arXiv subjects

Marcus Freitag

Publications and source records attributed to Marcus Freitag.

At least 19 recordsLinked to original sources

arXiv2026

PDE foundation models are skillful AI weather emulators for the Martian atmosphere

We show that AI foundation models that are pretrained on numerical solutions to a diverse corpus of partial differential equations can be adapted and fine-tuned to obtain skillful predictive weather emulators for the Martian atmosphere. We base our work on the Poseidon PDE foundation model for two-dimensional systems. We develop a method to extend Poseidon from two to three dimensions while keeping the pretraining information. Moreover, we investigate the performance of the model in the presence of sparse initial conditions. Our results make use of four Martian years (approx.~34 GB) of training data and a median compute budget of 13 GPU hours. We find that the combination of pretraining and model extension yields a performance increase of 34.4\% on a held-out year. This shows that PDEs-FMs can not only approximate solutions to (other) PDEs but also anchor models for real-world problems with complex interactions that lack a sufficient amount of training data or a suitable compute budget.

cs.LG↗

arXiv2025

SuryaBench: Benchmark Dataset for Advancing Machine Learning in Heliophysics and Space Weather Prediction

This paper introduces a high resolution, machine learning-ready heliophysics dataset derived from NASA's Solar Dynamics Observatory (SDO), specifically designed to advance machine learning (ML) applications in solar physics and space weather forecasting. The dataset includes processed imagery from the Atmospheric Imaging Assembly (AIA) and Helioseismic and Magnetic Imager (HMI), spanning a solar cycle from May 2010 to July 2024. To ensure suitability for ML tasks, the data has been preprocessed, including correction of spacecraft roll angles, orbital adjustments, exposure normalization, and degradation compensation. We also provide auxiliary application benchmark datasets complementing the core SDO dataset. These provide benchmark applications for central heliophysics and space weather tasks such as active region segmentation, active region emergence forecasting, coronal field extrapolation, solar flare prediction, solar EUV spectra prediction, and solar wind speed estimation. By establishing a unified, standardized data collection, this dataset aims to facilitate benchmarking, enhance reproducibility, and accelerate the development of AI-driven models for critical space weather prediction tasks, bridging gaps between solar physics, machine learning, and operational forecasting.

astro-ph.SR↗

arXiv2025

Surya: Foundation Model for Heliophysics

Heliophysics is central to understanding and forecasting space weather events and solar activity. Despite decades of high-resolution observations from the Solar Dynamics Observatory (SDO), most models remain task-specific and constrained by scarce labeled data, limiting their capacity to generalize across solar phenomena. We introduce Surya, a 366M parameter foundation model for heliophysics designed to learn general-purpose solar representations from multi-instrument SDO observations, including eight Atmospheric Imaging Assembly (AIA) channels and five Helioseismic and Magnetic Imager (HMI) products. Surya employs a spatiotemporal transformer architecture with spectral gating and long--short range attention, pretrained on high-resolution solar image forecasting tasks and further optimized through autoregressive rollout tuning. Zero-shot evaluations demonstrate its ability to forecast solar dynamics and flare events, while downstream fine-tuning with parameter-efficient Low-Rank Adaptation (LoRA) shows strong performance on solar wind forecasting, active region segmentation, solar flare forecasting, and EUV spectra. Surya is the first foundation model in heliophysics that uses time advancement as a pretext task on full-resolution SDO data. Its novel architecture and performance suggest that the model is able to learn the underlying physics behind solar evolution.

astro-ph.SR↗

arXiv2024

AI Foundation Model for Heliophysics: Applications, Design, and Implementation

Deep learning-based methods have been widely researched in the areas of language and vision, demonstrating their capacity to understand long sequences of data and their usefulness in numerous helio-physics applications. Foundation models (FMs), which are pre-trained on a large-scale datasets, form the basis for a variety of downstream tasks. These models, especially those based on transformers in vision and language, show exceptional potential for adapting to a wide range of downstream applications. In this paper, we provide our perspective on the criteria for designing an FM for heliophysics and associated challenges and applications using the Solar Dynamics Observatory (SDO) dataset. We believe that this is the first study to design an FM in the domain of heliophysics.

astro-ph.SR↗

arXiv2024

Climatic & Anthropogenic Hazards to the Nasca World Heritage: Application of Remote Sensing, AI, and Flood Modelling

