SEARCH · Search NASA
Results for “Cloud location, Images of Clouds”
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Automated construction of clear-sky dictionary from all-sky imager data
All-sky imagers (ASIs) have significant promise as scalable sensors for short-term solar irradiance forecasting. Many of the current computational techniques that use ASIs for this purpose rely on collections of clear-sky images indexed by time of day, solar angle, or both, called clear-sky dictionaries (CSDs). These CSDs act as baselines against which images can be compared to locate and classify clouds within the image frame. CSDs are often compiled by hand, where individuals visually inspect collections of images one at a time to find clear-sky images. This process is not scalable, and it is prone to error. This paper proposes an automated, nonparametric alternative that uses the principles of digital image processing to find clear-sky images within a set of images taken over several days. We use ground-truth measurements of the clearness index to assess the performance of our method, and we show that the images it selects accurately correspond to clear-sky images. We also compare our proposal, which is nonparametric, with a state-of-the-art parametric method. The numerical results indicate that the performance of the method proposed here is superior.
Quantitative Ceilometer-Radar Studies of Clouds Field Campaign Report
The Long Island Solar Farm (LISF) is a 32-megawatt solar photovoltaic power plant built through a collaboration including BP Solar, the Long Island Power Authority (LIPA), and the U.S. Department of Energy (DOE). The LISF, located on the Brookhaven National Laboratory site, began delivering power to the LIPA grid in November 2011, and is currently one of the largest solar photovoltaic power plants in the eastern United States. It is generating enough renewable energy to power approximately 4,500 homes and is helping New York State meet its clean energy and carbon reduction goals. Brookhaven National Laboratory (BN)L has the LISF instrumented and is capturing solar insolation and power data that can be used for research purposes. Additional information on LISF is available here: https://www.bnl.gov/lisf/. A network of nine high-definition (HD cameras) and 32 pyranometers are deployed within the LISF for the purpose of monitoring the location and characteristics of clouds and available global horizontal irradiance at the surface across the region. Information collected by this network every 30 s is currently used in BNL’s solar NowCasting algorithm, which forecasts near-term solar energy availability accounting for the behavior of clouds. The most important parameters for forecasting the near-term solar energy availability are accurate estimations of their horizontal extend, horizontal motion, and cloud base height. Currently, images from nearly located cameras are used to estimate the cloud base height. In order to evaluate the potential of this method for estimating the cloud base height, we decided to deploy a Vaisala ceilometer at the LISF.
The Illumination of Thunderclouds by Lightning: 4. Volumetric Thunderstorm Imagery
Optical instruments such as the Geostationary Lightning Mapper (GLM) detect lightning based on transient changes in cloud illumination. The horizontal location of lightning is determined from the coordinates of the pixels on the imaging array illuminated during the flash. However, the vertical position of the lightning pulses (approximated by GLM “groups”) below the cloud top cannot be routinely measured from a single space-based instrument. In our prior work, we have developed a machine learning algorithm that can infer optical source altitude for a given pulse based on how the optical energy is distributed across the group footprint and the local Advanced Baseline Imager Cloud-Top Height (CTH). In this fourth part of our thundercloud illumination study, we leverage these source altitudes to generate volumetric GLM imagery of a Colombia thunderstorm. We find that 3D versions of the current GLM meteorological imagery products (that describe thunderstorm kinematics) and thundercloud imagery products (that depict how the flashes appear from space) provide additional insights into lightning activity in the thunderstorm that are lost in the vertical integration used to generate the current 2D GLM gridded products. This new volumetric imaging capability provides a more comprehensive picture of where lightning occurs in the storm, how its physical characteristics vary across three-dimensional space, and how its optical emissions interact with surrounding the cloud medium.
