C Kreuzig, D Bischoff, N S Molinski, J N Brecher, A Kovalev, G Meier, J Oesert, S N Gorb, B Gundlach, J Blum
Abstract In this work, we present a comprehensive investigation into the production, characteristics, handling, and storage of micrometre-sized granular water-ice. The focus of this research is to provide well-characterized analogue samples for laboratory experiments simulating icy bodies found in the Solar System, particularly comets. These experiments are conducted as part of the CoPhyLab (Comet Physics Laboratory) project, an international collaboration aimed at studying cometary processes to gain insights into the underlying physics of cometary activity. Granular water-ice, along with other less abundant but more volatile ices, plays a crucial role in the ejection of gas and dust particles when comets approach the Sun. To facilitate large-scale laboratory experiments, an ice-particle machine was developed, capable of autonomously producing sufficient quantities of granular water-ice. Additionally, a cryogenic desiccator was designed to remove any residual moisture from the ice using liquid nitrogen. The resulting ice particles can be mixed with other materials and stored within the desiccator or a cryogenic transport can, enabling easy shipment to any laboratory, including via air transport. To analyse the ice grains, cryogenic scanning electron microscopy was employed to determine their particle shape and size-frequency distribution. These analyses contribute to a better understanding of the properties of granular water-ice and its behavior under cryogenic conditions, supporting the objectives of the CoPhyLab project.
{"title":"Micrometre-sized ice particles for planetary science experiments – CoPhyLab cryogenic granular sample production and storage","authors":"C Kreuzig, D Bischoff, N S Molinski, J N Brecher, A Kovalev, G Meier, J Oesert, S N Gorb, B Gundlach, J Blum","doi":"10.1093/rasti/rzad049","DOIUrl":"https://doi.org/10.1093/rasti/rzad049","url":null,"abstract":"Abstract In this work, we present a comprehensive investigation into the production, characteristics, handling, and storage of micrometre-sized granular water-ice. The focus of this research is to provide well-characterized analogue samples for laboratory experiments simulating icy bodies found in the Solar System, particularly comets. These experiments are conducted as part of the CoPhyLab (Comet Physics Laboratory) project, an international collaboration aimed at studying cometary processes to gain insights into the underlying physics of cometary activity. Granular water-ice, along with other less abundant but more volatile ices, plays a crucial role in the ejection of gas and dust particles when comets approach the Sun. To facilitate large-scale laboratory experiments, an ice-particle machine was developed, capable of autonomously producing sufficient quantities of granular water-ice. Additionally, a cryogenic desiccator was designed to remove any residual moisture from the ice using liquid nitrogen. The resulting ice particles can be mixed with other materials and stored within the desiccator or a cryogenic transport can, enabling easy shipment to any laboratory, including via air transport. To analyse the ice grains, cryogenic scanning electron microscopy was employed to determine their particle shape and size-frequency distribution. These analyses contribute to a better understanding of the properties of granular water-ice and its behavior under cryogenic conditions, supporting the objectives of the CoPhyLab project.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"33 3","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135874505","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Nikolaos Nikolaou, Ingo P. Waldmann, Angelos Tsiaras, Mario Morvan, Billy Edwards, Kai Hou Yip, Giovanna Tinetti, Subhajit Sarkar, James M. Dawson, Vadim Borisov, Gjergji Kasneci, Matej Petkovic, Tomaz Stepisnik, Tarek Al-Ubaidi, Rachel Louise Bailey, Michael Granitzer, Sahib Julka, Roman Kern, Patrick Ofner, Stefan Wagner, Lukas Heppe, Mirko Bunse, Katharina Morik
Abstract The last decade has witnessed a rapid growth of the field of exoplanet discovery and characterisation. However, several big challenges remain, many of which could be addressed using machine learning methodology. For instance, the most prolific method for detecting exoplanets and inferring several of their characteristics, transit photometry, is very sensitive to the presence of stellar spots. The current practice in the literature is to identify the effects of spots visually and correct for them manually or discard the affected data. This paper explores a first step towards fully automating the efficient and precise derivation of transit depths from transit light curves in the presence of stellar spots. The primary focus of the paper is to present in detail a diverse arsenal of methods for doing so. The methods and results we present were obtained in the context of the 1st Machine Learning Challenge organized for the European Space Agency’s upcoming Ariel mission. We first present the problem, the simulated Ariel-like data and outline the Challenge while identifying best practices for organizing similar challenges in the future. Finally, we present the solutions obtained by the top-5 winning teams, provide their code and discuss their implications. Successful solutions either construct highly non-linear (w.r.t. the raw data) models with minimal preprocessing –deep neural networks and ensemble methods– or amount to obtaining meaningful statistics from the light curves, constructing linear models on which yields comparably good predictive performance.
