, M.: Spatial and temporal variability of solar
energy, Foundations and Trends in Renewable Energy, 1, 1–44, https://doi.org/10.1561/2700000006, 2016. a
Radford, A., Metz, L., and Chintala, S.: Unsupervised representation learning
with deep convolutional generative adversarial networks, arXiv [preprint], arXiv:1511.06434, 2015. a
Ronneberger, O., Fischer, P., and Brox, T.: U-net: Convolutional networks for
biomedical image segmentation, in: International Conference on Medical image
computing and computer-assisted intervention, Springer, pp. 234–241,
https://doi.org/10.1007/978-3-319-24574-4_28, 2015. a
Rossow, W. and Zhang, Y.-C.: Calculation of surface and top of atmosphere
radiative fluxes from physical quantities based on ISCCP data sets: 2.
Validation and first results, J. Geophys. Res.-Atmos.,
100, 1167–1197, https://doi.org/10.1029/94JD02746, 1995. a
Rossow, W. B. and Schiffer, R. A.: Advances in understanding clouds from ISCCP,
B. Am. Meteorol. Soc., 80, 2261–2288,
https://doi.org/10.1175/1520-0477(1999)080<2261:AIUCFI>2.0.CO;2, 1999. a
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
Shi, C., Wang, Y., Wang, C., and Xiao, B.: Ground-based cloud detection using
graph model built upon superpixels, IEEE Geosci. Remote, 14, 719–723, https://doi.org/10.1109/LGRS.2017.2676007, 2017. a
Shi, W., Caballero, J., Huszár, F., Totz, J., Aitken, A. P., Bishop, R.,
Rueckert, D., and Wang, Z.: Real-time single image and video super-resolution
using an efficient sub-pixel convolutional neural network, in: Proceedings of
the IEEE conference on computer vision and pattern recognition, pp.
1874–1883, https://doi.org/10.1109/CVPR.2016.207, 2016. a
Shields, J., Karr, M., Tooman, T., Sowle, D., and Moore, S.: The whole sky
imager–a year of progress, in: Eighth Atmospheric Radiation Measurement
(ARM) Science Team Meeting, Tucson, Arizona, pp. 23–27, Citeseer, 1998. a
Smith, L. N.: Cyclical learning rates for training neural networks, in: 2017
IEEE Winter Conference on Applications of Computer Vision (WACV), pp.
464–472, IEEE, https://doi.org/10.1109/WACV.2017.58, 2017.
a
Smith, L. N.: A disciplined approach to neural network hyper-parameters: Part
1–learning rate, batch size, momentum, and weight decay, arXiv [preprint], arXiv:1803.09820, 2018. a
Song, Q., Cui, Z., and Liu, P.: An Efficient Solution for Semantic Segmentation
of Three Ground-based Cloud Datasets, Earth Space Sci., 7,
e2019EA001040, https://doi.org/10.1029/2019EA001040, 2020. a
Souza-Echer, M. P., Pereira, E. B., Bins, L., and Andrade, M.: A simple method
for the assessment of the cloud cover state in high-latitude regions by a
ground-based digital camera, J. Atmos. Ocean. Tech.,
23, 437–447, https://doi.org/10.1175/JTECH1833.1, 2006. a
Stubenrauch, C., Rossow, W., Kinne, S., Ackerman, S., Cesana, G., Chepfer, H.,
Di Girolamo, L., Getzewich, B., Guignard, A., Heidinger, A., et al.:
Assessment of global cloud datasets from satellites: Project and database
initiated by the GEWEX radiation panel, B. Am.
Meteorol. Soc., 94, 1031–1049, https://doi.org/10.1175/BAMS-D-12-00117.1,
2013. a
Taravat, A., Del Frate, F., Cornaro, C., and Vergari, S.: Neural networks and
support vector machine algorithms for automatic cloud classification of
whole-sky ground-based images, IEEE Geosci. Remote,
12, 666–670, https://doi.org/10.1109/LGRS.2014.2356616, 2014. a
West, S. R., Rowe, D., Sayeef, S., and Berry, A.: Short-term irradiance
forecasting using skycams: Motivation and development, Sol. Energy, 110,
188–207, https://doi.org/10.1016/j.solener.2014.08.038, 2014. a
Widener, K. and Long, C.: All sky imager, uS Patent App. 10/377,042, 2004. a
Wilbert, S., Nouri, B., Prahl, C., Garcia, G., Ramirez, L., Zarzalejo, L.,
Valenzuela, R., Ferrera, F., Kozonek, N., and Liria, J.: Application of whole
sky imagers for data selection for radiometer calibration, EU PVSEC 2016
Proceedings, pp. 1493–1498, https://doi.org/10.4229/EUPVSEC20162016-5AO.8.6, 2016. a
Xie, W., Liu, D., Yang, M., Chen, S., Wang, B., Wang, Z., Xia, Y., Liu, Y., Wang, Y., and Zhang, C.: SegCloud: a novel cloud image segmentation model using a deep convolutional neural network for ground-based all-sky-view camera observation, Atmos. Meas. Tech., 13, 1953–1961, https://doi.org/10.5194/amt-13-1953-2020, 2020. a, b
Ye, L., Cao, Z., and Xiao, Y.: DeepCloud: Ground-based cloud image
categorization using deep convolutional features,
IEEE T. Geosci. Remote, 55, 5729–5740,
https://doi.org/10.1109/TGRS.2017.2712809, 2017. a
Ye, L., Cao, Z., Xiao, Y., and Yang, Z.: Supervised Fine-Grained Cloud
Detection and Recognition in Whole-Sky Images, IEEE T. Geosci. Remote, 57, 7972–7985,
https://doi.org/10.1109/TGRS.2019.2917612, 2019. a, b, c
Zhang, J., Liu, P., Zhang, F., and Song, Q.: CloudNet: Ground-based cloud
classification with deep convolutional neural network, Geophys. Res.
Lett., 45, 8665–8672, https://doi.org/10.1029/2018GL077787, 2018. a
Zhuo, W., Cao, Z., and Xiao, Y.: Cloud classification of ground-based images
using texture–structure features, J. Atmos. Ocean.
Tech., 31, 79–92, https://doi.org/10.1175/JTECH-D-13-00048.1, 2014. a
Short summary
This work presents a new approach to exploit unlabeled image data from ground-based sky observations to train neural networks. We show that our model can detect cloud classes within images more accurately than models trained with conventional methods using small, labeled datasets only. Novel machine learning techniques as applied in this work enable training with much larger datasets, leading to improved accuracy in cloud detection and less need for manual image labeling.
This work presents a new approach to exploit unlabeled image data from ground-based sky...
