Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Convolutional neural networks”

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.

At least 1,009 records · Page 56
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Improve the Search of Very Metal-poor Stars Using the Deep Learning Method

Very metal-poor (VMP) stars have [Fe/H] < -2.0 dex. They are among the oldest stars in the universe, and their unique metallicity can help explore the enrichment mechanism and evolutionary history of the chemical elements of stars in the early universe. However, most current stellar parameter estimation methods do not perform well in determining the stellar parameters of VMP stars, which limits our ability to discover and exploit the properties of VMP stars. In this study, we propose a new model based on a convolutional neural network to determine the stellar atmospheric parameters of VMP stars. We tested our model on the LAMOST spectra; our model can determine the effective temperature (T {sub eff}), surface gravity (log g), metallicity ([Fe/H]), and carbon abundance ([C/Fe]) of the LAMOST spectra with precisions σ(T {sub eff}) = 134.82 K, σ(log g) = 0.33 dex, σ ([Fe/H]) = 0.20 dex, and σ([C/Fe]) = 0.35 dex. Furthermore, our model can distinguish VMP stars from normal stars with an accuracy of 88.65%. We also compared this model with other widely used methods, and found that this method performs better than other methods. It can be applied to the stellar parameter pipelines of upcoming large surveys such as 4MOST, WEAVES, and MOONS to search for VMP stars and identify carbon-enhanced metal-poor stars.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Star–Galaxy Image Separation with Computationally Efficient Gaussian Process Classification

Abstract We introduce a novel method for discerning optical telescope images of stars from those of galaxies using Gaussian processes (GPs). Although applications of GPs often struggle in high-dimensional data modalities such as optical image classification, we show that a low-dimensional embedding of images into a metric space defined by the principal components of the data suffices to produce high-quality predictions from real large-scale survey data. We develop a novel method of GP classification hyperparameter training that scales approximately linearly in the number of image observations, which allows for application of GP models to large-size Hyper Suprime-Cam Subaru Strategic Program data. In our experiments, we evaluate the performance of a principal component analysis embedded GP predictive model against other machine-learning algorithms, including a convolutional neural network and an image photometric morphology discriminator. Our analysis shows that our methods compare favorably with current methods in optical image classification while producing posterior distributions from the GP regression that can be used to quantify object classification uncertainty. We further describe how classification uncertainty can be used to efficiently parse large-scale survey imaging data to produce high-confidence object catalogs.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2020

A Hybrid Deep Learning Approach to Cosmological Constraints from Galaxy Redshift Surveys

We present a deep machine learning (ML)–based technique for accurately determining σ g and Ω m from mock 3D galaxy surveys. The mock surveys are built from the AbacusCosmos suite of N -body simulations, which comprises 40 cosmological volume simulations spanning a range of cosmological parameter values, and we account for uncertainties in galaxy formation scenarios through the use of generalized halo occupation distributions (HODs). We explore a trio of ML models: a 3D convolutional neural network (CNN), a power spectrum–based fully connected network, and a hybrid approach that merges the two to combine physically motivated summary statistics with flexible CNNs. We describe best practices for training a deep model on a suite of matched-phase simulations, and we test our model on a completely independent sample that uses previously unseen initial conditions, cosmological parameters, and HOD parameters. Despite the fact that the mock observations are quite small (~0.07 h -3 Gpc 3 ) and the training data span a large parameter space (six cosmological and six HOD parameters), the CNN and hybrid CNN can constrain estimates of σ g and Ω m to ~3% and ~4%, respectively.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

Gaussian Process Classification for Galaxy Blend Identification in LSST

Abstract A significant fraction of observed galaxies in the Rubin Observatory Legacy Survey of Space and Time (LSST) will overlap at least one other galaxy along the same line of sight, in a so-called “blend.” The current standard method of assessing blend likelihood in LSST images relies on counting up the number of intensity peaks in the smoothed image of a blend candidate, but the reliability of this procedure has not yet been comprehensively studied. Here we construct a realistic distribution of blended and unblended galaxies through high-fidelity simulations of LSST-like images, and from this we examine the blend classification accuracy of the standard peak-finding method. Furthermore, we develop a novel Gaussian process blend classifier model, and show that this classifier is competitive with both the peak finding method as well as with a convolutional neural network model. Finally, whereas the peak-finding method does not naturally assign probabilities to its classification estimates, the Gaussian process model does, and we show that the Gaussian process classification probabilities are generally reliable.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