{"title":"Lessons learned from the 1st Ariel Machine Learning Challenge: Correcting transiting exoplanet light curves for stellar spots","authors":"Nikolaos Nikolaou, Ingo P. Waldmann, Angelos Tsiaras, Mario Morvan, Billy Edwards, Kai Hou Yip, Giovanna Tinetti, Subhajit Sarkar, James M. Dawson, Vadim Borisov, Gjergji Kasneci, Matej Petkovic, Tomaz Stepisnik, Tarek Al-Ubaidi, Rachel Louise Bailey, Michael Granitzer, Sahib Julka, Roman Kern, Patrick Ofner, Stefan Wagner, Lukas Heppe, Mirko Bunse, Katharina Morik","doi":"10.1093/rasti/rzad050","DOIUrl":"https://doi.org/10.1093/rasti/rzad050","url":null,"abstract":"Abstract The last decade has witnessed a rapid growth of the field of exoplanet discovery and characterisation. However, several big challenges remain, many of which could be addressed using machine learning methodology. For instance, the most prolific method for detecting exoplanets and inferring several of their characteristics, transit photometry, is very sensitive to the presence of stellar spots. The current practice in the literature is to identify the effects of spots visually and correct for them manually or discard the affected data. This paper explores a first step towards fully automating the efficient and precise derivation of transit depths from transit light curves in the presence of stellar spots. The primary focus of the paper is to present in detail a diverse arsenal of methods for doing so. The methods and results we present were obtained in the context of the 1st Machine Learning Challenge organized for the European Space Agency’s upcoming Ariel mission. We first present the problem, the simulated Ariel-like data and outline the Challenge while identifying best practices for organizing similar challenges in the future. Finally, we present the solutions obtained by the top-5 winning teams, provide their code and discuss their implications. Successful solutions either construct highly non-linear (w.r.t. the raw data) models with minimal preprocessing –deep neural networks and ensemble methods– or amount to obtaining meaningful statistics from the light curves, constructing linear models on which yields comparably good predictive performance.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"19 4","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135874953","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
S Sullivan, M J Page, A A Breeveld, S Rosen, A Talavera
Abstract After nearly 18 years of successful and safe operations, on 2017 July 17th and 18th XMM-Newton’s Optical Monitor (OM) observed Jupiter – an object 10 magnitudes brighter than safe brightness limits – in it’s Visual (V) filter. The object was exposed 40 arc seconds from the nominal EPIC pn boresight in the negative Y direction, creating a patch of depleted sensitivity. Two exposures of 4000s each in the V filter left the sensitivity depleted in an area 70 by 40 arcsec; with the decrease in throughput varying from 34 per cent in the V filter to 15 per cent in the UV bands, but reaching a depth of 45 per cent in flat field exposures. The wavelength dependency suggests the majority of the detector damage is due to loss of sensitivity in the photocathode from damage inflicted by ion feedback, while up to 15 per cent could be due to gain depletion of the MCP. The physical mechanisms causing the damage to the detector are discussed as well as possible solutions and opportunities that exist for the future operation of the OM.
