Search for: All records

We propose a novel deep learning framework, named SYMHnet, which employs a graph neural network and a bidirectional long short‐term memory network to cooperatively learn patterns from solar wind and interplanetary magnetic field parameters for short‐term forecasts of the SYM‐H index based on 1‐ and 5‐min resolution data. SYMHnet takes, as input, the time series of the parameters' values provided by NASA's Space Science Data Coordinated Archive and predicts, as output, the SYM‐H index value at time pointt + whours for a given time pointtwherewis 1 or 2. By incorporating Bayesian inference into the learning framework, SYMHnet can quantify both aleatoric (data) uncertainty and epistemic (model) uncertainty when predicting future SYM‐H indices. Experimental results show that SYMHnet works well at quiet time and storm time, for both 1‐ and 5‐min resolution data. The results also show that SYMHnet generally performs better than related machine learning methods. For example, SYMHnet achieves a forecast skill score (FSS) of 0.343 compared to the FSS of 0.074 of a recent gradient boosting machine (GBM) method when predicting SYM‐H indices (1 hr in advance) in a large storm (SYM‐H = −393 nT) using 5‐min resolution data. When predicting the SYM‐H indices (2 hr in advance) in the large storm, SYMHnet achieves an FSS of 0.553 compared to the FSS of 0.087 of the GBM method. In addition, SYMHnet can provide results for both data and model uncertainty quantification, whereas the related methods cannot.

 

Coronal mass ejections (CMEs) are massive solar eruptions, which have a significant impact on Earth. In this paper, we propose a new method, called DeepCME, to estimate two properties of CMEs, namely, CME mass and kinetic energy. Being able to estimate these properties helps better understand CME dynamics. Our study is based on the CME catalog maintained at the Coordinated Data Analysis Workshops Data Center, which contains all CMEs manually identified since 1996 using the Large Angle and Spectrometric Coronagraph (LASCO) on board the Solar and Heliospheric Observatory. We use LASCO C2 data in the period between 1996 January and 2020 December to train, validate, and test DeepCME through 10-fold cross validation. The DeepCME method is a fusion of three deep-learning models, namely ResNet, InceptionNet, and InceptionResNet. Our fusion model extracts features from LASCO C2 images, effectively combining the learning capabilities of the three component models to jointly estimate the mass and kinetic energy of CMEs. Experimental results show that the fusion model yields a mean relative error (MRE) of 0.013 (0.009, respectively) compared to the MRE of 0.019 (0.017, respectively) of the best component model InceptionResNet (InceptionNet, respectively) in estimating the CME mass (kinetic energy, respectively). To our knowledge, this is the first time that deep learning has been used for CME mass and kinetic energy estimations.

 

The Sun constantly releases radiation and plasma into the heliosphere. Sporadically, the Sun launches solar eruptions such as flares and coronal mass ejections (CMEs). CMEs carry away a huge amount of mass and magnetic flux with them. An Earth-directed CME can cause serious consequences to the human system. It can destroy power grids/pipelines, satellites, and communications. Therefore, accurately monitoring and predicting CMEs is important to minimize damages to the human system. In this study we propose an ensemble learning approach, named CMETNet, for predicting the arrival time of CMEs from the Sun to the Earth. We collect and integrate eruptive events from two solar cycles, #23 and #24, from 1996 to 2021 with a total of 363 geoeffective CMEs. The data used for making predictions include CME features, solar wind parameters and CME images obtained from the SOHO/LASCO C2 coronagraph. Our ensemble learning framework comprises regression algorithms for numerical data analysis and a convolutional neural network for image processing. Experimental results show that CMETNet performs better than existing machine learning methods reported in the literature, with a Pearson product-moment correlation coefficient of 0.83 and a mean absolute error of 9.75 h. 