Physics-informed Machine Learning for Modeling Turbulence in Supernovae

Abstract Turbulence plays an important role in astrophysical phenomena, including core-collapse supernovae (CCSNe), but current simulations must rely on subgrid models, since direct numerical simulation is too expensive. Unfortunately, existing subgrid models are not sufficiently accurate. Recently, machine learning (ML) has shown an impressive predictive capability for calculating turbulence closure. We have developed a physics-informed convolutional neural network to preserve the realizability condition of the Reynolds stress that is necessary for accurate turbulent pressure prediction. The applicability of the ML subgrid model is tested here for magnetohydrodynamic turbulence in both the stationary and dynamic regimes. Our future goal is to utilize this ML methodology (available on GitHub) in the CCSN framework to investigate the effects of accurately modeled turbulence on the explosion of these stars.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

A Machine Learning–Based Approach to Time-series Wave Identification in the Solar Wind

The Wind spacecraft has yielded several decades of high-resolution magnetic field data, a large fraction of which displays small-scale structures. In particular, the solar wind is full of wavelike fluctuations that appear in both the field magnitude and its components. The nature of these fluctuations can be tied to the properties of other structures in the solar wind, such as shocks, that have implications for the time evolution of the solar wind. As such, having a large collection of wave events would facilitate further study of the effects that these fluctuations have on solar wind evolution. Given the large volume of magnetic field data available, machine learning is the most practical approach to classifying the myriad small-scale structures observed. To this end, a subset of Wind data is labeled and used as a training set for a multibranch 1D convolutional neural network aimed at classifying circularly polarized wave modes. Using this algorithm, a preliminary statistical study of 1 yr of data is performed, yielding about 300,000 wave intervals out of about 5,000,000 solar wind intervals. The wave intervals come about more often in the fast solar wind and at higher temperatures, and the number of waves per day is highly periodic. This machine learning–based approach to wave detection has the potential to be a powerful, inexpensive way to catalog waves throughout decades of spacecraft data.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Identification of Galaxy–Galaxy Strong Lens Candidates in the DECam Local Volume Exploration Survey Using Machine Learning

We perform a search for galaxy–galaxy strong lens systems using a convolutional neural network (CNN) applied to imaging data from the first public data release of the DECam Local Volume Exploration Survey, which contains ~520 million astronomical sources covering ~4000 deg 2 of the southern sky to a 5σ point–source depth of g = 24.3, r= 23.9, i = 23.3, and z = 22.8 mag. Following the methodology of similar searches using Dark Energy Camera data, we apply color and magnitude cuts to select a catalog of ~11 million extended astronomical sources. After scoring with our CNN, the highest-scoring 50,000 images were visually inspected and assigned a score on a scale from 0 (not a lens) to 3 (very probable lens). We present a list of 581 strong lens candidates, 562 of which are previously unreported. We categorize our candidates using their human-assigned scores, resulting in 55 Grade A candidates, 149 Grade B candidates, and 377 Grade C candidates. We additionally highlight eight potential quadruply lensed quasars from this sample. Due to the location of our search footprint in the northern Galactic cap (b > 10 deg) and southern celestial hemisphere (decl. < 0 deg), our candidate list has little overlap with other existing ground-based searches. Where our search footprint does overlap with other searches, we find a significant number of high-quality candidates that were previously unidentified, indicating a degree of orthogonality in our methodology. We report properties of our candidates including apparent magnitude and Einstein radius estimated from the image separation.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Target Selection and Sample Characterization for the DESI LOW-Z Secondary Target Program

