Articles | Volume 14, issue 1
https://doi.org/10.5194/amt-14-737-2021
© Author(s) 2021. 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-14-737-2021
© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Research article
| 
29 Jan 2021
Research article |
| 29 Jan 2021
| 29 Jan 2021
Improving cloud type classification of ground-based images using region covariance descriptors
Yuzhu Tang
School of Computer Science and Engineering, Nanjing University of
Science and Technology, Nanjing, 210094, China
State Key Laboratory of Satellite Ocean Environment Dynamics, Second Institute of Oceanography, Hangzhou, 310012, China
Pinglv Yang
College of Meteorology and Oceanology, National University of Defense
Technology, Nanjing, 211101, China
Zeming Zhou
CORRESPONDING AUTHOR
College of Meteorology and Oceanology, National University of Defense
Technology, Nanjing, 211101, China
Delu Pan
Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou), Guangzhou, 511458, China
State Key Laboratory of Satellite Ocean Environment Dynamics, Second Institute of Oceanography, Hangzhou, 310012, China
Jianyu Chen
CORRESPONDING AUTHOR
State Key Laboratory of Satellite Ocean Environment Dynamics, Second Institute of Oceanography, Hangzhou, 310012, China
Xiaofeng Zhao
College of Meteorology and Oceanology, National University of Defense
Technology, Nanjing, 211101, China
Related authors
No articles found.
Qixiang Luo, Yong Meng, Lei Liu, Xiaofeng Zhao, and Zeming Zhou
Atmos. Meas. Tech., 11, 5351–5361, https://doi.org/10.5194/amt-11-5351-2018, https://doi.org/10.5194/amt-11-5351-2018, 2018
Short summary
Short summary
In this paper, a novel cloud classification method is proposed to group images into five cloud types based on manifold and texture features. The proposed method is comprised of three stages: data pre-processing, feature extraction and classification. Compared to the recent cloud type recognition methods, the experimental results illustrate that the proposed method acquires a higher recognition rate with an increase of 2%–10% on the ground-based infrared datasets.
Related subject area
Subject: Clouds | Technique: Remote Sensing | Topic: Data Processing and Information Retrieval
High-spatial-resolution retrieval of cloud droplet size distribution from polarized observations of the cloudbow
Evaluation of the spectral misalignment on the Earth Clouds, Aerosols and Radiation Explorer/multi-spectral imager cloud product
Retrieval of terahertz ice cloud properties from airborne measurements based on the irregularly shaped Voronoi ice scattering models
Latent heating profiles from GOES-16 and its impacts on precipitation forecasts
A CO2-independent cloud mask from Infrared Atmospheric Sounding Interferometer (IASI) radiances for climate applications
Retrieval of ice water path from the Microwave Humidity Sounder (MWHS) aboard FengYun-3B (FY-3B) satellite polarimetric measurements based on a deep neural network
Intercomparison of Sentinel-5P TROPOMI cloud products for tropospheric trace gas retrievals
Near Global Distributions of Overshooting Tops Derived from Terra and Aqua MODIS Observations
The CHROMA cloud top pressure retrieval algorithm for the Plankton, Aerosol, Cloud, ocean Ecosytem (PACE) satellite mission
Improved spectral processing for a multi-mode pulse compression Ka–Ku-band cloud radar system
Uncertainty-bounded estimates of ash cloud properties using the ORAC algorithm: application to the 2019 Raikoke eruption
Ice water path retrievals from Meteosat-9 using quantile regression neural networks
Retrieving 3D distributions of atmospheric particles using Atmospheric Tomography with 3D Radiative Transfer – Part 1: Model description and Jacobian calculation
An optimal estimation algorithm for the retrieval of fog and low cloud thermodynamic and micro-physical properties
Identifying cloud droplets beyond lidar attenuation from vertically pointing cloud radar observations using artificial neural networks
Segmentation-based multi-pixel cloud optical thickness retrieval using a convolutional neural network
Top-of-the-atmosphere reflected shortwave radiative fluxes from GOES-R
Optimizing radar scan strategies for tracking isolated deep convection using observing system simulation experiments
A kriging-based analysis of cloud liquid water content using CloudSat data
High-resolution satellite-based cloud detection for the analysis of land surface effects on boundary layer clouds
Retrievals of ice microphysical properties using dual-wavelength polarimetric radar observations during stratiform precipitation events
The surface longwave cloud radiative effect derived from space lidar observations
Cloud phase and macrophysical properties over the Southern Ocean during the MARCUS field campaign
Detection of supercooled liquid water containing clouds with ceilometers: development and evaluation of deterministic and data-driven retrievals
An all-sky camera image classification method using cloud cover features
Determination of atmospheric column condensate using active and passive remote sensing technology
Phase correlation on the edge for estimating cloud motion
Improving discrimination between clouds and optically thick aerosol plumes in geostationary satellite data
Towards the use of conservative thermodynamic variables in data assimilation: a case study using ground-based microwave radiometer measurements
Empirical model of multiple-scattering effect on single-wavelength lidar data of aerosols and clouds
Analytic characterization of random errors in spectral dual-polarized cloud radar observations
Assessing synergistic radar and radiometer capability in retrieving ice cloud microphysics based on hybrid Bayesian algorithms
A Semi-Lagrangian Method for Detecting and Tracking Deep Convective Clouds in Geostationary Satellite Observations
Applying self-supervised learning for semantic cloud segmentation of all-sky images
Coincident in situ and triple-frequency radar airborne observations in the Arctic
Analysis of improvements in MOPITT observational coverage over Canada
Climatology of estimated LWC and scaling factor for warm clouds using radar – microwave radiometer synergy
Using artificial neural networks to predict riming from Doppler cloud radar observations
Evaluating cloud liquid detection against Cloudnet using cloud radar Doppler spectra in a pre-trained artificial neural network
Cloud optical properties retrieval and associated uncertainties using multi-angular and multi-spectral measurements of the airborne radiometer OSIRIS
PARAFOG v2.0: a near-real-time decision tool to support nowcasting fog formation events at local scales
Inpainting radar missing data regions with deep learning
Improved cloud detection for the Aura Microwave Limb Sounder (MLS): training an artificial neural network on colocated MLS and Aqua MODIS data
Triple-frequency radar retrieval of microphysical properties of snow
Retrieving microphysical properties of concurrent pristine ice and snow using polarimetric radar observations
Comparison of mid-latitude single- and mixed-phase cloud optical depth from co-located infrared spectrometer and backscatter lidar measurements
Physical characteristics of frozen hydrometeors inferred with parameter estimation
Cloud height measurement by a network of all-sky imagers
Increasing the spatial resolution of cloud property retrievals from Meteosat SEVIRI by use of its high-resolution visible channel: implementation and examples
Why we need radar, lidar, and solar radiance observations to constrain ice cloud microphysics
Veronika Pörtge, Tobias Kölling, Anna Weber, Lea Volkmer, Claudia Emde, Tobias Zinner, Linda Forster, and Bernhard Mayer
Atmos. Meas. Tech., 16, 645–667, https://doi.org/10.5194/amt-16-645-2023, https://doi.org/10.5194/amt-16-645-2023, 2023
Short summary
Short summary
In this work, we analyze polarized cloudbow observations by the airborne camera system specMACS to retrieve the cloud droplet size distribution defined by the effective radius (reff) and the effective variance (veff). Two case studies of trade-wind cumulus clouds observed during the EUREC4A field campaign are presented. The results are combined into maps of reff and veff with a very high spatial resolution (100 m × 100 m) that allow new insights into cloud microphysics.