Preservation of the Nasca geoglyphs at the UNESCO World Heritage Site in Peru is urgent as natural and human impact accelerates. More frequent weather extremes such as flashfloods threaten Nasca artifacts. We demonstrate that runoff models based on (sub-)meter scale, LiDAR-derived digital elevation data can highlight AI-detected geoglyphs that are in danger of erosion. We recommend measures of mitigation to protect the famous "lizard", "tree", and "hand" geoglyphs located close by, or even cut by the Pan-American Highway.

cs.CV↗

arXiv2023

TensorBank: Tensor Lakehouse for Foundation Model Training

Storing and streaming high dimensional data for foundation model training became a critical requirement with the rise of foundation models beyond natural language. In this paper we introduce TensorBank, a petabyte scale tensor lakehouse capable of streaming tensors from Cloud Object Store (COS) to GPU memory at wire speed based on complex relational queries. We use Hierarchical Statistical Indices (HSI) for query acceleration. Our architecture allows to directly address tensors on block level using HTTP range reads. Once in GPU memory, data can be transformed using PyTorch transforms. We provide a generic PyTorch dataset type with a corresponding dataset factory translating relational queries and requested transformations as an instance. By making use of the HSI, irrelevant blocks can be skipped without reading them as those indices contain statistics on their content at different hierarchical resolution levels. This is an opinionated architecture powered by open standards and making heavy use of open-source technology. Although, hardened for production use using geospatial-temporal data, this architecture generalizes to other use case like computer vision, computational neuroscience, biological sequence analysis and more.

cs.LG↗

arXiv2020

Event Clustering & Event Series Characterization on Expected Frequency

We present an efficient clustering algorithm applicable to one-dimensional data such as e.g. a series of timestamps. Given an expected frequency $\Delta T^{-1}$, we introduce an $\mathcal{O}(N)$-efficient method of characterizing $N$ events represented by an ordered series of timestamps $t_1,t_2,\dots,t_N$. In practice, the method proves useful to e.g. identify time intervals of "missing" data or to locate "isolated events". Moreover, we define measures to quantify a series of events by varying $\Delta T$ to e.g. determine the quality of an Internet of Things service.

cs.DC↗

arXiv2020

Learning and Recognizing Archeological Features from LiDAR Data

We present a remote sensing pipeline that processes LiDAR (Light Detection And Ranging) data through machine & deep learning for the application of archeological feature detection on big geo-spatial data platforms such as e.g. IBM PAIRS Geoscope. Today, archeologists get overwhelmed by the task of visually surveying huge amounts of (raw) LiDAR data in order to identify areas of interest for inspection on the ground. We showcase a software system pipeline that results in significant savings in terms of expert productivity while missing only a small fraction of the artifacts. Our work employs artificial neural networks in conjunction with an efficient spatial segmentation procedure based on domain knowledge. Data processing is constraint by a limited amount of training labels and noisy LiDAR signals due to vegetation cover and decay of ancient structures. We aim at identifying geo-spatial areas with archeological artifacts in a supervised fashion allowing the domain expert to flexibly tune parameters based on her needs.

cs.CV↗

arXiv2014

Substrate Sensitive Mid-Infrared Photoresponse in Graphene

We report mid-infrared photocurrent spectra of graphene nanoribbon arrays on SiO2 dielectrics showing dual signatures of the substrate interaction. First, hybrid polaritonic modes of graphene plasmons and dielectric surface polar phonons produce a thermal photocurrent in graphene with spectral features that are tunable by gate voltage, nanoribbon width, and light polarization. Secondly, phonon-polaritons associated with the substrate are excited, which indirectly heat up the graphene leading to a graphene photocurrent with fixed spectral features. Models for other commonly used substrates show that the responsivity of graphene infrared photodetectors can be tailored to specific mid-IR frequency bands by the choice of the substrate.

cond-mat.mtrl-sci↗

arXiv2013

Photocurrent in graphene harnessed by tunable intrinsic plasmons

Graphene's optical properties in the infrared and terahertz can be tailored and enhanced by patterning graphene into periodic metamaterials with sub-wavelength feature sizes. Here we demonstrate polarization sensitive and gate tunable photodetection in graphene nanoribbon arrays. The long-lived hybrid plasmon-phonon modes utilized are coupled excitations of electron density oscillations and substrate (SiO2) surface polar phonons. Their excitation by s-polarization leads to an in-resonance photocurrent an order of magnitude larger than the photocurrent observed for p-polarization, which excites electron-hole pairs. The plasmonic detectors exhibit photoinduced temperature increases up to four times as large as comparable 2D graphene detectors. Moreover, the photocurrent sign becomes polarization sensitive in the narrowest nanoribbon arrays due to differences in decay channels for photoexcited hybrid plasmon-phonons and electrons. Our work provides a path to light sensitive and frequency selective photodetectors based on graphene's plasmonic excitations.