Abstract We introduce the DESI LOW- Z Secondary Target Survey, which combines the wide-area capabilities of the Dark Energy Spectroscopic Instrument (DESI) with an efficient, low-redshift target selection method. Our selection consists of a set of color and surface brightness cuts, combined with modern machine-learning methods, to target low-redshift dwarf galaxies ( z < 0.03) between 19 < r < 21 with high completeness. We employ a convolutional neural network (CNN) to select high-priority targets. The LOW- Z survey has already obtained over 22,000 redshifts of dwarf galaxies ( M * < 10 9 M ⊙ ), comparable to the number of dwarf galaxies discovered in the Sloan Digital Sky Survey DR8 and GAMA. As a spare fiber survey, LOW- Z currently receives fiber allocation for just ∼50% of its targets. However, we estimate that our selection is highly complete: for galaxies at z < 0.03 within our magnitude limits, we achieve better than 95% completeness with ∼1% efficiency using catalog-level photometric cuts. We also demonstrate that our CNN selections z < 0.03 galaxies from the photometric cuts subsample at least 10 times more efficiently while maintaining high completeness. The full 5 yr DESI program will expand the LOW- Z sample, densely mapping the low-redshift Universe, providing an unprecedented sample of dwarf galaxies, and providing critical information about how to pursue effective and efficient low-redshift surveys.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2024

The Zwicky Transient Facility Bright Transient Survey. III. BTSbot: Automated Identification and Follow-up of Bright Transients with Deep Learning

Abstract The Bright Transient Survey (BTS) aims to obtain a classification spectrum for all bright ( m peak ≤ 18.5 mag) extragalactic transients found in the Zwicky Transient Facility (ZTF) public survey. BTS critically relies on visual inspection (“scanning”) to select targets for spectroscopic follow-up, which, while effective, has required a significant time investment over the past ∼5 yr of ZTF operations. We present BTSbot , a multimodal convolutional neural network, which provides a bright transient score to individual ZTF detections using their image data and 25 extracted features. BTSbot is able to eliminate the need for daily human scanning by automatically identifying and requesting spectroscopic follow-up observations of new bright transient candidates. BTSbot recovers all bright transients in our test split and performs on par with scanners in terms of identification speed (on average, ∼1 hr quicker than scanners). We also find that BTSbot is not significantly impacted by any data shift by comparing performance across a concealed test split and a sample of very recent BTS candidates. BTSbot has been integrated into Fritz and Kowalski , ZTF’s first-party marshal and alert broker, and now sends automatic spectroscopic follow-up requests for the new transients it identifies. Between 2023 December and 2024 May, BTSbot selected 609 sources in real time, 96% of which were real extragalactic transients. With BTSbot and other automation tools, the BTS workflow has produced the first fully automatic end-to-end discovery and classification of a transient, representing a significant reduction in the human time needed to scan.

Rehemtulla, Nabeel (ORCID:0000000256832389)↗
DOE OSTI2025

Introducing the DREAMS Project: DaRk mattEr and Astrophysics with Machine Learning and Simulations

We introduce the DaRk mattEr and Astrophysics with Machine learning and Simulations (DREAMS) project, an innovative approach to understanding the astrophysical implications of alternative dark matter (DM) models and their effects on galaxy formation and evolution. The DREAMS project will ultimately comprise thousands of cosmological hydrodynamic simulations that simultaneously vary over DM physics, astrophysics, and cosmology in modeling a range of systems—from galaxy clusters to ultra-faint satellites. Such extensive simulation suites can provide adequate training sets for machine-learning-based analyses. This paper introduces two new cosmological hydrodynamical suites of warm dark matter (WDM), each comprising 1024 simulations generated using the arepo code. One suite consists of uniform-box simulations covering a ${(25\,{h}^{-1}\,{\rm{Mpc}})}^{3}$ volume, while the other consists of Milky Way zoom-ins with sufficient resolution to capture the properties of classical satellites. For each simulation, the WDM particle mass is varied along with the initial density field and several parameters controlling the strength of baryonic feedback within the IllustrisTNG model. We provide two examples, separately utilizing emulators and convolutional neural networks, to demonstrate how such simulation suites can be used to disentangle the effects of DM and baryonic physics on galactic properties. The DREAMS project can be extended further to include different DM models, galaxy formation physics, and astrophysical targets. In this way, it will provide an unparalleled opportunity to characterize uncertainties on predictions for small-scale observables, leading to robust predictions for testing the particle physics nature of DM on these scales.

cosmological parameters↗
DOE OSTI2026

Gravitational Lenses in UNIONS and Euclid (GLUE). I. A Search for Strong Gravitational Lenses in UNIONS with Subaru, CFHT, and Pan-STARRS Data