{"title":"An investigation into the effect of exposing XMM-Newton’s Optical Monitor to the light of Jupiter","authors":"S Sullivan, M J Page, A A Breeveld, S Rosen, A Talavera","doi":"10.1093/rasti/rzad048","DOIUrl":"https://doi.org/10.1093/rasti/rzad048","url":null,"abstract":"Abstract After nearly 18 years of successful and safe operations, on 2017 July 17th and 18th XMM-Newton’s Optical Monitor (OM) observed Jupiter – an object 10 magnitudes brighter than safe brightness limits – in it’s Visual (V) filter. The object was exposed 40 arc seconds from the nominal EPIC pn boresight in the negative Y direction, creating a patch of depleted sensitivity. Two exposures of 4000s each in the V filter left the sensitivity depleted in an area 70 by 40 arcsec; with the decrease in throughput varying from 34 per cent in the V filter to 15 per cent in the UV bands, but reaching a depth of 45 per cent in flat field exposures. The wavelength dependency suggests the majority of the detector damage is due to loss of sensitivity in the photocathode from damage inflicted by ion feedback, while up to 15 per cent could be due to gain depletion of the MCP. The physical mechanisms causing the damage to the detector are discussed as well as possible solutions and opportunities that exist for the future operation of the OM.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"47 6","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135567782","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
C R Tinker, T D Glotch, L B Breitenfeld, A Ryan, L Li
Abstract Airless bodies in the Solar System are commonly dominated by complex regolith mixtures consisting of coarse and fine particulates. These materials often manifest as coatings with the potential to modify or obscure the spectral signatures of underlying substrates. This can make accurate spectral analysis of surface materials challenging, especially for thermal infrared (TIR) techniques of which the spectral properties concurrently depend on grain size and albedo. Further complexity is presented when these coatings occur as discontinuous patterns in which some substrate is exposed and some is masked. Discontinuous patterns are distinguished by scale as having macroscopic or microscopic discontinuity, with the former being patches of homogeneous dust covering portions of the substrate and the latter being randomly distributed individual particles on the substrate. Investigations of asteroid (101955) Bennu’s surface by NASA’s Origins, Spectral Interpretation, Resource Identification, and Security-Regolith Explorer (OSIRIS-REx) have revealed contradictions between spectral and thermophysical results that are hypothesized to indicate the presence of thin and/or laterally discontinuous dust coatings. To address this, we constructed an environment chamber that enables the controlled deposition of size-regulated dust particles in coatings with varying continuity and thickness. TIR spectra of coated substrates acquired in a simulated asteroid environment (SAE) are used to investigate the extent to which dust coatings of different thicknesses and arrangements contribute to orbital spectral signatures of airless body surfaces.
{"title":"Experimental and analytical methods for thermal infrared spectroscopy of complex dust coatings in a simulated asteroid environment","authors":"C R Tinker, T D Glotch, L B Breitenfeld, A Ryan, L Li","doi":"10.1093/rasti/rzad047","DOIUrl":"https://doi.org/10.1093/rasti/rzad047","url":null,"abstract":"Abstract Airless bodies in the Solar System are commonly dominated by complex regolith mixtures consisting of coarse and fine particulates. These materials often manifest as coatings with the potential to modify or obscure the spectral signatures of underlying substrates. This can make accurate spectral analysis of surface materials challenging, especially for thermal infrared (TIR) techniques of which the spectral properties concurrently depend on grain size and albedo. Further complexity is presented when these coatings occur as discontinuous patterns in which some substrate is exposed and some is masked. Discontinuous patterns are distinguished by scale as having macroscopic or microscopic discontinuity, with the former being patches of homogeneous dust covering portions of the substrate and the latter being randomly distributed individual particles on the substrate. Investigations of asteroid (101955) Bennu’s surface by NASA’s Origins, Spectral Interpretation, Resource Identification, and Security-Regolith Explorer (OSIRIS-REx) have revealed contradictions between spectral and thermophysical results that are hypothesized to indicate the presence of thin and/or laterally discontinuous dust coatings. To address this, we constructed an environment chamber that enables the controlled deposition of size-regulated dust particles in coatings with varying continuity and thickness. TIR spectra of coated substrates acquired in a simulated asteroid environment (SAE) are used to investigate the extent to which dust coatings of different thicknesses and arrangements contribute to orbital spectral signatures of airless body surfaces.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136078622","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Future surveys such as the Legacy Survey of Space and Time (LSST) of the Vera C. Rubin Observatory will observe an order of magnitude more astrophysical transient events than any previous survey before. With this deluge of photometric data, it will be impossible for all such events to be classified by humans alone. Recent efforts have sought to leverage machine learning methods to tackle the challenge of astronomical transient classification, with ever improving success. Transformers are a recently developed deep learning architecture, first proposed for natural language processing, that have shown a great deal of recent success. In this work we develop a new transformer architecture, which uses multi-head self attention at its core, for general multi-variate time-series data. Furthermore, the proposed time-series transformer architecture supports the inclusion of an arbitrary number of additional features, while also offering interpretability. We apply the time-series transformer to the task of photometric classification, minimising the reliance of expert domain knowledge for feature selection, while achieving results comparable to state-of-the-art photometric classification methods. We achieve a logarithmic-loss of 0.507 on imbalanced data in a representative setting using data from the Photometric LSST Astronomical Time-Series Classification Challenge (PLAsTiCC). Moreover, we achieve a micro-averaged receiver operating characteristic area under curve of 0.98 and micro-averaged precision-recall area under curve of 0.87.