Minrui Wang, Takashi Y. Nakajima, Woosub Roh, Masaki Satoh, Kentaroh Suzuki, Takuji Kubota, and Mayumi Yoshida
Atmos. Meas. Tech., 16, 603–623, https://doi.org/10.5194/amt-16-603-2023, https://doi.org/10.5194/amt-16-603-2023, 2023
Short summary
Short summary
SMILE (a spectral misalignment in which a shift in the center wavelength appears as a distortion in the spectral image) was detected during our recent work. To evaluate how it affects the cloud retrieval products, we did a simulation of EarthCARE-MSI forward radiation, evaluating the error in simulated scenes from a global cloud system-resolving model and a satellite simulator. Our results indicated that the error from SMILE was generally small and negligible for oceanic scenes.
Ming Li, Husi Letu, Hiroshi Ishimoto, Shulei Li, Lei Liu, Takashi Y. Nakajima, Dabin Ji, Huazhe Shang, and Chong Shi
Atmos. Meas. Tech., 16, 331–353, https://doi.org/10.5194/amt-16-331-2023, https://doi.org/10.5194/amt-16-331-2023, 2023
Short summary
Short summary
Influenced by the representativeness of ice crystal scattering models, the existing terahertz ice cloud remote sensing inversion algorithms still have significant uncertainties. We developed an ice cloud remote sensing retrieval algorithm of the ice water path and particle size from aircraft-based terahertz radiation measurements based on the Voronoi model. Validation revealed that the Voronoi model performs better than the sphere and hexagonal column models.
Yoonjin Lee, Christian D. Kummerow, and Milija Zupanski
Atmos. Meas. Tech., 15, 7119–7136, https://doi.org/10.5194/amt-15-7119-2022, https://doi.org/10.5194/amt-15-7119-2022, 2022
Short summary
Short summary
Vertical profiles of latent heating are derived from GOES-16 to be used in convective initialization. They are compared with other latent heating products derived from NEXRAD and GPM satellites, and the results show that their values are very similar to the radar-derived products. Finally, using latent heating derived from GOES-16 for convective initialization shows improvements in precipitation forecasts, which are comparable to the results using latent heating derived from NEXRAD.
Simon Whitburn, Lieven Clarisse, Marc Crapeau, Thomas August, Tim Hultberg, Pierre François Coheur, and Cathy Clerbaux
Atmos. Meas. Tech., 15, 6653–6668, https://doi.org/10.5194/amt-15-6653-2022, https://doi.org/10.5194/amt-15-6653-2022, 2022
Short summary
Short summary
With more than 15 years of measurements, the IASI radiance dataset is becoming a reference climate data record. Its exploitation for satellite applications requires an accurate and unbiased detection of cloud scenes. Here, we present a new cloud detection algorithm for IASI that is both sensitive and consistent over time. It is based on the use of a neural network, relying on IASI radiance information only and taking as a reference the last version of the operational IASI L2 cloud product.
Wenyu Wang, Zhenzhan Wang, Qiurui He, and Lanjie Zhang
Atmos. Meas. Tech., 15, 6489–6506, https://doi.org/10.5194/amt-15-6489-2022, https://doi.org/10.5194/amt-15-6489-2022, 2022
Short summary
Short summary
This paper uses a neural network approach to retrieve the ice water path from FY-3B/MWHS polarimetric measurements, focusing on its unique 150 GHz quasi-polarized channels. The Level 2 product of CloudSat is used as the reference value for the neural network. The results show that the polarization information is helpful for the retrieval in scenes with thicker cloud ice, and the 150 GHz channels give a significant improvement compared to using only 183 GHz channels.
Miriam Latsch, Andreas Richter, Henk Eskes, Maarten Sneep, Ping Wang, Pepijn Veefkind, Ronny Lutz, Diego Loyola, Athina Argyrouli, Pieter Valks, Thomas Wagner, Holger Sihler, Michel van Roozendael, Nicolas Theys, Huan Yu, Richard Siddans, and John P. Burrows
Atmos. Meas. Tech., 15, 6257–6283, https://doi.org/10.5194/amt-15-6257-2022, https://doi.org/10.5194/amt-15-6257-2022, 2022
Short summary
Short summary
The article investigates different S5P TROPOMI cloud retrieval algorithms for tropospheric trace gas retrievals. The cloud products show differences primarily over snow and ice and for scenes under sun glint. Some issues regarding across-track dependence are found for the cloud fractions as well as for the cloud heights.
Yulan Hong, Robert J. Trapp, Stephen W. Nesbitt, and Larry Di Girolamo
Atmos. Meas. Tech. Discuss., https://doi.org/10.5194/amt-2022-286, https://doi.org/10.5194/amt-2022-286, 2022
Revised manuscript accepted for AMT
Short summary
Short summary
Deep convective updrafts form overshooting tops (OT) when they extend into the upper troposphere and lower stratosphere. An OT often indicates hazardous weather conditions. Global distribution of OTs is useful for understanding global severe weather conditions. Moderate Resolution Imaging Spectroradiometer (MODIS) on both Aqua and Terra satellites has provide two-decade records of Earth-atmosphere system with stable orbits, which is used in this study to derive a 20-yr OT climatology.
Andrew Mark Sayer, Luca Lelli, Brian Cairns, Bastiaan van Diedenhoven, Amir Ibrahim, Kirk Knobelspiesse, Sergey Korkin, and P. Jeremy Werdell
Atmos. Meas. Tech. Discuss., https://doi.org/10.5194/amt-2022-201, https://doi.org/10.5194/amt-2022-201, 2022
Revised manuscript accepted for AMT
Short summary
Short summary
This paper presents a method to estimate the height of the top of clouds above Earth's surface using satellite measurements. It is based on light absorption by oxygen in Earth's atmosphere, which darkens the signal that a satellite will see at certain wavelengths of light. Clouds "shield" the satellite from some of this darkening, dependent on cloud height (and other factors), because clouds scatter light at these wavelengths. The method will be applied to the future NASA PACE mission.