cond-mat.mes-hall↗

arXiv2012

Mid-infrared plasmons in scaled graphene nanostructures

Plasmonics takes advantage of the collective response of electrons to electromagnetic waves, enabling dramatic scaling of optical devices beyond the diffraction limit. Here, we demonstrate the mid-infrared (4 to 15 microns) plasmons in deeply scaled graphene nanostructures down to 50 nm, more than 100 times smaller than the on-resonance light wavelength in free space. We reveal, for the first time, the crucial damping channels of graphene plasmons via its intrinsic optical phonons and scattering from the edges. A plasmon lifetime of 20 femto-seconds and smaller is observed, when damping through the emission of an optical phonon is allowed. Furthermore, the surface polar phonons in SiO2 substrate underneath the graphene nanostructures lead to a significantly modified plasmon dispersion and damping, in contrast to a non-polar diamond-like-carbon (DLC) substrate. Much reduced damping is realized when the plasmon resonance frequencies are close to the polar phonon frequencies. Our study paves the way for applications of graphene in plasmonic waveguides, modulators and detectors in an unprecedentedly broad wavelength range from sub-terahertz to mid-infrared.

cond-mat.mes-hall↗

arXiv2012

Cooling of photoexcited carriers in graphene by internal and substrate phonons

We investigate the energy relaxation of hot carriers produced by photoexcitation of graphene through coupling to both intrinsic and remote (substrate) surface polar phonons using the Boltzmann equation approach. We find that the energy relaxation of hot photocarriers in graphene on commonly used polar substrates, under most conditions, is dominated by remote surface polar phonons. We also calculate key characteristics of the energy relaxation process, such as the transient cooling time and steady state carrier temperatures and photocarriers densities, which determine the thermoelectric and photovoltaic photoresponse, respectively. Substrate engineering can be a promising route to efficient optoelectronic devices driven by hot carrier dynamics.

cond-mat.mes-hall↗

arXiv2012

Manipulating infrared photons using plasmons in transparent graphene superlattices

Superlattices are artificial periodic nanostructures which can control the flow of electrons. Their operation typically relies on the periodic modulation of the electric potential in the direction of electron wave propagation. Here we demonstrate transparent graphene superlattices which can manipulate infrared photons utilizing the collective oscillations of carriers, i.e., plasmons of the ensemble of multiple graphene layers. The superlattice is formed by depositing alternating wafer-scale graphene sheets and thin insulating layers, followed by patterning them all together into 3-dimensional photonic-crystal-like structures. We demonstrate experimentally that the collective oscillation of Dirac fermions in such graphene superlattices is unambiguously nonclassical: compared to doping single layer graphene, distributing carriers into multiple graphene layers strongly enhances the plasmonic resonance frequency and magnitude, which is fundamentally different from that in a conventional semiconductor superlattice. This property allows us to construct widely tunable far-infrared notch filters with 8.2 dB rejection ratio and terahertz linear polarizers with 9.5 dB extinction ratio, using a superlattice with merely five graphene atomic layers. Moreover, an unpatterned superlattice shields up to 97.5% of the electromagnetic radiations below 1.2 terahertz. This demonstration also opens an avenue for the realization of other transparent mid- and far-infrared photonic devices such as detectors, modulators, and 3-dimensional meta-material systems.

cond-mat.mes-hall↗

arXiv2012

Photoconductivity of biased graphene

Graphene is a promising candidate for optoelectronic applications such as photodetectors, terahertz imagers, and plasmonic devices. The origin of photoresponse in graphene junctions has been studied extensively and is attributed to either thermoelectric or photovoltaic effects. In addition, hot carrier transport and carrier multiplication are thought to play an important role. Here we report the intrinsic photoresponse in biased but otherwise homogeneous graphene. In this classic photoconductivity experiment, the thermoelectric effects are insignificant. Instead, the photovoltaic and a photo-induced bolometric effect dominate the photoresponse due to hot photocarrier generation and subsequent lattice heating through electron-phonon cooling channels respectively. The measured photocurrent displays polarity reversal as it alternates between these two mechanisms in a backgate voltage sweep. Our analysis yields elevated electron and phonon temperatures, with the former an order higher than the latter, confirming that hot electrons drive the photovoltaic response of homogeneous graphene near the Dirac point.

