Showing 1–50 of 64 results for author: Simon, J B

Exploring the Complex Ionization Environment of the Turbulent DM Tau Disk

Authors: Deryl E. Long, L. Ilsedore Cleeves, Fred C. Adams, Sean Andrews, Edwin A. Bergin, Viviana V. Guzmán, Jane Huang, A. Meredith Hughes, Chunhua Qi, Kamber Schwarz, Jacob B. Simon, David Wilner

Abstract: Ionization drives important chemical and dynamical processes within protoplanetary disks, including the formation of organics and water in the cold midplane and the transportation of material via accretion and magneto-hydrodynamic (MHD) flows. Understanding these ionization-driven processes is crucial for understanding disk evolution and planet formation. We use new and archival ALMA observations… ▽ More Ionization drives important chemical and dynamical processes within protoplanetary disks, including the formation of organics and water in the cold midplane and the transportation of material via accretion and magneto-hydrodynamic (MHD) flows. Understanding these ionization-driven processes is crucial for understanding disk evolution and planet formation. We use new and archival ALMA observations of HCO+, H13CO+, and N2H+ to produce the first forward-modeled 2D ionization constraints for the DM Tau protoplanetary disk. We include ionization from multiple sources and explore the disk chemistry under a range of ionizing conditions. Abundances from our 2D chemical models are post-processed using non-LTE radiative transfer, visibility sampling, and imaging, and are compared directly to the observed radial emission profiles. The observations are best fit by a modestly reduced CR ionization rate ($ζ_{CR}$ ~ 10$^{-18}$ s$^{-1}$) and a hard X-ray spectrum (hardness ratio [HR] = 0.3), which we associate with stellar flaring conditions. Our best-fit model under-produces emission in the inner disk, suggesting that there may be an additional mechanism enhancing ionization in DM Tau's inner disk. Overall, our findings highlight the complexity of ionization in protoplanetary disks and the need for high resolution multi-line studies. △ Less

Submitted 26 June, 2024; originally announced June 2024.

Comments: 18 pages, 12 figures, accepted to be published in The Astrophysical Journal (June 25, 2024)

Rapid protoplanet formation in vortices: three-dimensional local simulations with selfgravity

Authors: Wladimir Lyra, Chao-Chin Yang, Jacob B. Simon, Orkan M. Umurhan, Andrew N. Youdin

Abstract: Disk vortices, seen in numerical simulations of protoplanetary disks and found observationally in ALMA and VLA images of these objects, are promising sites for planet formation given their pebble trapping abilities. Previous works have shown strong concentration of pebbles in vortices, but gravitational collapse has only been shown in low-resolution, two-dimensional, global models. In this letter,… ▽ More Disk vortices, seen in numerical simulations of protoplanetary disks and found observationally in ALMA and VLA images of these objects, are promising sites for planet formation given their pebble trapping abilities. Previous works have shown strong concentration of pebbles in vortices, but gravitational collapse has only been shown in low-resolution, two-dimensional, global models. In this letter, we aim to study the pebble concentration and gravitational collapse of pebble clouds in vortices via high-resolution, three-dimensional, local models. We performed simulations of the dynamics of gas and solids in a local shearing box where the gas is subject to convective overstability, generating a persistent giant vortex. We find that the vortex produces objects of Moon and Mars mass, with mass function of power law $d\ln N/d\ln M=-1.6\pm 0.3$. The protoplanets grow rapidly, doubling in mass in about 5 orbits, following pebble accretion rates. The mass range and mass doubling rate are in broad agreement with previous low resolution global models. We conclude that Mars-mass planetary embryos are the natural outcome of planet formation inside the disk vortices seen in millimeter and radio images of protoplanetary disks. △ Less

Submitted 25 June, 2024; originally announced June 2024.