Han Ding, Haoran Li, and Liping Liu
Atmos. Meas. Tech., 15, 6181–6200, https://doi.org/10.5194/amt-15-6181-2022, https://doi.org/10.5194/amt-15-6181-2022, 2022
Short summary
Short summary
In this study, a framework for processing the Doppler spectra observations of a multi-mode pulse compression Ka–Ku cloud radar system is presented. We first proposed an approach to identify and remove the clutter signals in the Doppler spectrum. Then, we developed a new algorithm to remove the range sidelobe at the modes implementing the pulse compression technique. The radar observations from different modes were then merged using the shift-then-average method.
Andrew T. Prata, Roy G. Grainger, Isabelle A. Taylor, Adam C. Povey, Simon R. Proud, and Caroline A. Poulsen
Atmos. Meas. Tech., 15, 5985–6010, https://doi.org/10.5194/amt-15-5985-2022, https://doi.org/10.5194/amt-15-5985-2022, 2022
Short summary
Short summary
Satellite observations are often used to track ash clouds and estimate their height, particle sizes and mass; however, satellite-based techniques are always associated with some uncertainty. We describe advances in a satellite-based technique that is used to estimate ash cloud properties for the June 2019 Raikoke (Russia) eruption. Our results are significant because ash warning centres increasingly require uncertainty information to correctly interpret,
aggregate and utilise the data.
Adrià Amell, Patrick Eriksson, and Simon Pfreundschuh
Atmos. Meas. Tech., 15, 5701–5717, https://doi.org/10.5194/amt-15-5701-2022, https://doi.org/10.5194/amt-15-5701-2022, 2022
Short summary
Short summary
Geostationary satellites continuously image a given location on Earth, a feature that satellites designed to characterize atmospheric ice lack. However, the relationship between geostationary images and atmospheric ice is complex. Machine learning is used here to leverage such images to characterize atmospheric ice throughout the day in a probabilistic manner. Using structural information from the image improves the characterization, and this approach compares favourably to traditional methods.
Jesse Loveridge, Aviad Levis, Larry Di Girolamo, Vadim Holodovsky, Linda Forster, Anthony B. Davis, and Yoav Y. Schechner
Atmos. Meas. Tech. Discuss., https://doi.org/10.5194/amt-2022-251, https://doi.org/10.5194/amt-2022-251, 2022
Revised manuscript accepted for AMT
Short summary
Short summary
We describe a new method for measuring the 3D spatial variations of 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.
Alistair Bell, Pauline Martinet, Olivier Caumont, Frédéric Burnet, Julien Delanoë, Susana Jorquera, Yann Seity, and Vinciane Unger
Atmos. Meas. Tech., 15, 5415–5438, https://doi.org/10.5194/amt-15-5415-2022, https://doi.org/10.5194/amt-15-5415-2022, 2022
Short summary
Short summary
Cloud radars and microwave radiometers offer the potential to improve fog forecasts when assimilated into a high-resolution model. As this process can be complex, a retrieval of model variables is sometimes made as a first step. In this work, results from a 1D-Var algorithm for the retrieval of temperature, humidity and cloud liquid water content are presented. The algorithm is applied first to a synthetic dataset and then to a dataset of real measurements from a recent field campaign.
Willi Schimmel, Heike Kalesse-Los, Maximilian Maahn, Teresa Vogl, Andreas Foth, Pablo Saavedra Garfias, and Patric Seifert
Atmos. Meas. Tech., 15, 5343–5366, https://doi.org/10.5194/amt-15-5343-2022, https://doi.org/10.5194/amt-15-5343-2022, 2022
Short summary
Short summary
This study introduces the novel Doppler radar spectra-based machine learning approach VOODOO (reVealing supercOOled liquiD beyOnd lidar attenuatiOn). VOODOO is a powerful probability-based extension to the existing Cloudnet hydrometeor target classification, enabling the detection of liquid-bearing cloud layers beyond complete lidar attenuation via user-defined p* threshold. VOODOO performs best for (multi-layer) stratiform and deep mixed-phase clouds with liquid water path > 100 g m−2.
Vikas Nataraja, Sebastian Schmidt, Hong Chen, Takanobu Yamaguchi, Jan Kazil, Graham Feingold, Kevin Wolf, and Hironobu Iwabuchi
Atmos. Meas. Tech., 15, 5181–5205, https://doi.org/10.5194/amt-15-5181-2022, https://doi.org/10.5194/amt-15-5181-2022, 2022
Short summary
Short summary
A convolutional neural network (CNN) is introduced to retrieve cloud optical thickness (COT) from passive cloud imagery. The CNN, trained on large eddy simulations from the Sulu Sea, learns from spatial information at multiple scales to reduce cloud inhomogeneity effects. By considering the spatial context of a pixel, the CNN outperforms the traditional independent pixel approximation (IPA) across several cloud morphology metrics.
Rachel T. Pinker, Yingtao Ma, Wen Chen, Istvan Laszlo, Hongqing Liu, Hye-Yun Kim, and Jaime Daniels
Atmos. Meas. Tech., 15, 5077–5094, https://doi.org/10.5194/amt-15-5077-2022, https://doi.org/10.5194/amt-15-5077-2022, 2022
Short summary
Short summary
Scene-dependent narrow-to-broadband transformations are developed to facilitate the use of observations from the Advanced Baseline Imager (ABI), the primary instrument on GOES-R, to derive surface shortwave radiative fluxes. This is a first NOAA product at the high resolution of about 5 k over the contiguous United States (CONUS) region. The product is archived and can be downloaded from the NOAA Comprehensive Large Array-data Stewardship System (CLASS).
Mariko Oue, Stephen M. Saleeby, Peter J. Marinescu, Pavlos Kollias, and Susan C. van den Heever
Atmos. Meas. Tech., 15, 4931–4950, https://doi.org/10.5194/amt-15-4931-2022, https://doi.org/10.5194/amt-15-4931-2022, 2022
Short summary
Short summary
This study provides an optimization of radar observation strategies to better capture convective cell evolution in clean and polluted environments as well as a technique for the optimization. The suggested optimized radar observation strategy is to better capture updrafts at middle and upper altitudes and precipitation particle evolution of isolated deep convective clouds. This study sheds light on the challenge of designing remote sensing observation strategies in pre-field campaign periods.
Jean-Marie Lalande, Guillaume Bourmaud, Pierre Minvielle, and Jean-François Giovannelli
Atmos. Meas. Tech., 15, 4411–4429, https://doi.org/10.5194/amt-15-4411-2022, https://doi.org/10.5194/amt-15-4411-2022, 2022
Short summary
Short summary
In this paper we describe the implementation of an interpolation–prediction estimator applied to cloud properties derived from CloudSat observations. The objective is to evaluate the uncertainty associated with the estimated quantity. The model developed in this study can be valuable for satellite applications (GPS, telecommunication) as well as for cloud product comparisons. This paper is didactic and beneficial for anyone interested in kriging estimators.
