SEARCH · Engineering Papers

Results for “recurrence”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,135 records · Page 63

All sources NASA NTRS DOE OSTI

NASA NTRS2018

On the Natural Structure of Amino Acid Patterns in Families of Protein Sequences

All known terrestrial proteins are coded as continuous strings of ≈20 amino acids. The patterns formed by the repetitions of elements in groups of finite sequences describes the natural architectures of protein families. We present a method to search for patterns and groupings of patterns in protein sequences using a mathematically precise definition for “repetition”, an efficient algorithmic implementation and a robust scoring system with no adjustable parameters. We show that the sequence patterns can be well-separated into disjoint classes according to their recurrence in nested structures. The statistics of the occurrences of patterns indicate that short repetitions are sufficient to account for the differences between natural families and randomized groups of sequences by more than 10 standard deviations, while contiguous sequence patterns shorter than 5 residues are effectively random in their occurrences. A small subset of patterns is sufficient to account for a robust ”familiarity” definition between arbitrary sets of sequences.

Pablo Turjanski↗

NASA NTRS2020

Amino Acid Encoding for Deep Learning Applications

Background: The number of applications of deep learning algorithms in bioinformatics is increasing as they usually achieve superior performance over classical approaches, especially, when bigger training datasets are available. In deep learning applications, discrete data, e.g. words or n-grams in language, or amino acids or nucleotides in bioinformatics, are generally represented as a continuous vector through an embedding matrix. Recently, learning this embedding matrix directly from the data as part of the continuous iteration of the model to optimize the target prediction – a process called ‘end-to-end learning’ – has led to state-ofthe-art results in many fields. Although usage of embeddings is well described in the bioinformatics literature, the potential of end-to-end learning for single amino acids, as compared to more classical manually-curated encoding strategies, has not been systematically addressed. To this end, we compared classical encoding matrices, namely one-hot, VHSE8 and BLOSUM62, to end-to-end learning of amino acid embeddings for two different prediction tasks using three widely used architectures, namely recurrent neural networks (RNN), convolutional neural networks (CNN), and the hybrid CNN-RNN. Results: By using different deep learning architectures, we show that end-to-end learning is on par with classical encodings for embeddings of the same dimension even when limited training data is available, and might allow for a reduction in the embedding dimension without performance loss, which is critical when deploying the models to devices with limited computational capacities. We found that the embedding dimension is a major factor in controlling the model performance. Surprisingly, we observed that deep learning models are capable of learning from random vectors of appropriate dimension. Conclusion: Our study shows that end-to-end learning is a flexible and powerful method for amino acid encoding. Further, due to the flexibility of deep learning systems, amino acid encoding schemes should be benchmarked against random vectors of the same dimension to disentangle the information content provided by the encoding scheme from the distinguishability effect provided by the scheme.

Hesham ElAbd↗

NASA NTRS2023

Soil Respiration–Driven CO 2 Pulses Dominate Australia’s Flux Variability

The Australian continent contributes substantially to the year-to-year variability of the global terrestrial carbon dioxide (CO 2 ) sink. However, the scarcity of in situ observations in remote areas prevents the deciphering of processes that force the CO 2 flux variability. In this study, by examining atmospheric CO 2 measurements from satellites in the period 2009–2018, we find recurrent end-of-dry-season CO 2 pulses over the Australian continent. These pulses largely control the year-to-year variability of Australia’s CO 2 balance. They cause two to three times larger seasonal variations compared with previous top-down inversions and bottom-up estimates. The pulses occur shortly after the onset of rainfall and are driven by enhanced soil respiration preceding photosynthetic uptake in Australia’s semiarid regions. The suggested continental-scale relevance of soil-rewetting processes has substantial implications for our understanding and modeling of global climate–carbon cycle feedbacks.

