Search for: All records

ABSTRACT We present a spatially resolved study of the kinematical properties of known supernova remnants (SNRs) in the nearest galaxies of the SIGNALS survey, namely NGC 6822 (one object) and M33 (163 objects), based on data obtained with the SITELLE imaging Fourier transform spectrometer at the Canada–France–Hawaii Telescope. The purpose of this paper is to provide a better scheme of identification for extragalactic SNRs and, in particular, to distinguish between H ii regions and SNRs. For that we have used diagrams which involve both the [S ii]/H$$\alpha$$ ratio and the velocity dispersion ($$\sigma$$). We also introduce a new parameter, $$\xi = {[\rm S\, {\small II}] \over H\alpha } \times \sigma$$, which enhances still the contrast between SNRs and the rest of the ionized gas. More than 90 per cent of the SNRs in our entire sample show an integrated [S ii]/H$$\alpha$$ ratio larger than the canonical value (0.4). 86 per cent of the SNRs present in our field show a significant velocity dispersion. The spectral resolution of our observations allows us to observe the complex velocity structure of some SNRs. 

Abstract Robust quantification of predictive uncertainty is a critical addition needed for machine learning applied to weather and climate problems to improve the understanding of what is driving prediction sensitivity. Ensembles of machine learning models provide predictive uncertainty estimates in a conceptually simple way but require multiple models for training and prediction, increasing computational cost and latency. Parametric deep learning can estimate uncertainty with one model by predicting the parameters of a probability distribution but does not account for epistemic uncertainty. Evidential deep learning, a technique that extends parametric deep learning to higher-order distributions, can account for both aleatoric and epistemic uncertainties with one model. This study compares the uncertainty derived from evidential neural networks to that obtained from ensembles. Through applications of the classification of winter precipitation type and regression of surface-layer fluxes, we show evidential deep learning models attaining predictive accuracy rivaling standard methods while robustly quantifying both sources of uncertainty. We evaluate the uncertainty in terms of how well the predictions are calibrated and how well the uncertainty correlates with prediction error. Analyses of uncertainty in the context of the inputs reveal sensitivities to underlying meteorological processes, facilitating interpretation of the models. The conceptual simplicity, interpretability, and computational efficiency of evidential neural networks make them highly extensible, offering a promising approach for reliable and practical uncertainty quantification in Earth system science modeling. To encourage broader adoption of evidential deep learning, we have developed a new Python package, Machine Integration and Learning for Earth Systems (MILES) group Generalized Uncertainty for Earth System Science (GUESS) (MILES-GUESS) (https://github.com/ai2es/miles-guess), that enables users to train and evaluate both evidential and ensemble deep learning. Significance StatementThis study demonstrates a new technique, evidential deep learning, for robust and computationally efficient uncertainty quantification in modeling the Earth system. The method integrates probabilistic principles into deep neural networks, enabling the estimation of both aleatoric uncertainty from noisy data and epistemic uncertainty from model limitations using a single model. Our analyses reveal how decomposing these uncertainties provides valuable insights into reliability, accuracy, and model shortcomings. We show that the approach can rival standard methods in classification and regression tasks within atmospheric science while offering practical advantages such as computational efficiency. With further advances, evidential networks have the potential to enhance risk assessment and decision-making across meteorology by improving uncertainty quantification, a longstanding challenge. This work establishes a strong foundation and motivation for the broader adoption of evidential learning, where properly quantifying uncertainties is critical yet lacking. 

Abstract Artificial intelligence (AI) and machine learning (ML) pose a challenge for achieving science that is both reproducible and replicable. The challenge is compounded in supervised models that depend on manually labeled training data, as they introduce additional decision‐making and processes that require thorough documentation and reporting. We address these limitations by providing an approach to hand labeling training data for supervised ML that integrates quantitative content analysis (QCA)—a method from social science research. The QCA approach provides a rigorous and well‐documented hand labeling procedure to improve the replicability and reproducibility of supervised ML applications in Earth systems science (ESS), as well as the ability to evaluate them. Specifically, the approach requires (a) the articulation and documentation of the exact decision‐making process used for assigning hand labels in a “codebook” and (b) an empirical evaluation of the reliability” of the hand labelers. In this paper, we outline the contributions of QCA to the field, along with an overview of the general approach. We then provide a case study to further demonstrate how this framework has and can be applied when developing supervised ML models for applications in ESS. With this approach, we provide an actionable path forward for addressing ethical considerations and goals outlined by recent AGU work on ML ethics in ESS. 

