Search

An emerging trend in deep learning research focuses on the applications of graph neural networks (GNNs) for mesh-based continuum mechanics simulations. Most of these learning frameworks operate on graphs wherein each edge connects two nodes. Inspired by the data connectivity in the finite element method, we present a method to construct a hypergraph by connecting the nodes by elements rather than edges. A hypergraph message-passing network is defined on such a node-element hypergraph that mimics the calculation process of local stiffness matrices. We term this method a finite element-inspired hypergraph neural network, in short FEIH($\phi$)-GNN. We further equip the proposed network with rotation equivariance, and explore its capability for modeling unsteady fluid flow systems. The effectiveness of the network is demonstrated on two common benchmark problems, namely the fluid flow around a circular cylinder and airfoil configurations. Stabilized and accurate temporal roll-out predictions can be obtained using the $\phi$-GNN framework within the interpolation Reynolds number range. The network is also able to extrapolate moderately towards higher Reynolds number domain out of the training range.

Direct imaging of gas giant exoplanets provides key information on planetary atmospheres and the architectures of planetary systems. However, few planets have been detected in blind surveys used to achieve imaging detections. Using Gaia and Hipparcos astrometry we identified dynamical evidence for a gas giant planet around the nearby star HIP 99770 and then confirmed this planet by direct imaging with the Subaru Coronagraphic Extreme Adaptive Optics Project. HIP 99770 b orbits 17 astronomical units from its host star, with an insolation comparable to Jupiter's and a dynamical mass of 13.9--16.1 Jupiter masses. Its planet-to-star mass ratio (7--8$\times$10$^{-3}$) is comparable to that other directly-imaged planets. The planet's atmosphere resembles an older, less-cloudy analogue of the atmospheres of previously-imaged exoplanets around HR 8799., Comment: Authors' version of revised paper, published in Science on April 14, 2023 (passed independent peer review/was recommended for publication in Science by external referees on July 5, 2022). First joint direct imaging + astrometric discovery of an exoplanet. 49 pages, 18 figures, 6 tables

We study the empirical process arising from a multi-dimensional diffusion process with periodic drift and diffusivity. The smoothing properties of the generator of the diffusion are exploited to prove the Donsker property for certain classes of smooth functions. We partially generalise the finding from the one-dimensional case studied in [van der Vaart & van Zanten, 2005]: that the diffusion empirical process exhibits stronger regularity than in the classical case of i.i.d. observations. As an application, precise asymptotics are deduced for the Wasserstein-1 distance between the time-$T$ occupation measure and the invariant measure in dimensions $d\leq3$., Comment: 14 pages

Yrast and near-yrast states above the known 25/2$^{+}$ isomer in $^{207}$At are established for the first time. The level scheme is extended up to 47/2$\hbar$ and 6.5 MeV with the addition of about 60 new $\gamma$-ray transitions. The half-life of the 25/2$^{+}$ isomer is revisited and a value of $T_{1/2}$ = 107.5(9) ns is deduced. Evidence of a hitherto unobserved 29/2$^{+}$ isomer in $^{207}$At is presented. A systematic study of $B(E3)$ values for the transitions de-exciting the 29/2$^{+}$ isomer in the neighboring odd-$A$ At isotopes suggests a half-life in the 2$-$4.5 $\mu$s range for this state in $^{207}$At. The experimental results are compared with large-scale shell-model calculations performed using the KHM3Y effective interaction in the $Z$ = 50$-$126, $N$ = 82$-$184 model space and an overall good agreement is noted between the theory and the experiment. A qualitative comparison of the excited states and the isomers with analogous states in neighboring nuclei provides further insight into the structure of $^{207}$At., Comment: 16 pages, 14 figures