Eva-Marie Metz↗

NASA NTRS2023

Prediction of Stiffness and Fatigue Lives of Polymer Matrix Composite Laminates Using Artificial Neural Networks

Machine learning (ML) models are increasingly being used in many engineering fields due to the advancements in ML algorithms and availability of high-speed computing power. One of the most popular ML class of models is artificial neural networks (ANN). ML is increasingly being used in the design and analysis of composite materials and structures, specifically in the constitutive modeling of composite materials with the focus on greatly accelerating multiscale analyses of composite materials and structures through development of surrogate models. Towards that end, both Python and MATLAB-based neural nets have been developed to predict initial stiffness and fatigue life of an eight-ply symmetric polymer matrix composite laminate. Two types of neural networks, a Multilayer Perceptron (MLP) and a Recurrent Neural Network (RNN), have been developed for both platforms. Results show that the both neural net types can provide an excellent estimate of initial stiffness as well as fatigue life of eight-ply symmetric polymer matrix composite laminate. RNNs are better able to capture the shape of the fatigue curve of a laminate. This tool can be very useful for system level studies to obtain an estimate of desired properties and life of PMC composite laminates. The associated surrogate models could also be used in composite multiscale analyses to replace the actual physics-based calculations at lower scales and thereby significantly increase the computational efficiency of such analyses and thus make multiscale analyses a viable industrial tool for large scale structural problems.

Composite↗

NASA NTRS2022

A preliminary assessment of olivine phenocrysts from the monogenetic basalt of the McCartys Flow, Zuni-Bandera Volcanic Field, New Mexico

Monogenetic small-volume basaltic volcanoes are the most abundant subaerial volcanic landforms on Earth but are some of the most poorly understood systems. Their short durations, small volumes, and lack of recurrence make monitoring and hazard assessment difficult. The Zuni-Bandera volcanic field in western New Mexico contains small-volume basaltic centers erupting tholeiitic to alkalic basalts. Evidence shows no correlation of magma composition with eruption age, location, or volumetric output, prompting questions about the influence of magma ascent rates, magma storage conditions, and mantle source characteristics on lava compositions. Here, we present olivine major and minor element mineral chemistry from the 3200-year-old McCartys Flow, the youngest tholeiite basalt in the volcanic field. Olivine displays four phenocryst types with unique textures and major and minor element compositions. Multiple olivine types co-exist at the thin section scale. Major and minor element diffusion at frozen melt–phenocryst interfaces was modeled, revealing magma residence times ranging from 3–9 months. Type 3 olivine phenocrysts require step function initial conditions and record diffusion re-equilibration followed by magma mixing. These profiles indicate the magma resided in the reservoir for 10–15 years and accumulated from multiple batches of mixed magmas less than 10 days before the eruption. Our results show that primitive magmas in small-volume monogenetic volcanoes have complex lithospheric magmatic histories and stored in magma bodies influenced by an open system to develop different local chemical environments.

Gary S. Michelfelder↗

NASA NTRSDate not supplied

Stiffness and Fatigue Life Estimator for Polymer Composite Laminates Using Machine Learning

Machine learning (ML) models are increasingly being used in many engineering fields due to the advancements in ML algorithms and availability of high-speed computing power. One of the most popular ML class of models is artificial neural networks (ANN). ML is increasingly being used in the design and analysis of composite materials and structures, specifically in the constitutive modeling of composite materials with the focus on greatly accelerating multiscale analyses of composite materials and structures through development of surrogate models. Towards that end, Python-based neural nets have been developed to predict initial stiffness and fatigue life of an eight-ply symmetric polymer matrix composite laminate. Two types of neural networks, a Multilayer Perceptron (MLP) and a Recurrent Neural Network (RNN), have been established. Results show that both neural net type algorithms can provide an excellent estimate of initial laminate stiffness as well as fatigue life of eight-ply symmetric polymer matrix composite laminates (PMCs). RNNs are better able to capture the shape of the fatigue curve of a laminate. The resulting tool and GUI can be very useful for system level studies to obtain an estimate of desired properties and life of PMC composite laminates. Further, the associated surrogate models can also be used in composite multiscale analyses to replace the actual physics-based calculations at lower scales and thereby significantly increase the computational efficiency of such analyses and thus make micromechanics-based multiscale analyses a viable industrial tool for large scale structural problems.

multiscale analysis↗

NASA NTRSDate not supplied

Impact of Pacific Ocean Heatwaves on Phytoplankton Community Composition

Since 2013, marine heatwaves have become recurrent throughout the equatorial and northeastern Pacific Ocean and are expected to increase in intensity relative to historic norms. Among the ecological ramifications associated with these high temperature anomalies are increased mortality of higher trophic organisms such as marine mammals and seabirds, which are likely triggered by changes in the composition of phytoplankton, the base of the marine trophic food web. Here, we assimilated satellite ocean color data into an ocean biogeochemical model to describe changes in the abundance of phytoplankton functional types (PFTs) during the last decade’s (2010s) warm anomalies in the equatorial and northeastern Pacific Ocean. We find important changes associated with the “Blob” warm anomaly in the Gulf of Alaska, where reduced silica supply led to a switch in community composition from diatoms to dinoflagellates, resulting in an increase in surface ocean chlorophyll during the Summer–Fall of 2014. A more dramatic change was observed in the equatorial Pacific, where the extreme warm conditions of the 2016 El Niño resulted in a major decline of about 40% in surface chlorophyll, which was associated with a nearly total collapse in diatoms.