Julia Fuchs, Hendrik Andersen, Jan Cermak, Eva Pauli, and Rob Roebeling
Atmos. Meas. Tech., 15, 4257–4270, https://doi.org/10.5194/amt-15-4257-2022, https://doi.org/10.5194/amt-15-4257-2022, 2022
Short summary
Short summary
Two cloud-masking approaches, a local and a regional approach, using high-resolution satellite data are developed and validated for the region of Paris to improve applicability for analyses of urban effects on low clouds. We found that cloud masks obtained from the regional approach are more appropriate for the high-resolution analysis of locally induced cloud processes. Its applicability is tested for the analysis of typical fog conditions over different surface types.
Eleni Tetoni, Florian Ewald, Martin Hagen, Gregor Köcher, Tobias Zinner, and Silke Groß
Atmos. Meas. Tech., 15, 3969–3999, https://doi.org/10.5194/amt-15-3969-2022, https://doi.org/10.5194/amt-15-3969-2022, 2022
Short summary
Short summary
We use the C-band POLDIRAD and the Ka-band MIRA-35 to perform snowfall dual-wavelength polarimetric radar measurements. We develop an ice microphysics retrieval for mass, apparent shape, and median size of the particle size distribution by comparing observations to T-matrix ice spheroid simulations while varying the mass–size relationship. We furthermore show how the polarimetric measurements from POLDIRAD help to narrow down ambiguities between ice particle shape and size.
Assia Arouf, Hélène Chepfer, Thibault Vaillant de Guélis, Marjolaine Chiriaco, Matthew D. Shupe, Rodrigo Guzman, Artem Feofilov, Patrick Raberanto, Tristan S. L'Ecuyer, Seiji Kato, and Michael R. Gallagher
Atmos. Meas. Tech., 15, 3893–3923, https://doi.org/10.5194/amt-15-3893-2022, https://doi.org/10.5194/amt-15-3893-2022, 2022
Short summary
Short summary
We proposed new estimates of the surface longwave (LW) cloud radiative effect (CRE) derived from observations collected by a space-based lidar on board the CALIPSO satellite and radiative transfer computations. Our estimate appropriately captures the surface LW CRE annual variability over bright polar surfaces, and it provides a dataset more than 13 years long.
Baike Xi, Xiquan Dong, Xiaojian Zheng, and Peng Wu
Atmos. Meas. Tech., 15, 3761–3777, https://doi.org/10.5194/amt-15-3761-2022, https://doi.org/10.5194/amt-15-3761-2022, 2022
Short summary
Short summary
This study develops an innovative method to determine the cloud phases over the Southern Ocean (SO) using the combination of radar and lidar measurements during the ship-based field campaign of MARCUS. Results from our study show that the low-level, deep, and shallow cumuli are dominant, and the mixed-phase clouds occur more than single phases over the SO. The mixed-phase cloud properties are similar to liquid-phase (ice-phase) clouds in the midlatitudes (polar) region of the SO.
Adrien Guyot, Alain Protat, Simon P. Alexander, Andrew R. Klekociuk, Peter Kuma, and Adrian McDonald
Atmos. Meas. Tech., 15, 3663–3681, https://doi.org/10.5194/amt-15-3663-2022, https://doi.org/10.5194/amt-15-3663-2022, 2022
Short summary
Short summary
Ceilometers are instruments that are widely deployed as part of operational networks. They are usually not able to detect cloud phase. Here, we propose an evaluation of various methods to detect supercooled liquid water with ceilometer observations, using an extensive dataset from Davis, Antarctica. Our results highlight the possibility for ceilometers to detect supercooled liquid water in clouds.
Xiaotong Li, Baozhu Wang, Bo Qiu, and Chao Wu
Atmos. Meas. Tech., 15, 3629–3639, https://doi.org/10.5194/amt-15-3629-2022, https://doi.org/10.5194/amt-15-3629-2022, 2022
Short summary
Short summary
The all-sky camera images can reflect the local cloud cover, which is considerable for astronomical observatory site selection. Therefore, the realization of automatic classification of the images is very important. In this paper, three cloud cover features are proposed to classify the images. The proposed method is evaluated on a large dataset, and the method achieves an accuracy of 96.58 % and F1_score of 96.24 %, which greatly improves the efficiency of automatic processing of the images.
Huige Di, Yun Yuan, Qing Yan, Wenhui Xin, Shichun Li, Jun Wang, Yufeng Wang, Lei Zhang, and Dengxin Hua
Atmos. Meas. Tech., 15, 3555–3567, https://doi.org/10.5194/amt-15-3555-2022, https://doi.org/10.5194/amt-15-3555-2022, 2022
Short summary
Short summary
It is necessary to correctly evaluate the amount of cloud water resources in an area. Currently, there is a lack of effective observation methods for atmospheric column condensate evaluation. We propose a method for atmospheric column condensate by combining millimetre cloud radar, lidar and microwave radiometers. The method can realise determination of atmospheric column condensate. The variation of cloud before precipitation is considered, and the atmospheric column is deduced and obtained.
Bhupendra A. Raut, Scott Collis, Nicola Ferrier, Paytsar Muradyan, Rajesh Sankaran, Robert Jackson, Sean Shahkarami, Seongha Park, Dario Dematties, Yongho Kim, Joseph Swantek, Neal Conrad, Wolfgang Gerlach, Sergey Shemyakin, and Pete Beckman
Atmos. Meas. Tech. Discuss., https://doi.org/10.5194/amt-2022-159, https://doi.org/10.5194/amt-2022-159, 2022
Revised manuscript accepted for AMT
Short summary
Short summary
We tested the quality of cloud motion from sky camera images using the phase correlation method to optimize the algorithm in real time. Increased framerate and large image block considerably improved the quality, but image resolution and the color channel had a minor effect. In images from the revolving camera system, raindrop contamination is identified using the rotational motion of the drops.
Daniel Robbins, Caroline Poulsen, Steven Siems, and Simon Proud
Atmos. Meas. Tech., 15, 3031–3051, https://doi.org/10.5194/amt-15-3031-2022, https://doi.org/10.5194/amt-15-3031-2022, 2022
Short summary
Short summary
A neural network (NN)-based cloud mask for a geostationary satellite instrument, AHI, is developed using collocated data and is better at not classifying thick aerosols as clouds versus the Japanese Meteorological Association and the Bureau of Meteorology masks, identifying 1.13 and 1.29 times as many non-cloud pixels than each mask, respectively. The improvement during the day likely comes from including the shortest wavelength bands from AHI in the NN mask, which the other masks do not use.
Pascal Marquet, Pauline Martinet, Jean-François Mahfouf, Alina Lavinia Barbu, and Benjamin Ménétrier
Atmos. Meas. Tech., 15, 2021–2035, https://doi.org/10.5194/amt-15-2021-2022, https://doi.org/10.5194/amt-15-2021-2022, 2022
Short summary
Short summary
Two conservative thermodynamic variables (moist-air entropy potential temperature and total water content) are introduced into a one-dimensional EnVar data assimilation system to demonstrate their benefit for future operational assimilation schemes, with the use of microwave brightness temperatures from a ground-based radiometer installed during the field campaign SOFGO3D. Results show that the brightness temperatures analysed with the new variables are improved, including the liquid water.