There is a critical need for efficient and reliable active flow control strategies to reduce drag and noise in aerospace and marine engineering applications. While traditional full-order models based on the Navier-Stokes equations are not feasible, advanced model reduction techniques can be inefficient for active control tasks, especially with strong non-linearity and convection-dominated phenomena. Using convolutional recurrent autoencoder network architectures, deep learning-based reduced-order models have been recently shown to be effective while performing several orders of magnitude faster than full-order simulations. However, these models encounter significant challenges outside the training data, limiting their effectiveness for active control and optimization tasks. In this study, we aim to improve the extrapolation capability by modifying network architecture and integrating coupled space-time physics as an implicit bias. Reduced-order models via deep learning generally employ decoupling in spatial and temporal dimensions, which can introduce modeling and approximation errors. To alleviate these errors, we propose a novel technique for learning coupled spatial-temporal correlation using a 3D convolution network. We assess the proposed technique against a standard encoder-propagator-decoder model and demonstrate a superior extrapolation performance. To demonstrate the effectiveness of 3D convolution network, we consider a benchmark problem of the flow past a circular cylinder at laminar flow conditions and use the spatio-temporal snapshots from the full-order simulations. Our proposed 3D convolution architecture accurately captures the velocity and pressure fields for varying Reynolds numbers. Compared to the standard encoder-propagator-decoder network, the spatio-temporal-based 3D convolution network improves the prediction range of Reynolds numbers outside of the training data., Comment: 14 pages

A nonlinear electronic circuit comprising of three nodes with a feedback loop is analyzed. The system has two stable states, a uniform state and a sinusoidal oscillating state, and it transitions from one to another by means of a Hopf bifurcation. The stability of this system is analyzed with nonlinear equations derived from a repressilator-like transistor circuit. The apparatus is simple and inexpensive, and the experiment demonstrates aspects of nonlinear dynamical systems in an advanced undergraduate laboratory setting.

Tunneling, though a physical reality, is shrouded in mystery. Wave packets cannot be constructed under the barrier and group velocity cannot be defined. The tunneling particle can be observed on either sides of the barrier but its properties under the barrier has never been probed due to several problems related to quantum measurement. We show that there are ways to bypass these problems in mesoscopic systems and one can even derive an expression for the quantum mechanical current under the barrier. A general scheme is developed to derive this expression for any arbitrary system. One can use mesoscopic phenomenon to subject the expression to several theoretical and experimental cross checks. For demonstration we consider an ideal 1D quantum ring with Aharonov-Bohm flux $\Phi$, connected to a reservoir. It gives clear evidence that propagation occur under the barrier resulting in a current that can be measured non-invasively and theoretically cross checked. Time reversed states play a role but there is no evidence of violation of causality. The evanescent states are known to be largely stable and robust against phase fluctuations making them a possible candidate for device applications and so formalizing current under barrier is important., Comment: 13 pages, 5 figures

After 16 years of on-sky operation, Subaru Telescope's facility adaptive optics AO188 is getting several major upgrades to become the extreme-AO AO3000 (3000 actuators in the pupil compared to 188 previously). AO3000 will provide high-Strehl images for several instruments from visible to mid-infrared, notably the Infrared Camera and Spectrograph (IRCS), and the Subaru Coronagraphic Extreme Adaptive Optics (SCExAO). For this upgrade, the original 188-element deformable mirror (DM) will be replaced with ALPAO's $64\times64$ DM. The visible wavefront sensor will also be upgraded at a later date, but in the meantime we are adding a near-infrared Wavefront Sensor (NIR WFS), using either a double roof prism pyramid mode or a focal plane WFS mode. This new wavefront sensor will use for the first time First Light's C-RED ONE camera, allowing for a full control of the $64\times64$ DM at up to 1.6 kHz. One of the challenges is the use of non-destructive reads and a rolling shutter with the modulated pyramid. This upgrade will be particularly exciting for SCExAO, since the extreme-AO loop will focus more on creating high-contrast dark zones instead of correcting large atmospheric residuals. It will be the first time two extreme-AO loops will be combined on the same telescope. Finally, the setup AO3000+SCExAO+IRCS will serve as a perfect demonstrator for the Thirty Meter Telescope's Planetary Systems Imager (TMT-PSI). We will present here the design, integration and testing of AO3000, and show the first on-sky results., Comment: 14 pages, 12 figures, SPIE Astronomical Telescopes +Instrumentation, 2022, Montreal, Quebec, Canada