Rock salt caverns have been extensively used as reliable repositories for hazardous waste such as nuclear waste, oil or compressed gases. Undisturbed rock salt deposits in nature are usually impermeable and have very low porosity. However, rock salt formations under excavation stresses can develop crack networks, which increase their porosities; and in the case of a connected crack network within the media, rock salt may become permeable. Although the relationship between the permeability of rock salt and the applied stresses has been reported in the literature, a microscopic study that investigates the properties influencing this relationship, such as the evolution of texture and internal stresses, has yet to be conducted. This study employs in situ 3D synchrotron micro-computed tomography and 3D X-ray diffraction (3DXRD) on two small-scale polycrystalline rock salt specimens to investigate the evolution of the texture and internal stresses within the specimens. The 3DXRD technique measures the 3D crystal structure and lattice strains within rock salt grains. The specimens were prepared under 1D compression conditions and have shown an initial {111} preferred texture, a dominant {110}〈1 1 0〉 slip system and no fully connected crack network. The {111} preferred texture under the unconfined compression experiment became stronger, while the {111}〈1 1 0〉 slip system became more prominent. The specimens did not have a fully connected crack network until applied axial stresses reached about 30 MPa, at a point where the impermeability of the material becomes compromised due to the development of multiple major cracks. 

ABSTRACT We use comparisons between the Sydney-AAO Multi-object Integral Field Spectrograph (SAMI) Galaxy Survey and equilibrium galaxy models to infer the importance of disc fading in the transition of spirals into lenticular (S0) galaxies. The local S0 population has both higher photometric concentration and lower stellar spin than spiral galaxies of comparable mass and we test whether this separation can be accounted for by passive aging alone. We construct a suite of dynamically self-consistent galaxy models, with a bulge, disc, and halo using the galactics code. The dispersion-dominated bulge is given a uniformly old stellar population, while the disc is given a current star formation rate putting it on the main sequence, followed by sudden instantaneous quenching. We then generate mock observables (r-band images, stellar velocity, and dispersion maps) as a function of time since quenching for a range of bulge/total (B/T) mass ratios. The disc fading leads to a decline in measured spin as the bulge contribution becomes more dominant, and also leads to increased concentration. However, the quantitative changes observed after 5 Gyr of disc fading cannot account for all of the observed difference. We see similar results if we instead subdivide our SAMI Galaxy Survey sample by star formation (relative to the main sequence). We use EAGLE simulations to also take into account progenitor bias, using size evolution to infer quenching time. The EAGLE simulations suggest that the progenitors of current passive galaxies typically have slightly higher spin than present day star-forming disc galaxies of the same mass. As a result, progenitor bias moves the data further from the disc fading model scenario, implying that intrinsic dynamical evolution must be important in the transition from star-forming discs to passive discs. 

Tidal features in the outskirts of galaxies yield unique information about their past interactions and are a key prediction of the hierarchical structure formation paradigm. The Vera C. Rubin Observatory is poised to deliver deep observations for potentially millions of objects with visible tidal features, but the inference of galaxy interaction histories from such features is not straightforward. Utilizing automated techniques and human visual classification in conjunction with realistic mock images produced using the NewHorizon cosmological simulation, we investigate the nature, frequency, and visibility of tidal features and debris across a range of environments and stellar masses. In our simulated sample, around 80 per cent of the flux in the tidal features around Milky Way or greater mass galaxies is detected at the 10-yr depth of the Legacy Survey of Space and Time (30–31 mag arcsec−2), falling to 60 per cent assuming a shallower final depth of 29.5 mag arcsec−2. The fraction of total flux found in tidal features increases towards higher masses, rising to 10 per cent for the most massive objects in our sample (M⋆ ∼ 1011.5 M⊙). When observed at sufficient depth, such objects frequently exhibit many distinct tidal features with complex shapes. The interpretation and characterization of such features varies significantly with image depth and object orientation, introducing significant biases in their classification. Assuming the data reduction pipeline is properly optimized, we expect the Rubin Observatory to be capable of recovering much of the flux found in the outskirts of Milky Way mass galaxies, even at intermediate redshifts (z < 0.2).