Valery Shcherbakov, Frédéric Szczap, Alaa Alkasem, Guillaume Mioche, and Céline Cornet
Atmos. Meas. Tech., 15, 1729–1754, https://doi.org/10.5194/amt-15-1729-2022, https://doi.org/10.5194/amt-15-1729-2022, 2022
Short summary
Short summary
We performed extensive Monte Carlo (MC) simulations of lidar signals and developed an empirical model to account for the multiple scattering in the lidar signals. The simulations have taken into consideration four types of lidar configurations (the ground based, the airborne, the CALIOP, and the ATLID) and four types of particles (coarse aerosol, water cloud, jet-stream cirrus, and cirrus).
The empirical model has very good quality of MC data fitting for all considered cases.
Alexander Myagkov and Davide Ori
Atmos. Meas. Tech., 15, 1333–1354, https://doi.org/10.5194/amt-15-1333-2022, https://doi.org/10.5194/amt-15-1333-2022, 2022
Short summary
Short summary
This study provides equations to characterize random errors of spectral polarimetric observations from cloud radars. The results can be used for a broad spectrum of applications. For instance, accurate error characterization is essential for advanced retrievals of microphysical properties of clouds and precipitation. Moreover, error characterization allows for the use of measurements from polarimetric cloud radars to potentially improve weather forecasts.
Yuli Liu and Gerald G. Mace
Atmos. Meas. Tech., 15, 927–944, https://doi.org/10.5194/amt-15-927-2022, https://doi.org/10.5194/amt-15-927-2022, 2022
Short summary
Short summary
We propose a suite of Bayesian algorithms for synergistic radar and radiometer retrievals to evaluate the next-generation NASA Cloud, Convection and Precipitation (CCP) observing system. The algorithms address pixel-level retrievals using active-only, passive-only, and synergistic active–passive observations. Novel techniques in developing synergistic algorithms are presented. Quantitative assessments of the CCP observing system's capability in retrieving ice cloud microphysics are provided.
William K. Jones, Matthew W. Christensen, and Philip Stier
Atmos. Meas. Tech. Discuss., https://doi.org/10.5194/amt-2022-31, https://doi.org/10.5194/amt-2022-31, 2022
Revised manuscript accepted for AMT
Short summary
Short summary
Geostationary weather satellites have been used to detect storm clouds since their earliest applications. However, this task remains difficult as imaging satellites cannot observe the strong vertical winds that are characteristic of storm clouds. Here we introduce a new method that allows us to detect the early development of storms and continue to track them throughout their lifetime, allowing us to study how their early behaviour affects subsequent weather.
Yann Fabel, Bijan Nouri, Stefan Wilbert, Niklas Blum, Rudolph Triebel, Marcel Hasenbalg, Pascal Kuhn, Luis F. Zarzalejo, and Robert Pitz-Paal
Atmos. Meas. Tech., 15, 797–809, https://doi.org/10.5194/amt-15-797-2022, https://doi.org/10.5194/amt-15-797-2022, 2022
Short summary
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.
Cuong M. Nguyen, Mengistu Wolde, Alessandro Battaglia, Leonid Nichman, Natalia Bliankinshtein, Samuel Haimov, Kenny Bala, and Dirk Schuettemeyer
Atmos. Meas. Tech., 15, 775–795, https://doi.org/10.5194/amt-15-775-2022, https://doi.org/10.5194/amt-15-775-2022, 2022
Short summary
Short summary
An analysis of airborne triple-frequency radar and almost perfectly co-located coincident in situ data from an Arctic storm confirms the main findings of modeling work with radar dual-frequency ratios (DFRs) at different zones of the DFR plane associated with different ice habits. High-resolution CPI images provide accurate identification of rimed particles within the DFR plane. The relationships between the triple-frequency signals and cloud microphysical properties are also presented.
Heba S. Marey, James R. Drummond, Dylan B. A. Jones, Helen Worden, Merritt N. Deeter, John Gille, and Debbie Mao
Atmos. Meas. Tech., 15, 701–719, https://doi.org/10.5194/amt-15-701-2022, https://doi.org/10.5194/amt-15-701-2022, 2022
Short summary
Short summary
In this study, an analysis has been performed to understand the improvements in observational coverage over Canada in the new MOPITT V9 product. Temporal and spatial analysis of V9 indicates a general coverage gain of 15–20 % relative to V8, which varies regionally and seasonally; e.g., the number of successful MOPITT retrievals in V9 was doubled over Canada in winter. Also, comparison with the corresponding IASI instrument indicated generally good agreement, with about a 5–10 % positive bias.
Pragya Vishwakarma, Julien Delanoë, Susana Jorquera, Pauline Martinet, Frederic Burnet, Alistair Bell, and Jean-Charles Dupont
Atmos. Meas. Tech. Discuss., https://doi.org/10.5194/amt-2022-3, https://doi.org/10.5194/amt-2022-3, 2022
Revised manuscript accepted for AMT
Short summary
Short summary
Cloud observations are necessary to characterize the cloud properties at the local and global scales. These observations must be translated to cloud geophysical parameters. This paper presents the estimation of liquid water content (LWC) using radar and microwave radiometer (MWR) measurements. LWP from MWR scales the LWC and the scaling factor (lna) is retrieved. The retrievals are compared with in-situ observations. A climatology of lna is built to estimate LWC with only radar information.
Teresa Vogl, Maximilian Maahn, Stefan Kneifel, Willi Schimmel, Dmitri Moisseev, and Heike Kalesse-Los
Atmos. Meas. Tech., 15, 365–381, https://doi.org/10.5194/amt-15-365-2022, https://doi.org/10.5194/amt-15-365-2022, 2022
Short summary
Short summary
We are using machine learning techniques, a type of artificial intelligence, to detect graupel formation in clouds. The measurements used as input to the machine learning framework were performed by cloud radars. Cloud radars are instruments located at the ground, emitting radiation with wavelenghts of a few millimeters vertically into the cloud and measuring the back-scattered signal. Our novel technique can be applied to different radar systems and different weather conditions.
Heike Kalesse-Los, Willi Schimmel, Edward Luke, and Patric Seifert
Atmos. Meas. Tech., 15, 279–295, https://doi.org/10.5194/amt-15-279-2022, https://doi.org/10.5194/amt-15-279-2022, 2022
Short summary
Short summary
It is important to detect the vertical distribution of cloud droplets and ice in mixed-phase clouds. Here, an artificial neural network (ANN) previously developed for Arctic clouds is applied to a mid-latitudinal cloud radar data set. The performance of this technique is contrasted to the Cloudnet target classification. For thick/multi-layer clouds, the machine learning technique is better at detecting liquid than Cloudnet, but if lidar data are available Cloudnet is at least as good as the ANN.