We present an experimental study of photon statistics for high-contrast imaging with the Microwave Kinetic Inductance Detector (MKID) Exoplanet Camera (MEC) located behind the Subaru Coronagraphic Extreme Adaptive Optics System (SCExAO) at the Subaru Telescope. We show that MEC measures the expected distributions for both on-axis companion intensity and off-axis intensity which manifests as quasi-static speckles in the image plane and currently limits high-contrast imaging performance. These statistics can be probed by any MEC observation due to the photon-counting capabilities of MKID detectors. Photon arrival time statistics can also be used to directly distinguish companions from speckles using a post-processing technique called Stochastic Speckle Discrimination (SSD). Here, we we give an overview of the SSD technique and highlight the first demonstration of SSD on an extended source -- the protoplanetary disk AB Aurigae. We then present simulations that provide an in-depth exploration as to the current limitations of an extension of the SSD technique called Photon-Counting SSD (PCSSD) to provide a path forward for transitioning PCSSD from simulations to on-sky results. We end with a discussion of how to further improve the efficacy of such arrival time based post-processing techniques applicable to both MKIDs, as well as other high speed astronomical cameras., Comment: 11 pages, 9 figures, accepted for publication in AJ

Low wind and petaling effects, caused by the discontinuous apertures of telescopes, are poorly corrected -- if at all -- by commonly used workhorse wavefront sensors (WFSs). Wavefront petaling breaks the coherence of the point spread function core, splitting it into several side lobes, dramatically shutting off scientific throughput. We demonstrate the re-purposing of non-redundant sparse aperture masking (SAM) interferometers into low-order WFSs complementing the high-order pyramid WFS, on the SCExAO experimental platform at Subaru Telescope. The SAM far-field interferograms formed from a 7-hole mask are used for direct retrieval of petaling aberrations, which are almost invisible to the main AO loop. We implement a visible light dual-band SAM mode, using two disjoint 25 nm wide channels, that we recombine to overcome the one-lambda ambiguity of fringe-tracking techniques. This enables a control over petaling with sufficient capture range yet without conflicting with coronagraphic modes in the near-infrared. We present on-sky engineering results demonstrating that the design is able to measure petaling well beyond the range of a single-wavelength equivalent design., Comment: Proc. SPIE 12185 (Adaptive Optics Systems VIII), 285-297

We describe the design and initial results from a visible-light Lyot coronagraph for SCExAO/VAMPIRES. The coronagraph is comprised of four hard-edged, partially transmissive focal plane masks with inner working angles of 36 mas, 55 mas, 92 mas, and 129 mas, respectively. The Lyot stop is a reflective, undersized design with a geometric throughput of 65.7%. Our preliminary on-sky contrast is 1e-2 at 0.1" to 1e-4 at 0.75" for all mask sizes. The coronagraph was deployed in early 2022 and is available for open use., Comment: Proceedings of SPIE 2022 Astronomical Instrumentation and Telescopes conference (#12184-163)

Adaptive Optics projects at Subaru Telescope span a wide field of capabilities ranging from ground-layer adaptive optics (GLAO) providing partial correction over a 20 arcmin FOV to extreme adaptive optics (ExAO) for exoplanet imaging. We describe in this paper current and upcoming narrow field-of-view capabilities provided by the Subaru Extreme Adaptive Optics Adaptive Optics (SCExAO) system and its instrument modules, as well as the upcoming 3000-actuator upgrade of the Nasmyth AO system., Comment: 11 pages, to appear in SPIE Proceedings of Astronomical Telescopes + Instrumentation, 2022