Lionel A Quintero↗

NASA NTRS2023

New Evidence on the Origin of Solar Wind Microstreams/Switchbacks

Microstreams are fluctuations in the solar wind speed and density associated with polarity-reversing folds in the magnetic field (also denoted switchbacks). Despite their long heritage, the origin of these microstreams/switchbacks remains poorly understood. For the first time, we investigated periodicities in microstreams during Parker Solar Probe (PSP) Encounter 10 to understand their origin. Our analysis was focused on the inbound corotation interval on 2021 November 19–21, while the spacecraft dove toward a small area within a coronal hole (CH). Solar Dynamics Observatory remote-sensing observations provide rich context for understanding the PSP in situ data. Extreme ultraviolet images from the Atmospheric Imaging Assembly reveal numerous recurrent jets occurring within the region that was magnetically connected to PSP during intervals that contained microstreams. The periods derived from the fluctuating radial velocities in the microstreams (approximately 3, 5, 10, and 20 minutes) are consistent with the periods measured in the emission intensity of the jetlets at the base of the CH plumes, as well as in larger coronal jets and in the plume fine structures. Helioseismic and Magnetic Imager magnetograms reveal the presence of myriad embedded bipoles, which are known sources of reconnection-driven jets on all scales. Simultaneous enhancements in the PSP proton flux and ionic ( 3 He, 4 He, Fe, O) composition during the microstreams further support the connection with jetlets and jets. In keeping with prior observational and numerical studies of impulsive coronal activity, we conclude that quasiperiodic jets generated by interchange/ breakout reconnection at CH bright points and plume bases are the most likely sources of the microstreams/ switchbacks observed in the solar wind.

Pankaj Kumar↗

NASA NTRS2023

LunaNet Position, Navigation, and Timing Services and Signal, Enabling the Future of Lunar Exploration

The International Space Exploration Coordination Group established in 2018 the 3rd edition of the Global Exploration Roadmap (ISECG, 2018) that aims to achieve Mars human surface activities and identifies the exploration of the Moon as a critical intermediate step. A supplement covering updates on surface exploration scenarios was released in 2020 (ISECG, 2020). The Artemis Accords (NASA Artemis, 2020), first signed in October 2020, now includes over two dozen nations, in an agreement on the principles for best practices, including interoperability. September 2022 introduced the National Aeronautics and Space Administration’s (NASA) Moon to Mars Objectives highlighting recurring tenets of collaboration with international and industry partners and interoperability, along with infrastructure objectives for Position, Navigation, and Timing (PNT). The successful Artemis 1 mission paved the way to the ambitious plans to establish a sustainable human presence on the Moon. Just a few months after Artemis 1 launch (NASA, 2022), iSpace HAKUTO-R Mission1 (iSpace, 2022) launched, being the first-ever commercial mission, launched by a commercial launch service provider, aiming to land on the lunar surface. The NASA Artemis programme plans initial crewed landings and traverses in 2025, supported by the Lunar Gateway, followed by regular launches to build the lunar systems (NASA Artemis Plan, 2020), (NASA, 2022). NASA’s contracts with Commercial Lunar Payload Services (CLPS, (NASA, n.d.)) to deliver science and technology to the Moon, with launches starting in November 2023. The European Space Agency (ESA) Argonaut (ESA Argonaut, 2022) programme plans to have recurrent missions to bring payloads to the lunar surface, supporting lunar exploration. These are just a few examples of planned missions that will target the Earth’s natural satellite in the next decade, with forecasts of tens of missions per year (NSR, 2022), (Euroconsult, 2020). The large number of missions and the complexity of landing and operating are expected to demand a change of paradigm from the current Earth-based communication and navigation services. In recent years, several agencies have proposed to deploy cislunar communication and navigation services to support lunar missions (NASA LCRNS, 2022), (ESA Moonlight, 2022), (JAXA, 2022)). All these proposals seek to deploy service-providing satellites in lunar orbit to ease the user missions’ operations. The PNT services objective is to support all types of lunar users (e.g.: orbiters, landers, ascent vehicles, and surface crew and rovers). At the same time, NASA and ESA initiated an effort to define a common framework to ensure interoperability among different service providers: the LunaNet framework. The LunaNet Interoperability Specification (NASA and ESA, 2023) covers communication, PNT, and auxiliary services, by establishing a common set of requirements to ensure interoperability. This conference contribution will present the LunaNet PNT services, focusing on the Lunar Augmented Navigation Service (LANS) that resembles the Global Navigation Satellite System (GNSS) concept on Earth: constellations of satellites broadcasting a radio navigation signal synchronized to a common reference clock, with augmentations to accommodate users’ needs in an environment away from Earth. This includes a description of the high-level LANS concept and the basic principles defined to ensure interoperability. In addition, it will describe the common S-band PNT Augmented Forward Signal (AFS) and common messages to be adopted for compliance with the LunaNet framework, and the justification of the selected approach.