HAIMOS Ensemble Forecasts for Intra-day and Day- Ahead GHI, DNI and Ramps
The objective of this research is to develop a hybrid physics-based/data-driven forecast model to improve direct normal and global horizontal irradiance (DNI and GHI) prediction for horizons ranging from 1 to 72 hours. Project objectives also address key gaps in state-of-the-art solar forecasting: accurate probabilistic solar forecasts and the forecasting of large irradiance ramps (ramp onset and magnitude). The proposed model ensembles Numerical Weather Prediction (NWP) forecasts, determinist physics-based algorithms, and new-generation cloud cover products (high-resolution rapid refresh satellite images and Large Eddy Simulations). The result is the Hybrid Adaptive Input Model Objective Selection (HAIMOS) ensemble model. HAIMOS blends state of the art machine learning methodologies with physics-based models for cloud cover and cloud optical depth forecasts. The technical activities followed a two-pronged strategy. First, the preprocessing of data, the selection of inputs to the nonlinear approximators, the type of approximator and objective functions, and post-processing ensembling techniques included in HAIMOS were all optimized adaptively to find the best model for a specific goal (reduce DNI/GHI forecast error, improve the prediction of ramp onset, etc.). Second, a large effort was put in improving cloud identification and the forecast of cloud cover and cloud optical depth. To this end, new-generation cloud parametrization products were developed in this work. These include improved algorithms to assist in cloud identification, cloud classification and cloud parametrization from satellite images – three key factors in the accuracy of 1 to 6-hours irradiance forecasts and prediction of ramp onset. Furthermore, we also included cloud information extracted from high resolution rapid refresh satellite images (GOES-16) and Large Eddy Simulations (LES). LES was used to model the atmosphere in detail over locations of interest and produce cloud optical depth forecasts. Once these data streams were validated, they were used as input data to the HAIMOS forecast. The model was developed using data from several climatologically distinct locations with potential for high solar penetration. In the last year of the project, we conducted a validation campaign according to the guidelines stipulated by the Topic Area 1 project as described in the FOA. This effort brings, for the first time, proven machine-learning methodologies for generating state-of-the-art solar forecasts interweaved with detailed physics-based models for cloud detection, and cloud optical depth forecasts. HAIMOS will generate accurate irradiance probabilistic forecast to assist in reducing solar generation prediction error. Globally optimized solar forecast models are more likely to impact solar energy stakeholders. The goal of this project was to increase the state-of-the-art forecast skill from their present values of 10 to 35%. At the end of the project, we achieved between 30% and 50% forecast skill across a wide range of horizons for both GHI and DNI.
Demonstration of object location, classification, and characterization by developed deep learning dust ablation trail analysis code package using plasma jets
Based on deep learning, a Dust Ablation Trail Analysis (DATA) code package is developed to detect dust ablation trails in tokamaks, which is intended to analyze a large amount data of tokamak dusts. To validate and benchmark the DATA code package, 2440 plasma jet images are exploited for the training and test of the deep learning DATA code package, since plasma jets resemble the shape and size of dust ablation clouds in tokamaks. After being trained by 1920 plasma jet images, the DATA code package is able to locate 100% plasma jets, classify plasma jets with the accuracy of >99.9%, and output image skeleton information for classified plasma jets. The DATA code package trained by the plasma jet images is also used to analyze the dust ablation trails captured in the Experimental Advanced Superconducting (EAST) tokamak with the satisfactory performance, further verifying its applicability in the fusion dust ablation investigation. Based on its excellent performance presented here, it is demonstrated that our DATA code package is able to automatically identify and analyze dust ablation trails in tokamaks, which can be used for further detailed investigations, such as the three-dimensional reconstruction of dusts and their ablation trails.
The Illumination of Thunderclouds by Lightning: 3. Retrieving Optical Source Altitude
Abstract Optical space‐based lightning sensors such as the Geostationary Lightning Mapper (GLM) detect and geolocate lightning by recording rapid changes in cloud top illumination. While lightning locations can be determined to within a pixel on the GLM imaging array, these instruments are not individually able to natively report lightning altitude. It has previously been shown that thunderclouds are illuminated differently based on the altitude of the optical source. In this study, we examine how altitude information can be extracted from the spatial distributions of GLM energy recorded from each optical pulse. We match GLM “groups” with Lightning Mapping Array (LMA) source data that accurately report the 3‐D positions of coincident Radio‐Frequency (RF) emitters. We then use machine learning methods to predict the mean LMA source altitudes matched to GLM groups using metrics from the optical data that describe the amplitude, breadth, and texture of the group spatial energy distribution. The resulting model can predict the LMA mean source altitude from GLM group data with a median absolute error of <1.5 km, which is sufficient to determine the location of the charge layer where the optical energy originated. This model is able to capture changes to the source altitude distribution in response to convective processes in the thunderstorm, and the GLM predictions can reveal the vertical structure of individual flashes ‐ enabling 3‐D flash geolocation with GLM for the first time. Future work will account for differences in thunderstorm charge/precipitation structures and viewing angle across the GLM Field of View.