Speckle Noise is the dominant source of error in high contrast imaging with adaptive optics system. We discuss the potential for wavefront sensing telemetry to calibrate speckle noise with sufficient precision and accuracy so that it can be removed in post-processing of science images acquired by high contrast imaging instruments. In such a self-calibrating system, exoplanet detection would be limited by photon noise and be significantly more robust and efficient than in current systems. We show initial laboratory and on-sky tests, demonstrating over short timescale that residual speckle noise is indeed calibrated to an accuracy exceeding readout and photon noise in the high contrast region. We discuss immplications for the design of space and ground high-contrast imaging systems., Comment: 12 pages, 7 figures, To appear in SPIE Proceedings of Astronomical Telescopes + Instrumentation, 2022. arXiv admin note: text overlap with arXiv:2109.13958

We use deformation theory to study the big Hecke algebra acting on mod-2 modular forms of prime level $N$ and all weights, especially its local component at the trivial representation. For $N = 3, 5$, we prove that the maximal reduced quotient of this big Hecke algebra is isomorphic to the maximal reduced quotient of the corresponding universal deformation ring. Then we completely determine the structure of this big Hecke algebra. We also describe a natural grading on mod-$p$ Hecke algebras., Comment: Final version: 40 Pages, made some minor changes following referee's suggestions. Comments are Welcome!

We present the direct imaging discovery of a low-mass companion to the nearby accelerating F star, HIP 5319, using SCExAO coupled with the CHARIS, VAMPIRES, and MEC instruments in addition to Keck/NIRC2 imaging. CHARIS $JHK$ (1.1-2.4 $\mu$m) spectroscopic data combined with VAMPIRES 750 nm, MEC $Y$, and NIRC2 $L_{\rm p}$ photometry is best matched by an M3--M7 object with an effective temperature of T=3200 K and surface gravity log($g$)=5.5. Using the relative astrometry for HIP 5319 B from CHARIS and NIRC2 and absolute astrometry for the primary from $Gaia$ and $Hipparcos$ and adopting a log-normal prior assumption for the companion mass, we measure a dynamical mass for HIP 5319 B of $31^{+35}_{-11}M_{\rm J}$, a semimajor axis of $18.6^{+10}_{-4.1}$ au, an inclination of $69.4^{+5.6}_{-15}$ degrees, and an eccentricity of $0.42^{+0.39}_{-0.29}$. However, using an alternate prior for our dynamical model yields a much higher mass of 128$^{+127}_{-88}M_{\rm J}$. Using data taken with the LCOGT NRES instrument we also show that the primary HIP 5319 A is a single star in contrast to previous characterizations of the system as a spectroscopic binary. This work underscores the importance of assumed priors in dynamical models for companions detected with imaging and astrometry and the need to have an updated inventory of system measurements., Comment: 19 pages, 9 figuresm 5 tables

This paper investigates the use of retrospective approximation solution paradigm in solving risk-averse optimization problems effectively via importance sampling (IS). While IS serves as a prominent means for tackling the large sample requirements in estimating tail risk measures such as Conditional Value at Risk (CVaR), its use in optimization problems driven by CVaR is complicated by the need to tailor the IS change of measure differently to different optimization iterates and the circularity which arises as a consequence. The proposed algorithm overcomes these challenges by employing a univariate IS transformation offering uniform variance reduction in a retrospective approximation procedure well-suited for tuning the IS parameter choice. The resulting simulation based approximation scheme enjoys both the computational efficiency bestowed by retrospective approximation and logarithmically efficient variance reduction offered by importance sampling, Comment: 12 pages, 4 figures