LunaNet↗

NASA NTRS2023

Stiffness and Fatigue Life Estimator for Polymer Composite Laminates Using Machine Learning

Machine learning (ML) models are increasingly being used in many engineering fields due to the advancements in ML algorithms and availability of high-speed computing power. One of the most popular ML class of models is artificial neural networks (ANN). ML is increasingly being used in the design and analysis of composite materials and structures, specifically in the constitutive modeling of composite materials with the focus on greatly accelerating multiscale analyses of composite materials and structures through development of surrogate models. Towards that end, Python-based neural nets have been developed to predict initial stiffness and fatigue life of an eight-ply symmetric polymer matrix composite laminate. Two types of neural networks, a Multilayer Perceptron (MLP) and a Recurrent Neural Network (RNN), have been established. Results show that both neural net type algorithms can provide an excellent estimate of initial laminate stiffness as well as fatigue life of eight-ply symmetric polymer matrix composite laminates (PMCs). RNNs are better able to capture the shape of the fatigue curve of a laminate. The resulting tool and GUI can be very useful for system level studies to obtain an estimate of desired properties and life of PMC composite laminates. Further, the associated surrogate models can also be used in composite multiscale analyses to replace the actual physics-based calculations at lower scales and thereby significantly increase the computational efficiency of such analyses and thus make micromechanics-based multiscale analyses a viable industrial tool for large scale structural problems.

multiscale analysis↗

NASA NTRS2023

Lunanet Position, Navigation, and Timing Services and Signals, Enabling the Future of Lunar Exploration

The International Space Exploration Coordination Group established in 2018 the 3rd edition of the Global Exploration Roadmap (ISECG, 2018) that aims to achieve Mars human surface activities and identifies the exploration of the Moon as a critical intermediate step. A supplement covering updates on surface exploration scenarios was released in 2020 (ISECG, 2020). The Artemis Accords (NASA Artemis, 2020), first signed in October 2020, now includes over two dozen nations, in an agreement on the principles for best practices, including interoperability. In September 2022 the National Aeronautics and Space Administration (NASA) introduced the Moon to Mars Objectives highlighting recurring tenets of collaboration with international and industry partners and interoperability, along with infrastructure objectives for Position, Navigation, and Timing (PNT). The successful Artemis 1 mission paved the way to the ambitious plans to establish a sustainable human presence on the Moon. Just a few months after Artemis 1 launch (NASA, 2022), iSpace HAKUTO-R Mission1 (iSpace, 2022) launched, being the first-ever mission launched by a commercial launch service provider aiming to land on the lunar surface. The NASA Artemis program plans initial crewed landings and surface traverses in 2025, supported by the Lunar Gateway. Regular launches will follow to build the lunar systems for a sustained presence as presented in the Artemis Plan (NASA Artemis Plan, 2020), (NASA, 2022). NASA’s contracts with commercial providers through the Commercial Lunar Payload Services program (CLPS, (NASA, n.d.)) will deliver science and technology demonstration missions to the Moon starting in November 2023. The European Space Agency (ESA) Argonaut (ESA Argonaut, 2022) program plans to have recurrent missions to bring payloads to the lunar surface, supporting lunar exploration. These are just a few examples of planned missions that will target Earth’s natural satellite in the next decade, with forecasts of tens of missions per year (NSR, 2022), (Euroconsult, 2020). The large number of missions and the complexity of landing and operating are expected to demand a change of paradigm from the current Earth-based communication and navigation services, that may be combined with onboard sensors. In recent years, several agencies have proposed to deploy cislunar communication and navigation services to support lunar missions (NASA LCRNS, 2022), (ESA Moonlight, 2022), (JAXA, 2022)). All these proposals seek to deploy service-providing satellites in lunar orbit to ease the user missions’ operations. The PNT services objective is to support all types of lunar users (e.g.: orbiters, landers, ascent vehicles, surface crew, rovers, and deployed science payloads). At the same time, NASA and ESA initiated an effort to define a common framework to ensure interoperability among different service providers: the LunaNet framework. The LunaNet Interoperability Specification (NASA and ESA, 2023) covers communication, PNT, and auxiliary services, by establishing a common set of requirements to ensure interoperability. This conference contribution will present the LunaNet PNT services, focusing on the Lunar Augmented Navigation Service (LANS) that would be provided by a system that resembles the Global Navigation Satellite System (GNSS) concept on Earth: constellations of satellites broadcasting a radio navigation signal synchronized to a common reference clock, with augmentations to accommodate users’ needs in an environment away from Earth. This paper includes a description of the high-level LANS concept, and the basic principles defined to ensure interoperability. In addition, it will describe the common S-band PNT Augmented Forward Signal (AFS) and common messages to be adopted for compliance with the LunaNet framework, and the justification of the selected approach.