未来的调查，如Vera C. Rubin天文台的遗产时空调查(LSST)，将比以往任何调查都能观测到更多的天体物理瞬变事件。有了这些海量的光度数据，单靠人类对所有这些事件进行分类是不可能的。最近的努力试图利用机器学习方法来解决天文瞬态分类的挑战，并取得了越来越大的成功。变形金刚是最近开发的一种深度学习架构，最初是为自然语言处理提出的，最近取得了很大的成功。在这项工作中，我们开发了一种新的变压器架构，它以多头自关注为核心，用于一般的多变量时间序列数据。此外，所建议的时间序列转换器体系结构支持包含任意数量的附加特性，同时还提供可解释性。我们将时间序列转换器应用于光度分类任务，最大限度地减少了特征选择对专家领域知识的依赖，同时实现了与最先进的光度分类方法相当的结果。我们使用来自Photometric LSST天文时间序列分类挑战(PLAsTiCC)的数据，在代表性设置中对不平衡数据实现了0.507的对数损失。此外，我们还实现了微平均接收机工作特征曲线下面积为0.98，微平均精确召回面积为0.87。
{"title":"Paying attention to astronomical transients: Introducing the time-series transformer for photometric classification","authors":"Tarek Allam, Jason D McEwen","doi":"10.1093/rasti/rzad046","DOIUrl":"https://doi.org/10.1093/rasti/rzad046","url":null,"abstract":"Future surveys such as the Legacy Survey of Space and Time (LSST) of the Vera C. Rubin Observatory will observe an order of magnitude more astrophysical transient events than any previous survey before. With this deluge of photometric data, it will be impossible for all such events to be classified by humans alone. Recent efforts have sought to leverage machine learning methods to tackle the challenge of astronomical transient classification, with ever improving success. Transformers are a recently developed deep learning architecture, first proposed for natural language processing, that have shown a great deal of recent success. In this work we develop a new transformer architecture, which uses multi-head self attention at its core, for general multi-variate time-series data. Furthermore, the proposed time-series transformer architecture supports the inclusion of an arbitrary number of additional features, while also offering interpretability. We apply the time-series transformer to the task of photometric classification, minimising the reliance of expert domain knowledge for feature selection, while achieving results comparable to state-of-the-art photometric classification methods. We achieve a logarithmic-loss of 0.507 on imbalanced data in a representative setting using data from the Photometric LSST Astronomical Time-Series Classification Challenge (PLAsTiCC). Moreover, we achieve a micro-averaged receiver operating characteristic area under curve of 0.98 and micro-averaged precision-recall area under curve of 0.87.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"85 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135045770","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract Kuiper belt objects smaller than a few kilometres are difficult to observe directly. They can be detected when they randomly occult a background star. Close to the ecliptic plane, each star is occulted once every tens of thousands of hours, and occultations typically last for less than a second. We present an algorithm, and companion pipeline, for detection of diffractive occultation events. Our approach includes: cleaning the data; an efficient and optimal matched filtering of the light curves with a template bank of diffractive occultations; treating the red-noise in the light curves; injection of simulated events for efficiency estimation; and applying data quality cuts. We discuss human vetting of the candidate events in a blinded way to reduce bias caused by the human-in-the-loop. We present Markov Chain Monte Carlo tools to estimate the parameters of candidate occultations, and test them on simulated events. This pipeline is used by the W-FAST. The methods discussed here can be applied to searches for other Trans-Neptunian objects, albeit with larger radii that correspond to a larger diffraction length scale.