Let $p \geq 5$ be a prime, $N$ be an integer not divisible by $p$, $\bar\rho_0$ be a reducible, odd and semi-simple representation of $G_{\mathbb{Q},Np}$ of dimension $2$ and $\{\ell_1,\cdots,\ell_r\}$ be a set of primes not dividing $Np$. After assuming that a certain Selmer group has dimension at most $1$, we find sufficient conditions for the existence of a cuspidal eigenform $f$ of level $N\prod_{i=1}^{r}\ell_i$ and appropriate weight lifting $\bar\rho_0$ such that $f$ is new at every $\ell_i$. Moreover, suppose $p \mid \ell_{i_0}+1$ for some $1 \leq i_0 \leq r$. Then, after assuming that a certain Selmer group vanishes, we find sufficient conditions for the existence of a cuspidal eigenform of level $N\ell_{i_0}^2 \prod_{j \neq i_0} \ell_j$ and appropriate weight which is new at every $\ell_i$ and which lifts $\bar\rho_0$. As a consequence, we prove a conjecture of Billerey--Menares in many cases., Comment: v3: 42 Pages, significantly improved the main results, incorporated the changes suggested by the referees. Comments are welcome

We present the direct-imaging discovery of a substellar companion in orbit around a Sun-like star member of the Hyades open cluster. So far, no other substellar companions have been unambiguously confirmed via direct imaging around main-sequence stars in Hyades. The star HIP 21152 is an accelerating star as identified by the astrometry from the Gaia and Hipparcos satellites. We have detected the companion, HIP 21152 B, in multi-epoch using the high-contrast imaging from SCExAO/CHARIS and Keck/NIRC2. We have also obtained the stellar radial-velocity data from the Okayama 188cm telescope. The CHARIS spectroscopy reveals that HIP 21152 B's spectrum is consistent with the L/T transition, best fit by an early T dwarf. Our orbit modeling determines the semi-major axis and the dynamical mass of HIP 21152 B to be 17.5$^{+7.2}_{-3.8}$ au and 27.8$^{+8.4}_{-5.4}$ $M_{\rm{Jup}}$, respectively. The mass ratio of HIP 21152 B relative to its host is $\approx$2\%, near the planet/brown dwarf boundary suggested from recent surveys. Mass estimates inferred from luminosity evolution models are slightly higher (33--42 $M_{\rm{Jup}}$). With a dynamical mass and a well-constrained age due to the system's Hyades membership, HIP 21152 B will become a critical benchmark in understanding the formation, evolution, and atmosphere of a substellar object as a function of mass and age. Our discovery is yet another key proof-of-concept for using precision astrometry to select direct imaging targets., Comment: 21 pages (11 pages in main body), 8 figures (4 figures in main body). Accepted for Publication in ApJL at July 9, 2022 (UT)

We develop a novel hyperspectral imaging system using structured illumination in an SLM-based Michelson interferometer. In our design, we use a reflective SLM as a mirror in one of the arms of a Michelson interferometer, and scan the interferometer by varying the phase across the SLM display. For achieving the latter, we apply a checkerboard phase mask on the SLM display where the gray value varies between 0-255, thereby imparting a dynamic phase of up to 262{\deg} to the incident light beam. We couple a supercontinuum source into the interferometer in order to mimic an astronomical object such as the Sun, and choose a central wavelength of 637.4 nm akin to the strong emission line of Fe X present in the solar spectrum. We use a bandwidth of 30 nm, and extract fringes corresponding to a spectral resolution of 3.8 nm which is limited by the reflectivity of the SLM. We also demonstrate a maximum wavelength tunability of ~8 nm by varying the phase over the phase mask with a spectral sampling of around 0.03 nm between intermediate fringes. The checkerboard phase mask can be adapted close to real time on time-scales of a few tens of milliseconds to obtain spectral information for other near-contiguous wavelengths. The compactness, potential low cost, low power requirements, real-time tunability and lack of moving mechanical parts in the setup implies that it can have very useful applications in settings which require near real-time, multi-wavelength spectroscopic applications, and is especially relevant in space astronomy., Comment: 14 pages, 8 figures, accepted for publication in Optics Express