Comments: 9 pages, 5 figures, accepted for publication in ApJ letters

Evidence for Non-zero Turbulence in the Protoplanetary disc around IM Lup

Authors: Kevin Flaherty, A. Meredith Hughes, Jacob B. Simon, Alicia Smith Reina, Chunhua Qi, Xue-Ning Bai, Sean M. Andrews, David J. Wilner, Agnes Kospal

Abstract: The amount of turbulence in protoplanetary discs around young stars is critical for determining the efficiency, timeline, and outcomes of planet formation. It is also difficult to measure. Observations are still limited, but direct measurements of the non-thermal, turbulent gas motion are possible with the Atacama Large Millimeter/submillimeter Array (ALMA). Using CO(2-1)/$^{13}$CO(2-1)/C$^{18}$O(… ▽ More The amount of turbulence in protoplanetary discs around young stars is critical for determining the efficiency, timeline, and outcomes of planet formation. It is also difficult to measure. Observations are still limited, but direct measurements of the non-thermal, turbulent gas motion are possible with the Atacama Large Millimeter/submillimeter Array (ALMA). Using CO(2-1)/$^{13}$CO(2-1)/C$^{18}$O(2-1) ALMA observations of the disc around IM Lup at ~0.4" (~60 au) resolution we find evidence of significant turbulence, at the level of $δv_{\rm turb}=(0.18-0.30)$c$_s$. This result is robust against systematic uncertainties (e.g., amplitude flux calibration, midplane gas temperature, disc self-gravity). We find that gravito-turbulence as the source of the gas motion is unlikely based on the lack of an imprint on the rotation curve from a massive disc, while magneto-rotational instabilities and hydrodynamic instabilities are still possible, depending on the unknown magnetic field strength and the cooling timescale in the outer disc. △ Less

Submitted 11 June, 2024; originally announced June 2024.

Comments: Accepted by MNRAS, 17 pages, 12 figures

The Limited Role of the Streaming Instability During Moon and Exomoon Formation

Authors: Miki Nakajima, Jeremy Atkins, Jacob B. Simon, Alice C. Quillen

Abstract: It is generally accepted that the Moon accreted from the disk formed by an impact between the proto-Earth and impactor, but its details are highly debated. Some models suggest that a Mars-sized impactor formed a silicate melt-rich (vapor-poor) disk around Earth, whereas other models suggest that a highly energetic impact produced a silicate vapor-rich disk. Such a vapor-rich disk, however, may not… ▽ More It is generally accepted that the Moon accreted from the disk formed by an impact between the proto-Earth and impactor, but its details are highly debated. Some models suggest that a Mars-sized impactor formed a silicate melt-rich (vapor-poor) disk around Earth, whereas other models suggest that a highly energetic impact produced a silicate vapor-rich disk. Such a vapor-rich disk, however, may not be suitable for the Moon formation, because moonlets, building blocks of the Moon, of 100 m-100 km may experience strong gas drag and fall onto Earth on a short timescale, failing to grow further. This problem may be avoided if large moonlets ($\gg 100$ km) form very quickly by streaming instability, which is a process to concentrate particles enough to cause gravitational collapse and rapid formation of planetesimals or moonlets. Here, we investigate the effect of the streaming instability in the Moon-forming disk for the first time and find that this instability can quickly form $\sim 100$ km-sized moonlets. However, these moonlets are not large enough to avoid strong drag and they still fall onto Earth quickly. This suggests that the vapor-rich disks may not form the large Moon, and therefore the models that produce vapor-poor disks are supported. This result is applicable to general impact-induced moon-forming disks, supporting the previous suggestion that small planets ($<1.6 R_\oplus$) are good candidates to host large moons because their impact-induced disks would be likely vapor-poor. We find a limited role of streaming instability in a satellite formation in an impact-induced disk, whereas it plays a key role during planet formation. △ Less

Submitted 28 April, 2024; originally announced April 2024.

Comments: Accepted for Publication in PSJ

Journal ref: 2024, The Planetary Science Journal, 5, 145

Magnetically Driven Turbulence in the Inner Regions of Protoplanetary Disks

Authors: David G. Rea, Jacob B. Simon, Daniel Carrera, Geoffroy Lesur, Wladimir Lyra, Debanjan Sengupta, Chao-Chin Yang, Andrew N. Youdin