We present the results of our pipeline for discovering strong gravitational lenses in the ongoing Ultraviolet Near-Infrared Optical Northern Survey (UNIONS). We successfully train a deep residual convolutional neural network based on CMU-Deeplens architecture, which is designed to detect strong lenses in ground-based imaging surveys. We train on images of real strong lenses and deploy on a sample of 8 million galaxies in areas with full coverage in the g, r, and i filters—the first multiband search for strong gravitational lenses in UNIONS. Following human inspection and grading, we report the discovery of a total of 1346 new strong-lens candidates, of which 146 are Grade A, 199 are Grade B, and 1001 are Grade C. Of these candidates, 283 have lens galaxy spectroscopic redshifts from the Sloan Digital Sky Survey, and an additional 297 have them from the Dark Energy Spectroscopic Instrument Data Release 1. We find 15 of these systems display evidence of both lens and source galaxy redshifts in spectral superposition. We also report the spectroscopic confirmation of seven lensed sources in high-quality systems, all with z > 2.1, using the Keck Near Infrared Echellette Spectrograph and the Gemini Near-Infrared Spectrograph.

Storfer, Christopher J. [University of Hawaii, Hon↗
DOE OSTI2026

Quasars Acting as Strong Lenses Found in DESI DR1

Quasars acting as strong gravitational lenses offer a rare opportunity to probe the redshift evolution of scaling relations between supermassive black holes and their host galaxies, particularly the M$_{BH}$–M$_{host}$ relation. Using these powerful probes, the mass of the host galaxy can be precisely inferred from the Einstein radius θ$_{E}$. Using 812,118 quasars from DESI DR1 (0.03 ≤ z ≤ 1.8), we searched for quasars lensing higher-redshift galaxies by identifying background emission-line features in their spectra. To detect these rare systems, we trained a convolutional neural network (CNN) on mock lenses constructed from real DESI spectra of quasars and emission-line galaxies (ELGs), achieving a high classification performance (AUC = 0.99). We also trained a regression network to estimate the redshift of the background ELG. Applying this pipeline, we identified seven high-quality (Grade A) lens candidates, each exhibiting a strong [O II] doublet at a higher redshift than the foreground quasar; four candidates additionally show Hβ, [O III] λ4959, and [O III] λ5007 emission. These results significantly expand the sample of quasar lens candidates beyond the 12 identified and 3 confirmed in previous work and demonstrate the potential for scalable, data-driven discovery of quasars as strong lenses in upcoming spectroscopic surveys.

McArthur, Everett [Stanford U., Phys. Dept.; KIPAC↗
DOE OSTI2021

Classifying Seyfert Galaxies with Deep Learning

The traditional classification for a subclass of the Seyfert galaxies is visual inspection or using a quantity defined as a flux ratio between the Balmer line and forbidden line. One algorithm of deep learning is the convolution neural network (CNN), which has shown successful classification results. We build a one-dimensional CNN model to distinguish Seyfert 1.9 spectra from Seyfert 2 galaxies. We find that our model can recognize Seyfert 1.9 and Seyfert 2 spectra with an accuracy of over 80% and pick out an additional Seyfert 1.9 sample that was missed by visual inspection. We use the new Seyfert 1.9 sample to improve the performance of our model and obtain a 91% precision of Seyfert 1.9. These results indicate that our model can pick out Seyfert 1.9 spectra among Seyfert 2 spectra. We decompose the Hα emission line of our Seyfert 1.9 galaxies by fitting two Gaussian components and derive the line width and flux. We find that the velocity distribution of the broad Hα component of the new Seyfert 1.9 sample has an extending tail toward the higher end, and the luminosity of the new Seyfert 1.9 sample is slightly weaker than the original Seyfert 1.9 sample. This result indicates that our model can pick out the sources that have a relatively weak broad Hα component. In addition, we check the distributions of the host galaxy morphology of our Seyfert 1.9 samples and find that the distribution of the host galaxy morphology is dominated by a large bulge galaxy. In the end, we present an online catalog of 1297 Seyfert 1.9 galaxies with measurements of the Hα emission line.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

The Completed Sloan Digital Sky Survey IV Extended Baryon Oscillation Spectroscopic Survey: The Damped Lyα Systems Catalog