AEPF: Attention-Enabled Point Fusion for 3D Object Detection
Current state-of-the-art (SOTA) LiDAR-only detectors perform well for 3D object detection tasks, but point cloud data are typically sparse and lacks semantic information. Detailed semantic information obtained from camera images can be added with existing LiDAR-based detectors to create a robust 3D detection pipeline. With two different data types, a major challenge in developing multi-modal sensor fusion networks is to achieve effective data fusion while managing computational resources. With separate 2D and 3D feature extraction backbones, feature fusion can become more challenging as these modes generate different gradients, leading to gradient conflicts and suboptimal convergence during network optimization. To this end, we propose a 3D object detection method, Attention-Enabled Point Fusion (AEPF). AEPF uses images and voxelized point cloud data as inputs and estimates the 3D bounding boxes of object locations as outputs. An attention mechanism is introduced to an existing feature fusion strategy to improve 3D detection accuracy and two variants are proposed. These two variants, AEPF-Small and AEPF-Large, address different needs. AEPF-Small, with a lightweight attention module and fewer parameters, offers fast inference. AEPF-Large, with a more complex attention module and increased parameters, provides higher accuracy than baseline models. Experimental results on the KITTI validation set show that AEPF-Small maintains SOTA 3D detection accuracy while inferencing at higher speeds. AEPF-Large achieves mean average precision scores of 91.13, 79.06, and 76.15 for the car class’s easy, medium, and hard targets, respectively, in the KITTI validation set. Results from ablation experiments are also presented to support the choice of model architecture.
Principles of stereo reconstruction of aerial objects using stationary cameras
An overview is given here of the principles and mathematics of stereo reconstruction of objects in the sky using stationary cameras with an emphasis on meteorological applications. Through its Atmospheric Radiation Measurement program, the Department of Energy has operated stereo-photogrammetric cameras since 2017 as part of an effort to measure the life-cycle properties of clouds. At the core of that technology is stereo reconstruction, which calculates the real-world position of an object from the location of the object’s image in two cameras’ photographs. Here, stereo reconstruction is stripped down to its basic elements and presented using conventions tailored to applications in atmospheric science. In addition, the resulting equations are used to illustrate the high sensitivity of reconstructed cloud positions to errors in the cameras’ Euler angles. The interested reader will find here a self-contained guide to performing stereo reconstructions using distortion-corrected images from a pair of calibrated, stationary cameras, as well as a demonstration of the need for high accuracy in the measurement of camera properties and orientations.
Geometric Interpretation of the Cluster Location Problem Part II: Application to the Pahala, Hawaii, Earthquake Sequence
In the companion “Theory” article, we presented a new framing of the seismic location problem in terms of differential geometry (Harris et al., 2025). From that viewpoint, we developed a “project and correct” approach for estimating the relative locations of earthquakes. Here, in this study, we use project and correct to estimate high-precision relative locations of events from an earthquake sequence beneath the town of Pahala, Hawaii, using high-precision correlation-derived picks. The sequence was active from 2020 through 2022 and produced many highly correlated signals at Hawaii Volcano Observatory (HVO) stations on the island of Hawaii. The data we inverted consisted of 2882 events with observations at 5 HVO stations. For comparison with the travel-time image, we also produced conventional hypocenter solutions using both the Bayesloc program (Myers et al., 2007, 2009) and a purpose-built double-difference code. There were obvious structural elements in the resulting image, the resolution of which we used to test the performance of the project and the correct algorithm. For the projection step, we first produced a 3D local basis using an singular value decomposition (SVD) of the 2882 groups of times. Projection of the travel-time vectors into this basis resulted in an image with structures similar to those produced by our conventional locators, but with distortion as predicted by theory. Removing the distortion requires an inverse operator generated from the metric tensor at the geometric centroid of the events. We compared two approaches to obtaining such an inverse operator. The first uses an estimate of the geographic centroid of the event cloud from the centroid of the travel-time data. The second approach uses the centroid of the conventionally produced locations. The first approach produces a corrected image very similar to the conventional results, but with a rotation. The corrected image produced using the conventionally derived centroid is a near-exact match to the conventional locations.
High-Volume Precipitation Spectrometer (HVPS) Instrument Handbook
The high-volume precipitation spectrometer (HVPS) is a U.S. Department of Energy (DOE) Atmospheric Radiation Measurement (ARM) user facility instrument for measuring hydrometeors larger than 150 μm. It is an open-path instrument located on the outside of the aircraft, typically below the wing. A laser sheet throws a shadowgraph image of the hydrometeor (cloud droplets or ice crystals) in the sample volume on a photodiode array. As the hydrometeor moves through the sample volume, this optical array records the moving shape and a two-dimensional picture can be recorded. At an air speed of 100 m/s, the sample volume is 310 l/s. The recorded images can be analyzed for size distributions, number concentrations, and further cloud properties detailed below. Hydrometeors in the range of 0.3-19.2 mm can be fully recorded at a 150 μm resolution; larger particles can only be sized along the direction of flight.