Christian Matar, Céline Cornet, Frédéric Parol, Laurent C.-Labonnote, Frédérique Auriol, and Jean-Marc Nicolas
Atmos. Meas. Tech. Discuss., https://doi.org/10.5194/amt-2021-414, https://doi.org/10.5194/amt-2021-414, 2022
Revised manuscript accepted for AMT
Short summary
Short summary
The uncertainties in cloud remote sensing can propagate to the retrieved cloud properties and they need to be quantified. We present the formalism of error extraction and we apply it on the cloud properties retrieved from the measurements of the airborne radiometer OSIRIS. We show that errors related to measurement uncertainties reach 10 %. Errors related to the simplified model assuming that the clouds are plane-parallel and homogeneous lead to uncertainties exceeding 10 %.
Jean-François Ribaud, Martial Haeffelin, Jean-Charles Dupont, Marc-Antoine Drouin, Felipe Toledo, and Simone Kotthaus
Atmos. Meas. Tech., 14, 7893–7907, https://doi.org/10.5194/amt-14-7893-2021, https://doi.org/10.5194/amt-14-7893-2021, 2021
Short summary
Short summary
PARAFOG is a near-real-time decision tool that aims to retrieve pre-fog alert levels minutes to hours prior to fog onset. The second version of PARAFOG allows us to discriminate between radiation and stratus lowering fog situations. It is based upon the combination of visibility observations and automatic lidar and ceilometer measurements. The overall performance of the second version of PARAFOG over more than 300 fog cases at five different locations presents a good perfomance.
Andrew Geiss and Joseph C. Hardin
Atmos. Meas. Tech., 14, 7729–7747, https://doi.org/10.5194/amt-14-7729-2021, https://doi.org/10.5194/amt-14-7729-2021, 2021
Short summary
Short summary
Radars can suffer from missing or poor-quality data regions for several reasons: beam blockage, instrument failure, and near-ground blind zones, etc. Here, we demonstrate how deep convolutional neural networks can be used for filling in radar-missing data regions and that they can significantly outperform conventional approaches in terms of realism and accuracy.
Frank Werner, Nathaniel J. Livesey, Michael J. Schwartz, William G. Read, Michelle L. Santee, and Galina Wind
Atmos. Meas. Tech., 14, 7749–7773, https://doi.org/10.5194/amt-14-7749-2021, https://doi.org/10.5194/amt-14-7749-2021, 2021
Short summary
Short summary
In this study we present an improved cloud detection scheme for the Microwave Limb Sounder, which is based on a feedforward artificial neural network. This new algorithm is shown not only to reliably detect high and mid-level convection containing even small amounts of cloud water but also to distinguish between high-reaching and mid-level to low convection.
Kamil Mroz, Alessandro Battaglia, Cuong Nguyen, Andrew Heymsfield, Alain Protat, and Mengistu Wolde
Atmos. Meas. Tech., 14, 7243–7254, https://doi.org/10.5194/amt-14-7243-2021, https://doi.org/10.5194/amt-14-7243-2021, 2021
Short summary
Short summary
A method for estimating microphysical properties of ice clouds based on radar measurements is presented. The algorithm exploits the information provided by differences in the radar response at different frequency bands in relation to changes in the snow morphology. The inversion scheme is based on a statistical relation between the radar simulations and the properties of snow calculated from in-cloud sampling.
Nicholas J. Kedzuf, J. Christine Chiu, V. Chandrasekar, Sounak Biswas, Shashank S. Joshil, Yinghui Lu, Peter Jan van Leeuwen, Christopher Westbrook, Yann Blanchard, and Sebastian O'Shea
Atmos. Meas. Tech., 14, 6885–6904, https://doi.org/10.5194/amt-14-6885-2021, https://doi.org/10.5194/amt-14-6885-2021, 2021
Short summary
Short summary
Ice clouds play a key role in our climate system due to their strong controls on precipitation and the radiation budget. However, it is difficult to characterize co-existing ice species using radar observations. We present a new method that separates the radar signals of pristine ice embedded in snow aggregates and retrieves their respective abundances and sizes for the first time. The ability to provide their quantitative microphysical properties will open up many research opportunities.
Gianluca Di Natale, Marco Barucci, Claudio Belotti, Giovanni Bianchini, Francesco D'Amato, Samuele Del Bianco, Marco Gai, Alessio Montori, Ralf Sussmann, Silvia Viciani, Hannes Vogelmann, and Luca Palchetti
Atmos. Meas. Tech., 14, 6749–6758, https://doi.org/10.5194/amt-14-6749-2021, https://doi.org/10.5194/amt-14-6749-2021, 2021
Short summary
Short summary
The importance of cirrus and mixed-phase clouds in the Earth radiation budget has been proven by many studies. In this paper the properties that characterize these clouds are retrieved from lidar and far-infrared spectral measurements performed in winter 2018/19 on the Zugspitze (Germany). The synergy of lidar and spectrometer measurements allowed us to assess the exponent k of the power-law relationship between the backscattering and the extinction coefficients.
Alan J. Geer
Atmos. Meas. Tech., 14, 5369–5395, https://doi.org/10.5194/amt-14-5369-2021, https://doi.org/10.5194/amt-14-5369-2021, 2021
Short summary
Short summary
Satellite observations sensitive to cloud and precipitation help improve the quality of weather forecasts. However, they are sensitive to things that models do not forecast, such as the shapes and sizes of snow and ice particles. These details can be estimated from the observations themselves and then incorporated in the satellite simulators used in weather forecasting. This approach, known as parameter estimation, will be increasingly useful to build models of poorly known physical processes.
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.
Hartwig Deneke, Carola Barrientos-Velasco, Sebastian Bley, Anja Hünerbein, Stephan Lenk, Andreas Macke, Jan Fokke Meirink, Marion Schroedter-Homscheidt, Fabian Senf, Ping Wang, Frank Werner, and Jonas Witthuhn
Atmos. Meas. Tech., 14, 5107–5126, https://doi.org/10.5194/amt-14-5107-2021, https://doi.org/10.5194/amt-14-5107-2021, 2021
Short summary
Short summary
The SEVIRI instrument flown on the European geostationary Meteosat satellites acquires multi-spectral images at a relatively coarse pixel resolution of 3 × 3 km2, but it also has a broadband high-resolution visible channel with 1 × 1 km2 spatial resolution. In this study, the modification of an existing cloud property and solar irradiance retrieval to use this channel to improve the spatial resolution of its output products as well as the resulting benefits for applications are described.
Florian Ewald, Silke Groß, Martin Wirth, Julien Delanoë, Stuart Fox, and Bernhard Mayer
Atmos. Meas. Tech., 14, 5029–5047, https://doi.org/10.5194/amt-14-5029-2021, https://doi.org/10.5194/amt-14-5029-2021, 2021
Short summary
Short summary
In this study, we show how solar radiance observations can be used to validate and further constrain ice cloud microphysics retrieved from the synergy of radar–lidar measurements. Since most radar–lidar retrievals rely on a global assumption about the ice particle shape, ice water content and particle size biases are to be expected in individual cloud regimes. In this work, we identify and correct these biases by reconciling simulated and measured solar radiation reflected from these clouds.