We present the characteristics of the damped Lyα (DLA) systems found in data release DR16 of the extended Baryon Oscillation Spectroscopic Survey of the Sloan Digital Sky Survey. The DLAs were identified using the convolutional neural network of Parks et al. (2018). A total of 117,458 absorber candidates were found with 2 ≤ $z_{DLA}$ ≤ 5.5 and 19.7 ≤ log(N(H I )/cm -2 ) ≤ 22, including 57,136 DLA candidates with log(N(H I )/cm -2 ) ≥ 20.3. Mock quasar spectra were used to estimate the DLA detection efficiency and the purity of the resulting catalog. Restricting the quasar sample to bright forests, i.e., those with mean forest fluxes $\bar{f_{λ}}$ > 2 × 10 -19 W m -2 nm -1 , the efficiency and purity are greater than 90% for DLAs with column densities in the range 20.1 ≤ log(N(H I )/cm -2 ) ≤ 22.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

Deep Learning of Dark Energy Spectroscopic Instrument Mock Spectra to Find Damped Lyα Systems

We have updated and applied a convolutional neural network (CNN) machine-learning model to discover and characterize damped Ly α systems (DLAs) based on Dark Energy Spectroscopic Instrument (DESI) mock spectra. We have optimized the training process and constructed a CNN model that yields a DLA classification accuracy above 99% for spectra that have signal-to-noise ratios (S/N) above 5 per pixel. The classification accuracy is the rate of correct classifications. This accuracy remains above 97% for lower S/N ≈1 spectra. This CNN model provides estimations for redshift and H i column density with standard deviations of 0.002 and 0.17 dex for spectra with S/N above 3 pixel -1 . Also, this DLA finder is able to identify overlapping DLAs and sub-DLAs. Further, the impact of different DLA catalogs on the measurement of baryon acoustic oscillations (BAO) is investigated. The cosmological fitting parameter result for BAO has less than 0.61% difference compared to analysis of the mock results with perfect knowledge of DLAs. This difference is lower than the statistical error for the first year estimated from the mock spectra: above 1.7%. We also compared the performances of the CNN and Gaussian Process (GP) models. Our improved CNN model has moderately 14% higher purity and 7% higher completeness than an older version of the GP code, for S/N > 3. Both codes provide good DLA redshift estimates, but the GP produces a better column density estimate by 24% less standard deviation. A credible DLA catalog for the DESI main survey can be provided by combining these two algorithms.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2020

Mapping of Benthic Habitats at Marine Renewable Energy Sites Using Multibeam Echosounder and Sediment Profile Imaging Technologies

Objectives/Scope: The goal of this work was to develop a consistent and semi-automated seafloor survey method for generating high-resolution, benthic habitat maps for environmental assessments and monitoring of marine renewable energy sites. Sediment profile and plan view imaging (SPI/PV) technology was combined with multibeam bathymetry and acoustic backscatter methods to demonstrate a rapid, cost-effective benthic mapping protocol. A key technical innovation of this project was the development of an image processing platform that automatically measures key features from the images. Methods: Multibeam echosounder (MBES) acoustic and SPI/PV surveys were conducted at three coastal areas off the U.S. west coast, including the PacWave South energy test site off of Newport, Oregon. Point data on physical and biological sediment conditions obtained from the SPI/PV imagery were used to efficiently ground-truth the high-resolution MBES bathymetry and backscatter mosaics. The SPI camera is an optical corer that obtains an undisturbed 21 by 15 cm, high-resolution, cross-sectional image of the sediment–water interface and upper sediment column. The plan view camera attached to the SPI camera frame captures a downward looking view of the seabed immediately before the SPI image is obtained. As part of this project, we developed a computer vision image processing platform (iSPI) that uses deep convolutional neural networks and other approaches to automatically identify and measure key features in the images, such as grain size and surface relief. Results: Detailed benthic habitat maps were generated using this mapping approach for three different marine settings. The sites mapped included a silt-dominated embayment; a sloping, nearshore, transitional very fine to coarse sand bottom; and a medium, sand-dominated continental shelf marine energy test site. In each case, acoustic and imaging surveys were completed in less than week, and detailed benthic habitat maps were generated within 60 days. The results were placed within the Coastal and Marine Ecological Classification Standard (CMECS) habitat mapping framework, and various combinations of the bathymetry, backscatter, and SPI and PV data were used to generate CMECS component maps. Novel Information: This project developed and demonstrated a repeatable and cost-effective approach for efficiently mapping benthic habitat conditions over broad areas of the seafloor by combining state-of-the-art acoustic and imaging techniques. The primary survey tools used in this project have been used previously in both the offshore renewable (wind) and oil and gas sectors to map and monitor benthic environments. This project’s innovations include 1) the focused use of high-resolution SPI/PV imagery to ground-truth acoustic mosaics, and 2) the development of a computer automated image analysis processing tool that both streamlines and standardizes the generation of data from the imagery and makes the data extraction process more cost-effective and repeatable.