Holes in Optical Lightning Flashes: Identifying Poorly Transmissive Clouds in Lightning Imager Data
Space-based optical lightning sensors including the lightning imaging sensor (LIS) and geostationary lightning mapper (GLM) are pixelated imagers that detect lightning as transient increases in cloud top illumination. Detection requires optical emissions to escape the cloud top to space with sufficient energy to trigger a pixel on the imaging array. Through scattering and absorption, certain clouds are able to block most light from reaching the instrument, causing a reduction in detection efficiency (DE) and possibly location accuracy (LA). Radiant lightning emissions that illuminate large cloud top areas are used to examine scenarios where clouds block light from reaching orbit. In some cases, these anomalies in the spatial radiance distribution from the lightning pulse lead to “holes” in the optical lightning flash where certain pixels fail to trigger. Such holes are identified algorithmically in the Tropical Rainfall Measuring Mission satellite LIS record and the microphysical properties of the coincident storm region are queried. We find that holes primarily occur in tall (IR T b < 235 K) convection (87%) and overhanging anvil clouds (10%). The remaining 3% of holes occur in moderate-to-weak convection or in clear air breaks between stormclouds. We further demonstrate how an algorithm that assesses the spatial radiance patterns from energetic lightning pulses might be used to construct an optical transmission gridded stoplight product for GLM that could help operators identify clouds with a potentially reduced DE and LA.
Influence of covariance of aerosol and meteorology on co-located precipitating and non-precipitating clouds over the Indo-Gangetic Plain
Aerosol–cloud–precipitation interaction (ACPI) plays a pivotal role in the global and regional water cycle and the earth's energy budget; however, it remains highly uncertain due to the different underlying physical mechanisms. Therefore, this study aims to systematically analyze the effects of aerosols and meteorological factors on ACPI in the co-located precipitating clouds (PCs) and non-precipitating clouds (NPCs) in winter and summer seasons by employing the long-term (2001–2021) retrievals from the Moderate Resolution Imaging Spectroradiometer (MODIS) and Tropical Rainfall Measuring Mission (TRMM) coupled with the National Center for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) reanalysis-II datasets over the Indo-Gangetic Plain (IGP). The results exhibit a decadal increase in aerosol optical depth (AOD) over Lahore (5.2 %), Delhi (9 %), Kanpur (10.7 %), and Gandhi College (22.7 %) and a decrease over Karachi (-1.9 %) and Jaipur (-0.5 %). The most stable meteorology with high values of lower-tropospheric stability (LTS) is found in both seasons over Karachi. In the summer season, the occurrence frequency of clouds is high (74 %) over Gandhi College, 60 % of which are PCs. Conversely, the lowest number of PCs is found over Karachi. Similarly, in the winter season, the frequency of cloud occurrence is low over Karachi and high over Lahore and Gandhi College. The analysis of cloud top pressure (CTP) and cloud optical thickness (COT) indicates high values of cloud fraction (CF) for thick and high-level clouds over all study areas except Karachi. The microphysical properties such as cloud effective radius (CER) and cloud droplet number concentration (CDNC) bear high values (CER ≳ 15 µm and CDNC ≳ 50 cm -3 ) for both NPCs and PCs in summer. The AOD–CER correlation is good (weak) for PCs (NPCs) in winter. Similarly, the sensitivity value of the first indirect effect (FIE) is high (ranging from 0.2 ± 0.13 to 0.3 ± 0.01 in winter and from 0.19 ± 0.03 to 0.32 ± 0.05 in summer) for PCs and low for NPCs. The sensitivity value for the second indirect effect (SIE) is relatively higher (such as 0.6 ± 0.14 in winter and 0.4 ± 0.04 in summer) than FIE. Sensitivity values of the aerosol-cloud interaction (ACI) are low (i.e., -0.06 ± 0.09) for PCs in summer. Furthermore, the precipitation rate (PR) exhibits high values in the summer season, primarily due to the significant contribution from optically thick clouds with lower CDNC (≲ 50 cm -3 ) and larger CER and the intermediate contribution from optically thick clouds with higher CDNC ( ≳ 50 cm -3 ).