Abstract We present JWST observations of the Crab Nebula, the iconic remnant of the historical SN 1054. The observations include NIRCam and MIRI imaging mosaics plus MIRI/MRS spectra that probe two select locations within the ejecta filaments. We derive a high-resolution map of dust emission and show that the grains are concentrated in the innermost, high-density filaments. These dense filaments coincide with multiple synchrotron bays around the periphery of the Crab's pulsar wind nebula (PWN). We measure synchrotron spectral index changes in small-scale features within the PWN’s torus region, including the well-known knot and wisp structures. The index variations are consistent with Doppler boosting of emission from particles with a broken power-law distribution, providing the first direct evidence that the curvature in the particle injection spectrum is tied to the acceleration mechanism at the termination shock. We detect multiple nickel and iron lines in the ejecta filaments and use photoionization models to derive nickel-to-iron abundance ratios that are a factor of 3–8 higher than the solar ratio. We also find that the previously reported order-of-magnitude higher Ni/Fe values from optical data are consistent with the lower values from JWST when we reanalyze the optical emission using updated atomic data and account for local extinction from dust. We discuss the implications of our results for understanding the nature of the explosion that produced the Crab Nebula and conclude that the observational properties are most consistent with a low-mass Fe core-collapse supernova, even though an electron-capture explosion cannot be ruled out. 

Abstract We  present new high-spectral-resolution observations (R=λ/Δλ= 7000) of the filamentary nebula surrounding NGC 1275, the central galaxy of the Perseus cluster. These observations have been obtained with SITELLE, an imaging Fourier transform spectrometer installed on the Canada–France–Hawai Telescope with a field of view of $11\prime \times 11\prime$, encapsulating the entire filamentary structure of ionized gas despite its large size of 80 kpc × 50 kpc. Here, we present renewed fluxes, velocities, and velocity dispersion maps that show in great detail the kinematics of the optical nebula at [Sii]λ6716, [Sii]λ6731, [Nii]λ6584, Hα(6563 Å), and [Nii]λ6548. These maps reveal the existence of a bright flattened disk-shaped structure in the core extending tor∼10 kpc and dominated by a chaotic velocity field. This structure is located in the wake of X-ray cavities and characterized by a high mean velocity dispersion of 134 km s⁻¹. The disk-shaped structure is surrounded by an extended array of filaments spread out tor∼ 50 kpc that are 10 times fainter in flux, remarkably quiescent, and have a uniform mean velocity dispersion of 44 km s⁻¹. This stability is puzzling given that the cluster core exhibits several energetic phenomena. Based on these results, we argue that there are two mechanisms that form multiphase gas in clusters of galaxies: a first triggered in the wake of X-ray cavities leading to more turbulent multiphase gas and a second, distinct mechanism, that is gentle and leads to large-scale multiphase gas spreading throughout the core. 

Understanding and describing the diversity of living organisms is a great challenge. Fungi have for a long time been, and unfortunately still are, underestimated when it comes to taxonomic research. The foundations were laid by the first mycologists through field observations. These important fundamental works have been and remain vital reference works. Nevertheless, a non-negligible part of the studied funga escaped their attention. Thanks to modern developments in molecular techniques, the study of fungal diversity has been revolutionized in terms of tools and knowledge. Despite a number of disadvantages inherent to these techniques, traditional field-based inventory work has been increasingly superseded and neglected. This perspective aims to demonstrate the central importance of field-based research in fungal diversity studies, and encourages researchers not to be blinded by the sole use of molecular methods. 

When monitoring bats, the greatest yield in capture rate for survey effort can often be found in riparian and lentic habitats. However, capturing bats over large bodies of water is usually challenging due to the logistics of deploying equipment and extracting bats whilst ensuring the safety of surveyors. We present a novel technique – the “skynet” – as one solution to this problem, allowing fast and safe deployment of a suspended mist net between two anchor points over open water. Preliminary fieldwork in a Croatian scrub-dominated landscape yielded a capture of 22 bats of five species over a 1600 m2 pond. Our results demonstrate that the method is effective compared to a simultaneous net positioning on the bank of the same water body, which yielded no bats. System design and recommendations for bespoke alterations, alternative equipment options, and future investigations are presented here.