cond-mat.mes-hall↗

arXiv2011

Infrared Spectroscopy of Wafer-Scale Graphene

We report on spectroscopy results from the mid- to far-infrared on wafer-scale graphene, grown either epitaxially on silicon carbide, or by chemical vapor deposition. The free carrier absorption (Drude peak) is simultaneously obtained with the universal optical conductivity (due to interband transitions), and the wavelength at which Pauli blocking occurs due to band filling. From these the graphene layer number, doping level, sheet resistivity, carrier mobility, and scattering rate can be inferred. The mid-IR absorption of epitaxial two-layer graphene shows a less pronounced peak at 0.37\pm0.02 eV compared to that in exfoliated bilayer graphene. In heavily chemically-doped single layer graphene, a record high transmission reduction due to free carriers approaching 40% at 250 \mum (40 cm-1) is measured in this atomically thin material, supporting the great potential of graphene in far-infrared and terahertz optoelectronics.

cond-mat.mes-hall↗

arXiv2011

Layer Number Determination and Thickness-dependent Properties of Graphene Grown on SiC

The electronic properties of few-layer graphene grown on the carbon-face of silicon carbide (SiC) are found to be strongly dependent on the number of layers. The carrier mobility is larger in thicker graphene because substrate-related scattering is reduced in the higher layers. The carrier density dependence of the mobility is qualitatively different in thin and thick graphene, with the transition occurring at about 2 layers. The mobility increases with carrier density in thick graphene, similar to multi-layer graphene exfoliated from natural graphite, suggesting that the individual layers are still electrically coupled in spite of reports recording non-Bernal stacking order in C-face grown graphene. The Hall coefficient peak value is reduced in thick graphene due to the increased density of states. A reliable and rapid characterization tool for the layer number is therefore highly desirable. To date, AFM height determination and Raman scattering are typically used since the optical contrast of graphene on SiC is weak. However, both methods suffer from low throughput. We show that the scanning electron microscopy (SEM) contrast can give similar results with much higher throughput.

cond-mat.mes-hall↗

arXiv2010

Wafer-scale Epitaxial Graphene Growth on the Si-face of Hexagonal SiC (0001) for High Frequency Transistors

Up to two layers of epitaxial graphene have been grown on the Si-face of two-inch SiC wafers exhibiting room-temperature Hall mobilities up to 1800 cm^2/Vs, measured from ungated, large, 160 micron x 200 micron Hall bars, and up to 4000 cm^2/Vs, from top-gated, small, 1 micron x 1.5 micron Hall bars. The growth process involved a combination of a cleaning step of the SiC in a Si-containing gas, followed by an annealing step in Argon for epitaxial graphene formation. The structure and morphology of this graphene has been characterized using AFM, HRTEM, and Raman spectroscopy. Furthermore, top-gated radio frequency field effect transistors (RF-FETs) with a peak cutoff frequency fT of 100 GHz for a gate length of 240 nm were fabricated using epitaxial graphene grown on the Si face of SiC that exhibited Hall mobilities up to 1450 cm^2/Vs from ungated Hall bars and 1575 cm^2/Vs from top-gated ones. This is by far the highest cut-off frequency measured from any kind of graphene.

cond-mat.mtrl-sci↗

arXiv2010

Thermal infrared emission reveals the Dirac point movement in biased graphene

Graphene is a 2-dimensional material with high carrier mobility and thermal conductivity, suitable for high-speed electronics. Conduction and valence bands touch at the Dirac point. The absorptivity of single-layer graphene is 2.3%, nearly independent of wavelength. Here we investigate the thermal radiation from biased graphene transistors. We find that the emission spectrum of single-layer graphene follows that of a grey body with constant emissivity (1.6 \pm 0.8)%. Most importantly, we can extract the temperature distribution in the ambipolar graphene channel, as confirmed by Stokes/anti-Stokes measurements. The biased graphene exhibits a temperature maximum whose location can be controlled by the gate voltage. We show that this peak in temperature reveals the spatial location of the minimum in carrier density, i.e. the Dirac point.

cond-mat.mes-hall↗