LunaNet↗

NASA NTRS2023

Self-Similar Outflows at the Source of the Fast Solar Wind: A Smoking Gun of Multiscale Impulsive Reconnection?

We present results of a quantitative analysis of structured plasma outflows above a polar coronal hole observed by the Atmospheric Imaging Assembly (AIA) on board the Solar Dynamics Observatory (SDO) spacecraft. In a 6 hr interval of continuous high-cadence SDO/AIA images, we identified more than 2300 episodes of small-scale plasma flows in the polar corona. The mean upward flow speed measured by the surfing transform technique is estimated to be 122 ± 34 km s−1, which is comparable to the local sound speed. The typical recurrence period of the flow episodes is 10–30 minutes, and the mean duration and transverse size of each episode are about 3–5 minutes and 3–4 Mm, respectively. The largest identifiable episodes last for tens of minutes and reach widths up to 40 Mm. For the first time, we demonstrate that the polar coronal-hole outflows obey a family of power-law probability distributions characteristic of impulsive interchange magnetic reconnection. Turbulent photospheric driving may play a crucial role in releasing magnetically confined plasma onto open field. The estimated occurrence rate of the detected self-similar coronal outflows is sufficient for them to make a dominant contribution to the fast-wind mass and energy fluxes and to account for the wind's small-scale structure.

jets↗

NASA NTRSDate not supplied

Classification of Notices to Airmen using Natural Language Processing

This paper establishes the feasibility of using Natural Language Processing (NLP) to classify NOTAMs or Notices to Airmen – a pilot messaging framework to gather real-time situational awareness. Present day air mobility operations heavily rely on NOTAMs. However, pilots often have difficulty interpreting NOTAMs due to the sheer volume of inapplicable messages and unclear abbreviations. Using NLP, the presented study analyzes the accuracy of classifying NOTAMs and, thereby, the efficiency of generating actionable interpretations in real time. To this effect, efficacies of four NLP neural network architectures were analyzed, including three Recurrent Neural Networks (RNNs) with GloVe, Word2Vec, and FastText word embeddings, and one trained Bi-Directional Encoder Representations from Transformers (BERT) model. The four neural networks were trained and evaluated on three open-source datasets of varying text lengths, vocabularies, and grammars, taken from e-commerce product descriptions, social media tweets, and unstructured descriptions for data and analytics services on open data marketplaces such as NASA’s Data and Reasoning Fabric (DRF) platform. This provided cross-analysis of each neural network architecture’s performance per text type. The best performing architecture, BERT, was then fine-tuned on a collection of open-source NOTAM data. Post-training, a real-time NOTAM classification service was implemented to draw inference on new NOTAMs using the trained model, which demonstrated close to 99% accuracy in classification. This modular classification service is envisioned to be integrated with a data and analytics delivery platform, such as the DRF, thus availing real-time contextualization of NOTAMs to air mobility clients, humans, and machines for enhanced decision making.