Cited articles
Beleites, C., Neugebauer, U., Bocklitz, T., Krafft, C., and Popp, J.:
Sample size planning for classification models, Anal. Chim. Acta,
760C, 25–33, https://doi.org/10.1016/j.aca.2012.11.007, 2013.
Calbó, J. and Sabburg, J.: Feature Extraction from Whole-Sky
Ground-Based Images for Cloud-Type Recognition, J. Atmos. Ocean. Technol.,
25, 3–14, https://doi.org/10.1175/2007JTECHA959.1, 2008.
Carreira, J., Caseiro, R., Batista, J., and Sminchisescu, C.: Free-Form
Region Description with Second-Order Pooling, IEEE T. Pattern Anal., 37, 1177–1189, https://doi.org/10.1109/TPAMI.2014.2361137, 2015.
Chang, C.-C. and Lin, C.-J.: LIBSVM: A library for support vector machines,
ACM T. Intell. Syst. Technol., 2, 1–39, https://doi.org/10.1145/1961189.1961199, 2007.
Chen, T., Rossow, W. B., and Zhang, Y.: Radiative Effects of Cloud-Type
Variations, J. Clim., 13, 264–286, https://doi.org/10.1175/1520-0442(2000)013<0264:reoctv>2.0.co;2, 2000.
Cheng, H.-Y. and Yu, C.-C.: Block-based cloud classification with statistical features and distribution of local texture features, Atmos. Meas. Tech., 8, 1173–1182, https://doi.org/10.5194/amt-8-1173-2015, 2015.
Dev, S., Savoy, F. M., Lee, Y. H., and Winkler, S.: WAHRSIS: A low-cost
high-resolution whole sky imager with near-infrared capabilities, in:
Proceedings of SPIE – The International Society for Optical Engineering
9071:90711L on SPIE Defense + Security, Baltimore, Maryland, USA, 5–9 May
2014, 90711L, https://doi.org/10.1117/12.2052982, 2014.
Dev, S., Lee, Y. H., and Winkler, S.:
Categorization of cloud image patches using an improved texton-based approach, in: Proceedings of 2015 IEEE International Conference on Image Processing (ICIP), Québec City, Canada, 27–30 September, 422–426, https://doi.org/10.1109/ICIP.2015.7350833, 2015 (data available at: http://vintage.winklerbros.net/swimcat.html, last access: 23 January 2021).
Fang, L., He, N., Li, S., Plaza, A. J., and Plaza, J.: A New
Spatial–Spectral Feature Extraction Method for Hyperspectral Images Using
Local Covariance Matrix Representation, IEEE T. Geosci. Remote,
56, 3534–3546, https://doi.org/10.1109/TGRS.2018.2801387, 2018.
Faraki, M., Harandi, M. T., and Porikli, F.: More about VLAD: A leap from
Euclidean to Riemannian manifolds, in: Proceedings of 2015 IEEE Conference
on Computer Vision and Pattern Recognition, Boston, USA, 7–12 June 2015, 4951–4960, 2015a.
Faraki, M., Palhang, M., and Sanderson, C.: Log-Euclidean Bag of Words for
Human Action Recognition, IET Comput. Vision, 9, 331–339,
https://doi.org/10.1049/iet-cvi.2014.0018, 2015b.
Guo, K., Ishwar, P., and Konrad, J.: Action Recognition Using Sparse
Representation on Covariance Manifolds of Optical Flow, in: Proceedings of
the 7th IEEE International Conference on Advanced Video and Signal Based
Surveillance, Boston, USA, 29 August–1 September 2010, 188–195, 2010.
Hartmann, D. L., Ockert-Bell, M. E., and Michelsen, M. L.: The Effect of
Cloud Type on Earth's Energy Balance: Global Analysis, J. Clim., 5,
1281–1304, https://doi.org/10.1175/1520-0442(1992)005<1281:Teocto>2.0.Co;2, 1992.
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.
Hsu, C.-W. and Lin, C.-J.: A Comparison of Methods for Multiclass Support
Vector Machines, IEEE T. Neur. Net. Lear., 13, 415–425, https://doi.org/10.1109/72.991427, 2002.
Huang, Z., Wang, R., Shan, S., Gool, L. V., and Chen, X.: Cross
Euclidean-to-Riemannian Metric Learning with Application to Face Recognition
from Video, IEEE T. Pattern Anal., 40, 2827–2840,
https://doi.org/10.1109/TPAMI.2017.2776154, 2018.
Huertas, J. R.-B., Pozo, D., Aler, R., and Galván, I. M.: Genetic
programming to extract features from the whole-sky camera for cloud type
classification, in: Proceedings of the International Conference on Renewable
Energies and Power Quality, Malaga, Spain, 4–6 April 2017, 132–136, 2017.
Jégou, H., Perronnin, F., Douze, M., Sánchez, J., Pérez, P., and
Schmid, C.: Aggregating local image descriptors into compact codes, IEEE
T. Pattern Anal., 34, 1704–1716, https://doi.org/10.1109/TPAMI.2011.235, 2012.
Joubert, A.: The astronomical image – Toward an objective analysis,
Lastronomie, 93, 3–30, 1978.
Kazantzidis, A., Tzoumanikas, P., Bais, A., Fotopoulos, S., and Economou,
G.: Cloud Detection and Classification with the Use of Whole-Sky
Ground-Based Images, Atmos. Res., 113, 8-0-88,
https://doi.org/10.1016/j.atmosres.2012.05.005, 2012.
Knerr, S., Personnaz, L., and Dreyfus, G.: Single-layer learning revisited:
a stepwise procedure for building and training a neural network, in:
Neurocomputing – Algorithms, Architectures and Applications, Springer,
Berlin, Heidelberg, Germany, 41–50, 1990.
Kreßel, U. H.-G.: Pairwise classification and support vector machines, in: Advances in kernel methods: support vector learning, MIT Press,
Cambridge, 255–268, 1999.
Krizhevsky, A., Sutskever, I., and Hinton, G. E.: ImageNet Classification
with Deep Convolutional Neural Networks, in: Proceedings of the 25th
International Conference on Neural Information Processing Systems - Volume
1, Lake Tahoe, Nevada, USA, 3–8 December 2012, 1097–1105, 2012.
Li, M., Liu, S., and Zhang, Z.: Deep tensor fusion network for multimodal
ground-based cloud classification in weather station networks, Ad Hoc
Networks, 96, 101991, https://doi.org/10.1016/j.adhoc.2019.101991, 2020.