Landscape-scale Characterization of Arctic Tundra Vegetation Composition, Structure, and Function with a Multi-sensor Unoccupied Aerial System: Supporting Data.
Airborne remote sensing data collected using the Brookhaven National Laboratory's (BNL) heavy-lift unoccupied aerial system (UAS) octocopter platform, the Osprey, operated by the Terrestrial Ecosystem Science and Technology (TEST) group (https://www.bnl.gov/testgroup). This package includes data from three flights flown over the NGEE-Arctic Council, Kougarok and Teller sites in July, 2018. The Osprey is a multi-sensor UAS platform that simultaneously measures very high spatial resolution optical red/green/blue (RGB) and thermal infrared (TIR) surface 'skin' temperature imagery, as well as surface reflectance at 1 nm intervals in the visible to near-infrared spectral range 350 - 1000 nm measured at regular intervals along each flight path. Derived image products include ortho-mosaiced RGB and TIR images, an RGB-based digital surface model (DSM) using the structure from motion (SfM) technique, digital terrain model (DTM), and a canopy height model (CHM). In addition, a VNIR surface reflectance file is provided for the trigger locations collected during each flight campaign. Ancillary aircraft data, flight mission parameters, and general flight conditions provided by the onboard flight and data collection computers are also included. Unprocessed and processed data products are included in this package (processing levels 0-3). Data and metadata are provided as text (*.txt, *.json, *.kml, *hdr, *.enp), tabular (*.dat, *.csv, *.waypoint, ENVI format (no extension)), point cloud (*.laz) and image (*.jpg, *.tif, *png) formats. This metadata document contains flight campaign, instrument and file metadata, along with a description of data processing levels, data products and file naming scheme.The Next-Generation Ecosystem Experiments: Arctic (NGEE Arctic), a research effort to reduce uncertainty in Earth System Models by developing a predictive understanding of carbon-rich Arctic ecosystems and feedbacks to climate. NGEE Arctic was supported by the Department of Energy?s Office of Biological and Environmental Research.The NGEE Arctic project had two field research sites: 1) located within the Arctic polygonal tundra coastal region on the Barrow Environmental Observatory (BEO) and the North Slope near Utqiagvik (Barrow), Alaska and 2) multiple areas on the discontinuous permafrost region of the Seward Peninsula north of Nome, Alaska.Through observations, experiments, and synthesis with existing datasets, NGEE Arctic provided an enhanced knowledge base for multi-scale modeling and contributed to improved process representation at global pan-Arctic scales within the Department of Energy's Earth system Model (the Energy Exascale Earth System Model, or E3SM), and specifically within the E3SM Land Model component (ELM).
Impact of Aerosol and Cloud on the Solar Energy Potential over the Central Gangetic Himalayan Region
We examine the impact of atmospheric aerosols and clouds on the surface solar radiation and solar energy at Nainital, a high-altitude remote location in the central Gangetic Himalayan region (CGHR). For this purpose, we exploited the synergy of remote-sensed data in terms of ground-based AERONET Sun Photometer and satellite observations from the MODerate Resolution Imaging Spectroradiometer (MODIS) and the Meteosat Second Generation (MSG), with radiative transfer model (RTM) simulations and 1 day forecasts from the Copernicus Atmosphere Monitoring Service (CAMS). Clouds and aerosols are one of the most common sources of solar irradiance attenuation and hence causing performance issues in the photovoltaic (PV) and concentrated solar power (CSP) plant installations. The outputs of RTM results presented with high accuracy under clear, cloudy sky and dust conditions for global horizontal (GHI) and beam horizontal irradiance (BHI). On an annual basis the total aerosol attenuation was found to be up to 105 kWh m –2 for the GHI and 266 kWh m –2 for BHI, respectively, while the cloud effect is much stronger with an attenuation of 245 and 271 kWh m –2 on GHI and BHI. The results of this study will support the Indian solar energy producers and electricity handling entities in order to quantify the energy and financial losses due to cloud and aerosol presence.