{"title":"A reduction procedure and pipeline for the detection of trans-Neptunian objects using occultations","authors":"Guy Nir, Eran O Ofek, Barak Zackay","doi":"10.1093/rasti/rzad040","DOIUrl":"https://doi.org/10.1093/rasti/rzad040","url":null,"abstract":"Abstract Kuiper belt objects smaller than a few kilometres are difficult to observe directly. They can be detected when they randomly occult a background star. Close to the ecliptic plane, each star is occulted once every tens of thousands of hours, and occultations typically last for less than a second. We present an algorithm, and companion pipeline, for detection of diffractive occultation events. Our approach includes: cleaning the data; an efficient and optimal matched filtering of the light curves with a template bank of diffractive occultations; treating the red-noise in the light curves; injection of simulated events for efficiency estimation; and applying data quality cuts. We discuss human vetting of the candidate events in a blinded way to reduce bias caused by the human-in-the-loop. We present Markov Chain Monte Carlo tools to estimate the parameters of candidate occultations, and test them on simulated events. This pipeline is used by the W-FAST. The methods discussed here can be applied to searches for other Trans-Neptunian objects, albeit with larger radii that correspond to a larger diffraction length scale.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135784048","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
M C Bezuidenhout, C J Clark, R P Breton, B W Stappers, E D Barr, M Caleb, W Chen, F Jankowski, M Kramer, K Rajwade, M Surnis
Abstract Multi-element interferometers such as MeerKAT, which observe with high time resolution and have a wide field of view, provide an ideal opportunity to perform real-time, untargeted transient and pulsar searches. However, because of data storage limitations, it is not always feasible to store the baseband data required to image the field of a discovered transient or pulsar. This limits the ability of surveys to effectively localize their discoveries and may restrict opportunities for follow-up science, especially of one-off events like some fast radio bursts. Here, we present a novel maximum-likelihood estimation approach to localizing transients and pulsars detected in multiple MeerKAT tied-array beams at once, which we call tied-array beam localization, as well as a Python implementation of the method named SeeKAT. We provide real-world examples of SeeKAT’s use as well as a Monte Carlo analysis to show that it is capable of localizing single pulses detected in beamformed MeerKAT data to (sub)arcsec precision.
{"title":"Tied-array beam localization of radio transients and pulsars","authors":"M C Bezuidenhout, C J Clark, R P Breton, B W Stappers, E D Barr, M Caleb, W Chen, F Jankowski, M Kramer, K Rajwade, M Surnis","doi":"10.1093/rasti/rzad007","DOIUrl":"https://doi.org/10.1093/rasti/rzad007","url":null,"abstract":"Abstract Multi-element interferometers such as MeerKAT, which observe with high time resolution and have a wide field of view, provide an ideal opportunity to perform real-time, untargeted transient and pulsar searches. However, because of data storage limitations, it is not always feasible to store the baseband data required to image the field of a discovered transient or pulsar. This limits the ability of surveys to effectively localize their discoveries and may restrict opportunities for follow-up science, especially of one-off events like some fast radio bursts. Here, we present a novel maximum-likelihood estimation approach to localizing transients and pulsars detected in multiple MeerKAT tied-array beams at once, which we call tied-array beam localization, as well as a Python implementation of the method named SeeKAT. We provide real-world examples of SeeKAT’s use as well as a Monte Carlo analysis to show that it is capable of localizing single pulses detected in beamformed MeerKAT data to (sub)arcsec precision.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"75 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135126920","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Marc Huertas-Company, Regina Sarmiento, Johan H Knapen
Abstract Reliable tools to extract patterns from high-dimensionality spaces are becoming more necessary as astronomical data sets increase both in volume and complexity. Contrastive Learning is a self-supervised machine learning algorithm that extracts informative measurements from multidimensional data sets, which has become increasingly popular in the computer vision and Machine Learning communities in recent years. To do so, it maximizes the agreement between the information extracted from augmented versions of the same input data, making the final representation invariant to the applied transformations. Contrastive Learning is particularly useful in astronomy for removing known instrumental effects and for performing supervised classifications and regressions with a limited amount of available labels, showing a promising avenue towards Foundation Models. This short review paper briefly summarizes the main concepts behind contrastive learning and reviews the first promising applications to astronomy. We include some practical recommendations on which applications are particularly attractive for contrastive learning.
{"title":"A brief review of contrastive learning applied to astrophysics","authors":"Marc Huertas-Company, Regina Sarmiento, Johan H Knapen","doi":"10.1093/rasti/rzad028","DOIUrl":"https://doi.org/10.1093/rasti/rzad028","url":null,"abstract":"Abstract Reliable tools to extract patterns from high-dimensionality spaces are becoming more necessary as astronomical data sets increase both in volume and complexity. Contrastive Learning is a self-supervised machine learning algorithm that extracts informative measurements from multidimensional data sets, which has become increasingly popular in the computer vision and Machine Learning communities in recent years. To do so, it maximizes the agreement between the information extracted from augmented versions of the same input data, making the final representation invariant to the applied transformations. Contrastive Learning is particularly useful in astronomy for removing known instrumental effects and for performing supervised classifications and regressions with a limited amount of available labels, showing a promising avenue towards Foundation Models. This short review paper briefly summarizes the main concepts behind contrastive learning and reviews the first promising applications to astronomy. We include some practical recommendations on which applications are particularly attractive for contrastive learning.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135989061","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Minghan Chen, Jason J Wang, Timothy D Brandt, Thayne Currie, Julien Lozi, Jeffrey Chilcote, Maria Vincent
Abstract We present the pyKLIP-CHARIS post-processing pipeline, a Python library that reduces high contrast imaging data for the CHARIS integral field spectrograph used with the SCExAO project on the Subaru Telescope. The pipeline is a part of the pyklip package, a Python library dedicated to the reduction of direct imaging data of exoplanets, brown dwarfs, and discs. For PSF subtraction, the pyKLIP-CHARIS post-processing pipeline relies on the core algorithms implemented in pyklip but uses image registration and calibrations that are unique to CHARIS. We describe the pipeline procedures, calibration results, and capabilities in processing imaging data acquired via the angular differential imaging and spectral differential imaging observing techniques. We showcase its performance on extracting spectra of injected synthetic point sources as well as compare the extracted spectra from real data sets on HD 33632 and HR 8799 to results in the literature. The pipeline is a python-based complement to the SCExAO project supported, widely used (and currently IDL-based) CHARIS data post-processing pipeline (CHARIS DPP) and provides an additional approach to reducing CHARIS data and extracting calibrated planet spectra.