Aiden C. Szeto↗

NASA NTRSDate not supplied

Comparison Study of Machine Learning Techniques to Predict Flight Energy Consumption for Advanced Air Mobility

This paper addresses the need to predict the flight energy consumption of aerial vehicles in the presence of wind using machine learning techniques. The presented work is critical to achieving sustainable and efficient operations for Advanced Air Mobility (AAM) and to evaluating the readiness of the ground-supporting energy infrastructure, e.g., electric grid and AAM portals. The flight energy consumption is described using the "energy per meter" (EPM) metric. We present a comparison study of influential machine learning techniques in predicting EPM using real-world flight test data. We presented new results of using the Decision Tree, Random Forest, and linear regression techniques, along with our previous results using the Recurrent Neural Network and Feed Forward Neural Network techniques. The comparison results show that the Linear Regression method outperforms other methods on the basis of the Mean Squared Error and error variance.

Machine Learning↗

NASA NTRSDate not supplied

Classification of Notices to Airmen using Natural Language Processing

Aiden Szeto↗

NASA NTRSDate not supplied

Impact of Pacific Ocean Heatwaves on Phytoplankton Community Composition

Since 2013, marine heatwaves have become recurrent throughout the equatorial and northeastern Pacific Ocean and are expected to increase in intensity relative to historic norms. Among the ecological ramifications associated with these high temperature anoma- lies are increased mortality of higher trophic organisms such as marine mammals and sea- birds, which are likely triggered by changes in the composition of phytoplankton, the base of the marine trophic food web. Here, we assimilated satellite ocean color data into an ocean biogeochemical model to describe changes in the abundance of phytoplankton functional types (PFTs) during the last decade’s (2010s) warm anomalies in the equatorial and northeastern Pacific Ocean. We find important changes associated with the “Blob” warm anomaly in the Gulf of Alaska, where reduced silica supply led to a switch in community composition from diatoms to dinoflagellates, resulting in an increase in surface ocean chlorophyll during the Summer–Fall of 2014. A more dramatic change was observed in the equatorial Pacific, where the extreme warm conditions of the 2016 El Niño resulted in a major decline of about 40% in surface chlorophyll, which was associated with a nearly total collapse in diatoms.

Lionel A. Arteaga↗

NASA NTRSDate not supplied

Predictive Workload Model for Air Traffic Controllers during UAM Operations

The effect of airspace factors on air traffic controller (ATC) workload has been an active area of study for almost three decades due to the importance of safety considerations necessary to design and maintain operations. Existing literature has examined several traffic-related (e.g., number of aircraft under control, loss of separation) contributors to ATC workload and proposed mathematical functions to best describe controller response. However, future air traffic continues to increase in complexity with the introduction of urban air mobility (UAM) – or the transportation of humans and cargo using electric vertical takeoff and landing (eVTOL) aircraft. UAM aims to alleviate congestion for existing ground transportation systems and improve mobility within urban centers and other high-demand locations. This shift in the traditional airspace paradigm necessitates an evolved understanding of model use and development for ATC workload prediction. This study aimed to develop an ATC workload forecasting model based on human-in-the-loop (HITL) simulation data for UAM operations at large airports. Data collected from the HITL simulation served as the training and testing data for a Long Short-Term Memory recurrent neural network and enabled time-series forecasting of ATC workload from traffic characteristics. Results demonstrated the potential of LSTM models for forecasting ATC workload 40 minutes into the future and highlighted important considerations for future development.

predictive model↗

NASA NTRS2022

Revisiting the Ground Magnetic Field Perturbations Challenge: A Machine Learning Perspective

Forecasting ground magnetic field perturbations has been a long-standing goal of the space weather community. The availability of ground magnetic field data and its potential to be used in geomagnetically induced current studies, such as risk assessment, have resulted in several forecasting efforts over the past few decades. One particular community effort was the Geospace Environment Modeling (GEM) challenge of ground magnetic field perturbations that evaluated the predictive capacity of several empirical and first principles models at both mid- and high-latitudes in order to choose an operative model. In this work, we use three different deep learning models-a feed-forward neural network, a long short-term memory recurrent network and a convolutional neural network-to forecast the horizontal component of the ground magnetic field rate of change (dB H /dt) over 6 different ground magnetometer stations and to compare as directly as possible with the original GEM challenge. We find that, in general, the models are able to perform at similar levels to those obtained in the original challenge, although the performance depends heavily on the particular storm being evaluated. We then discuss the limitations of such a comparison on the basis that the original challenge was not designed with machine learning algorithms in mind.

Victor A. Pinto↗