Evolution of titanium particle combustion in potassium perchlorate and air
Understanding titanium particle combustion processes is critical not only for characterizing existing pyrotechnic systems but also for creating new igniter designs. In order to characterize titanium particle combustion processes, morphologies, and temperatures, simultaneous spatially-resolved electric field holography and imaging pyrometry techniques were used to capture post-ignition data at up to 7 kHz. Due to the phase and thermal distortions present in the combustion cloud, traditional digital in-line holography techniques fail to capture accurate data. Here in this work, electric field holography techniques are used in order to cancel distortions and capture the three-dimensional spatial locations and diameters of the particles. In order to estimate the projected surface temperatures of the titanium particles, an imaging pyrometry method that ratios emission at 750 and 850 nm is utilized. Using these diagnostics, joint statistics are collected for particle size, morphology, velocity, and temperature. Results show that, early in the combustion process, the titanium particles are primarily oxidized by potassium perchlorate inside the igniter cup, resulting in projected surface temperatures near 3000 K. Later in the process, the particles interact with ambient air, resulting in lower surface temperatures around 2400 K and the formation of flame zones. These results are consistent with adiabatic flame temperature predictions as well as particle morphology observations of a titanium core with a TiO 2 surface. Late stage particle expansion, star fragmentation, and molten droplet breakup events are also observed using the time-resolved morphology and temperature diagnostics. These results illustrate the different stages of titanium particle combustion in pyrotechnic environments, which can be used to inform improvements in next-generation igniters.
Panel-Segmentation [SWR-21-18]
Panel-Segmentation contains the scripts for automated metadata extraction of solar PV installations, using satellite imagery coupled with computer vision techniques. In this package, the user can perform the following actions: *Automatically generate a satellite image using a set of lat-long coordinates, and a Google Maps API key. Users would need to set up a Google Cloud account and get a Maps Static API key. Please refer to Setting Up Google Maps Static API Key section for this process. *Perform image segmentation on the satellite image, to locate the solar array(s) in the image on a pixel-by-pixel basis, using an image segmentation model (panel_detection_model.pth). Get classification of the installation (rooftop, ground mounted fixed-tilt or tracking, carport, etc). *Perform azimuth estimation on each solar array cluster in the masked image. *Detect solar panels and get its latitude, longitude, and address within a geographic bounding box through the SOL-Searcher Pipeline. *Detect and calculate hurricane damage on solar installations given pre-hurricane and post-hurricane satellite imagery through the Hurricane Detection Pipeline. *Detect and calculate hail damage on solar installations given satellite imagery through the Hail Detection pipeline. *Convert NOAA MESH (Maximum Estimated Size of Hail) grib2 files into kml or geojson files. *Estimate tilt and azimuth of a solar array by processing USGS LiDAR data for the array’s location.
AICCA: AI-Driven Cloud Classification Atlas
Clouds play an important role in the Earth’s energy budget, and their behavior is one of the largest uncertainties in future climate projections. Satellite observations should help in understanding cloud responses, but decades and petabytes of multispectral cloud imagery have to date received only limited use. This study describes a new analysis approach that reduces the dimensionality of satellite cloud observations by grouping them via a novel automated, unsupervised cloud classification technique based on a convolutional autoencoder, an artificial intelligence (AI) method good at identifying patterns in spatial data. Our technique combines a rotation-invariant autoencoder and hierarchical agglomerative clustering to generate cloud clusters that capture meaningful distinctions among cloud textures, using only raw multispectral imagery as input. Cloud classes are therefore defined based on spectral properties and spatial textures without reliance on location, time/season, derived physical properties, or pre-designated class definitions. We use this approach to generate a unique new cloud dataset, the AI-driven cloud classification atlas (AICCA), which clusters 22 years of ocean images from the Moderate Resolution Imaging Spectroradiometer (MODIS) on NASA’s Aqua and Terra instruments—198 million patches, each roughly 100 km × 100 km (128 × 128 pixels)—into 42 AI-generated cloud classes, a number determined via a newly-developed stability protocol that we use to maximize richness of information while ensuring stable groupings of patches. AICCA thereby translates 801 TB of satellite images into 54.2 GB of class labels and cloud top and optical properties, a reduction by a factor of 15,000. The 42 AICCA classes produce meaningful spatio-temporal and physical distinctions and capture a greater variety of cloud types than do the nine International Satellite Cloud Climatology Project (ISCCP) categories—for example, multiple textures in the stratocumulus decks along the West coasts of North and South America. We conclude that our methodology has explanatory power, capturing regionally unique cloud classes and providing rich but tractable information for global analysis. AICCA delivers the information from multi-spectral images in a compact form, enables data-driven diagnosis of patterns of cloud organization, provides insight into cloud evolution on timescales of hours to decades, and helps democratize climate research by facilitating access to core data.