13 HYDRO ENERGY↗
DOE OSTI2026

Advancing Industry 4.0: Multimodal Sensor Fusion for AI-Based Fault Detection in 3D Printing

Additive manufacturing, particularly fused deposition modeling, is transforming modern production by enabling rapid prototyping and complex part fabrication. However, its layer-by-layer process remains vulnerable to faults such as nozzle clogging, filament runout, and layer misalignment, which compromise print quality and reliability. Traditional inspection methods are costly, time-intensive, and often limited to post-process analysis, making them unsuitable for real-time intervention. In this current study, the authors developed a novel, low-cost, and portable faultdetection system that leverages multimodal sensor fusion and artificial intelligence for real-time monitoring in FDM-based 3D printing. The system integrates acoustic, vibration, and thermal sensing into a non-intrusive architecture, capturing complementary data streams that reflect both mechanical and process-related anomalies. Acoustic and thermal sensors operate in a fully contactless manner, while the vibration sensor requires minimal attachment such that it will not interfere with printer hardware, thereby preserving portability and ease of deployment. The multimodal signals are processed into spectrograms and time-frequency features, which are classified using convolutional neural networks for intelligent fault detection. The proposed system advances Industry 4.0 objectives by offering an affordable, scalable, and practical monitoring solution that improves faultdetection accuracy, reduces waste, and supports sustainable, adaptive manufacturing.

42 ENGINEERING↗
DOE OSTI2025

Classifying thermodynamic cloud phase using machine learning models

Vertically resolved thermodynamic cloud-phase classifications are essential for studies of atmospheric cloud and precipitation processes. The Department of Energy (DOE) Atmospheric Radiation Measurement (ARM) Thermodynamic Cloud Phase (THERMOCLDPHASE) value-added product (VAP) uses a multi-sensor approach to classify the thermodynamic cloud phase by combining lidar backscatter and depolarization, radar reflectivity, Doppler velocity, spectral width, microwave-radiometer-derived liquid water path, and radiosonde temperature measurements. The measured pixels are classified as ice, snow, mixed phase, liquid (cloud water), drizzle, rain, and liq_driz (liquid+drizzle). We use this product as the ground truth to train three machine learning (ML) models to predict the thermodynamic cloud phase from multi-sensor remote sensing measurements taken at the ARM North Slope of Alaska (NSA) observatory: a random forest (RF), a multi-layer perceptron (MLP), and a convolutional neural network (CNN) with a U-Net architecture. Evaluations against the outputs of the THERMOCLDPHASE VAP with 1 year of data show that the CNN outperforms the other two models, achieving the highest test accuracy, F1 score, and mean intersection over union (IOU). Analysis of ML confidence scores shows that ice, rain, and snow have higher confidence scores, followed by liquid, while mixed, drizzle, and liq_driz have lower scores. Feature importance analysis reveals that the mean Doppler velocity and vertically resolved temperature are the most influential data streams for ML thermodynamic cloud-phase predictions. Lidar measurements exhibit lower feature importance due to rapid signal attenuation caused by the frequent presence of persistent low-level clouds at the NSA site. The ML models' generalization capacity is further evaluated by applying them at another Arctic ARM site in Norway using data taken during the ARM Cold-Air Outbreaks in the Marine Boundary Layer Experiment (COMBLE) field campaign. The models demonstrated similar performance to that observed at the NSA site. Finally, we evaluate the ML models' response to simulated instrument outages and signal degradation and show that a CNN U-Net model trained with input channel dropouts performs better when input fields are missing.

ARM Aerial Facility↗