{"title":"Post-processing CHARIS integral field spectrograph data with pyKLIP","authors":"Minghan Chen, Jason J Wang, Timothy D Brandt, Thayne Currie, Julien Lozi, Jeffrey Chilcote, Maria Vincent","doi":"10.1093/rasti/rzad039","DOIUrl":"https://doi.org/10.1093/rasti/rzad039","url":null,"abstract":"Abstract We present the pyKLIP-CHARIS post-processing pipeline, a Python library that reduces high contrast imaging data for the CHARIS integral field spectrograph used with the SCExAO project on the Subaru Telescope. The pipeline is a part of the pyklip package, a Python library dedicated to the reduction of direct imaging data of exoplanets, brown dwarfs, and discs. For PSF subtraction, the pyKLIP-CHARIS post-processing pipeline relies on the core algorithms implemented in pyklip but uses image registration and calibrations that are unique to CHARIS. We describe the pipeline procedures, calibration results, and capabilities in processing imaging data acquired via the angular differential imaging and spectral differential imaging observing techniques. We showcase its performance on extracting spectra of injected synthetic point sources as well as compare the extracted spectra from real data sets on HD 33632 and HR 8799 to results in the literature. The pipeline is a python-based complement to the SCExAO project supported, widely used (and currently IDL-based) CHARIS data post-processing pipeline (CHARIS DPP) and provides an additional approach to reducing CHARIS data and extracting calibrated planet spectra.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"127 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135102199","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Steven J Gibbons, Ashley P Willis, Chris Davies, David Gubbins
Abstract We present a set of codes for calculating and displaying solutions to diverse problems within thermal convection and magnetic field generation in rotating fluid-filled spheres and spherical shells. There are diverse programs for the kinematic dynamo problem, the onset of thermal convection, and boundary-locked thermal convection, and time-stepping codes for non-magnetic convection and the dynamo with either homogeneous or spatially varying thermal boundary conditions. Where possible, all programs have been benchmarked against other codes and tested by reproducing previously published results. Each program comes with the complete source code, a pdf instruction manual, and at least one example run with a sample input file and all necessary files for describing an initial condition. The only prerequisite for running most of the codes is a FORTRAN compiler. The plotting programs require in addition the PGPLOT graphics library. All source code, examples, input files, solutions, and instructions are available for download from github and Zenodo.
{"title":"A set of codes for numerical convection and geodynamo calculations","authors":"Steven J Gibbons, Ashley P Willis, Chris Davies, David Gubbins","doi":"10.1093/rasti/rzad043","DOIUrl":"https://doi.org/10.1093/rasti/rzad043","url":null,"abstract":"Abstract We present a set of codes for calculating and displaying solutions to diverse problems within thermal convection and magnetic field generation in rotating fluid-filled spheres and spherical shells. There are diverse programs for the kinematic dynamo problem, the onset of thermal convection, and boundary-locked thermal convection, and time-stepping codes for non-magnetic convection and the dynamo with either homogeneous or spatially varying thermal boundary conditions. Where possible, all programs have been benchmarked against other codes and tested by reproducing previously published results. Each program comes with the complete source code, a pdf instruction manual, and at least one example run with a sample input file and all necessary files for describing an initial condition. The only prerequisite for running most of the codes is a FORTRAN compiler. The plotting programs require in addition the PGPLOT graphics library. All source code, examples, input files, solutions, and instructions are available for download from github and Zenodo.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134889104","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：