Li, P., Wang, Q., Zuo, W., and Zhang, L.: Log-Euclidean Kernels for Sparse
Representation and Dictionary Learning, in: Proceedings of 2013 IEEE
International Conference on Computer Vision, Sydney, NSW, Australia, 1–8 December 2013, 1601–1608, 2013.
Liu, L., Sun, X., Chen, F., Zhao, S., and Gao, T.: Cloud Classification
Based on Structure Features of Infrared Images, J. Atmos. Ocean. Technol.,
28, 410–417, https://doi.org/10.1175/2010jtecha1385.1, 2011.
Liu, L., Sun, X., Gao, T., and Zhao, S.: Comparison of Cloud Properties from
Ground-Based Infrared Cloud Measurement and Visual Observations, J. Atmos.
Ocean. Technol., 30, 1171–1179, https://doi.org/10.1175/jtech-d-12-00157.1, 2013.
Liu, S., Duan, L., Zhang, Z., and Cao, X.: Hierarchical Multimodal Fusion
for Ground-Based Cloud Classification in Weather Station Networks, IEEE
Access, 7, 85688–85695, https://doi.org/10.1109/ACCESS.2019.2926092, 2019.
Liu, S. and Li, M.: Deep multimodal fusion for ground-based cloud
classification in weather station networks, EURASIP J. Wirel. Comm., 2018, 48, https://doi.org/10.1186/s13638-018-1062-0, 2018.
Liu, S., Zhang, Z., and Mei, X.: Ground-based cloud classification using
weighted local binary patterns, J. Appl. Remote Sens., 9, 095062,
https://doi.org/10.1117/1.JRS.9.095062, 2015.
Liu, S., Li, M., Zhang, Z., Xiao, B., and Cao, X.: Multimodal Ground-Based
Cloud Classification Using Joint Fusion Convolutional Neural Network, Remote
Sens., 10, 822, https://doi.org/10.3390/rs10060822, 2018.
Luo, Q., Meng, Y., Liu, L., Zhao, X., and Zhou, Z.: Cloud classification of ground-based infrared images combining manifold and texture features, Atmos. Meas. Tech., 11, 5351–5361, https://doi.org/10.5194/amt-11-5351-2018, 2018.
Pang, Y., Yuan, Y., and Li, X.: Gabor-Based Region Covariance Matrices for
Face Recognition, IEEE T. Circ. Syst. Vid., 18, 989–993,
https://doi.org/10.1109/TCSVT.2008.924108, 2008.
Pennec, X.: Intrinsic Statistics on Riemannian Manifolds: Basic Tools for
Geometric Measurements, J. Math. Imaging Vis., 25, 127–154,
https://doi.org/10.1007/s10851-006-6228-4, 2006.
Pennec, X., Fillard, P., and Ayache, N.: A Riemannian Framework for Tensor
Computing, Int. J. Comput. Vision, 66, 41–66, https://doi.org/10.1007/s11263-005-3222-z,
2006.
Ramanathan, V., Cess, R., Harrison, E. F., Minnis, P., Barkstrom, R. B.,
Ahmad, E., and Hartmann, D.: Cloud-radiative forcing and climate: Results
from the Earth's radiation budget, Science, 243, 57–63,
https://doi.org/10.1126/science.243.4887.57, 1989.
Shi, C., Wang, C., Wang, Y., and Xiao, B.: Deep Convolutional
Activations-Based Features for Ground-Based Cloud Classification, IEEE
Geosci. Remote S., 14, 816–820, https://doi.org/10.1109/LGRS.2017.2681658, 2017.
Simonyan, K. and Zisserman, A.: Very Deep Convolutional Networks for
Large-Scale Image Recognition, in: Proceedings of the 3rd International
Conference on Learning Representations, San Diego, CA, USA, 7–9 May 2015,
1–14, 2015.
Sra, S.: A new metric on the manifold of kernel matrices with application to
matrix geometric means, in: Proceedings of the 25th International Conference
on Neural Information Processing Systems – Volume 1, Lake Tahoe, Nevada,
USA, 3–8 December 2012, 144–152, 2012.
Taravat, A., Frate, F. D., 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 S., 12, 666–670,
https://doi.org/10.1109/LGRS.2014.2356616, 2015.
Tuzel, O., Porikli, F., and Meer, P.: Region Covariance: A Fast Descriptor
for Detection and Classification, in: Proceedings of the 9th European
Conference on Computer Vision, Berlin, Heidelberg, 7–13 May 2006, 589–600,
2006.
Wang, R., Guo, H., Davis, L. S., and Dai, Q.: Covariance discriminative
learning: A natural and efficient approach to image set classification, in:
Proceedings of 2012 IEEE Conference on Computer Vision and Pattern
Recognition, Providence, RI, USA, 16–21 June 2012, 2496–2503, 2012.
Wang, W., Wang, R., Huang, Z., Shan, S., and Chen, X.: Discriminant Analysis
on Riemannian Manifold of Gaussian Distributions for Face Recognition with
Image Sets, IEEE T. Image Process., 27, 151–163,
https://doi.org/10.1109/TIP.2017.2746993, 2018.
Wang, Y., Shi, C., Wang, C., and Xiao, B.: Ground-based cloud classification
by learning stable local binary patterns, Atmos. Res., 207, 74–89,
https://doi.org/10.1016/j.atmosres.2018.02.023, 2018.
Xiao, Y., Cao, Z., Zhuo, W., Ye, L., and Zhu, L.: mCLOUD: A Multi-view
Visual Feature Extraction Mechanism for Ground-based Cloud Image
Categorization, J. Atmos. Ocean. Technol., 33, 789,
https://doi.org/10.1175/JTECH-D-15-0015.1, 2016.
Ye, L., Cao, Z., and Xiao, Y.: DeepCloud: Ground-Based Cloud Image
Categorization Using Deep Convolutional Features, IEEE T. Geosci. Remote
S., 55, 5729–5740, https://doi.org/10.1109/TGRS.2017.2712809, 2017.
Zhang, Z., Li, D., and Liu, S.: Salient Dual Activations Aggregation for
Ground-Based Cloud Classification in Weather Station Networks, IEEE Access,
6, 59173–59181, https://doi.org/10.1109/ACCESS.2018.2874994, 2018.
Zhuo, W., Cao, Z.-G., and Xiao, Y.: Cloud Classification of Ground-Based
Images Using Texture–Structure Features, J. Atmos. Ocean. Technol., 31,
79–92, https://doi.org/10.1175/JTECH-D-13-00048.1, 2014.
Short summary
An automatic cloud classification method on whole-sky images is presented. We first extract multiple pixel-level features to form region covariance descriptors (RCovDs) and then encode RCovDs by the Riemannian bag-of-feature (BoF) method to output the histogram representation. Reults show that a very high prediction accuracy can be obtained with a small number of training samples, which validate the proposed method and exhibit the competitive performance against state-of-the-art methods.
An automatic cloud classification method on whole-sky images is presented. We first extract...