Abstract: Given the important role turbulence plays in the settling and growth of dust grains in protoplanetary disks, it is crucial that we determine whether these disks are turbulent and to what extent. Protoplanetary disks are weakly ionized near the mid-plane, which has led to a paradigm in which largely laminar magnetic field structures prevail deeper in the disk, with angular momentum being transporte… ▽ More Given the important role turbulence plays in the settling and growth of dust grains in protoplanetary disks, it is crucial that we determine whether these disks are turbulent and to what extent. Protoplanetary disks are weakly ionized near the mid-plane, which has led to a paradigm in which largely laminar magnetic field structures prevail deeper in the disk, with angular momentum being transported via magnetically launched winds. Yet, there has been little exploration on the precise behavior of the gas within the bulk of the disk. We carry out 3D, local shearing box simulations that include all three low-ionization effects (Ohmic diffusion, ambipolar diffusion, and the Hall effect) to probe the nature of magnetically driven gas dynamics 1-30 AU from the central star. We find that gas turbulence can persist with a generous yet physically motivated ionization prescription (order unity Elsasser numbers). The gas velocity fluctuations range from 0.03-0.09 of the sound speed $c_s$ at the disk mid-plane to $\sim c_s$ near the disk surface, and are dependent on the initial magnetic field strength. However, the turbulent velocities do not appear to be strongly dependent on the field polarity, and thus appear to be insensitive to the Hall effect. The mid-plane turbulence has the potential to drive dust grains to collision velocities exceeding their fragmentation limit, and likely reduces the efficacy of particle clumping in the mid-plane, though it remains to be seen if this level of turbulence persists in disks with lower ionization levels. △ Less

Submitted 10 April, 2024; originally announced April 2024.

Comments: 23 pages, 19 figures, 2 tables, submitted to ApJ

A solution for the density dichotomy problem of Kuiper Belt objects with multi-species streaming instability and pebble accretion

Authors: Manuel H. Cañas, Wladimir Lyra, Daniel Carrera, Leonardo Krapp, Debanjan Sengupta, Jacob B. Simon, Orkan M. Umurhan, Chao-Chin Yang, Andrew Youdin

Abstract: Kuiper belt objects show an unexpected trend, whereby large bodies have increasingly higher densities, up to five times greater than their smaller counterparts. Current explanations for this trend assume formation at constant composition, with the increasing density resulting from gravitational compaction. However, this scenario poses a timing problem to avoid early melting by decay of $^{26}$Al.… ▽ More Kuiper belt objects show an unexpected trend, whereby large bodies have increasingly higher densities, up to five times greater than their smaller counterparts. Current explanations for this trend assume formation at constant composition, with the increasing density resulting from gravitational compaction. However, this scenario poses a timing problem to avoid early melting by decay of $^{26}$Al. We aim to explain the density trend in the context of streaming instability and pebble accretion. Small pebbles experience lofting into the atmosphere of the disk, being exposed to UV and partially losing their ice via desorption. Conversely, larger pebbles are shielded and remain more icy. We use a shearing box model including gas and solids, the latter split into ices and silicate pebbles. Self-gravity is included, allowing dense clumps to collapse into planetesimals. We find that the streaming instability leads to the formation of mostly icy planetesimals, albeit with an unexpected trend that the lighter ones are more silicate-rich than the heavier ones. We feed the resulting planetesimals into a pebble accretion integrator with a continuous size distribution, finding that they undergo drastic changes in composition as they preferentially accrete silicate pebbles. The density and masses of large KBOs are best reproduced if they form between 15 and 22\,AU. Our solution avoids the timing problem because the first planetesimals are primarily icy, and $^{26}$Al is mostly incorporated in the slow phase of silicate pebble accretion. Our results lend further credibility to the streaming instability and pebble accretion as formation and growth mechanisms. △ Less

Submitted 8 January, 2024; originally announced January 2024.

Comments: 24 pages, 13 figures, accepted to The Planetary Science Journal

Streaming Instability and Turbulence: Conditions for Planetesimal Formation

Authors: Jeonghoon Lim, Jacob B. Simon, Rixin Li, Philip J. Armitage, Daniel Carrera, Wladimir Lyra, David G. Rea, Chao-Chin Yang, Andrew N. Youdin

Abstract: The streaming instability (SI) is a leading candidate for planetesimal formation, which can concentrate solids through two-way aerodynamic interactions with the gas. The resulting concentrations can become sufficiently dense to collapse under particle self-gravity, forming planetesimals. Previous studies have carried out large parameter surveys to establish the critical particle to gas surface den… ▽ More The streaming instability (SI) is a leading candidate for planetesimal formation, which can concentrate solids through two-way aerodynamic interactions with the gas. The resulting concentrations can become sufficiently dense to collapse under particle self-gravity, forming planetesimals. Previous studies have carried out large parameter surveys to establish the critical particle to gas surface density ratio ($Z$), above which SI-induced concentration triggers planetesimal formation. The threshold $Z$ depends on the dimensionless stopping time ($τ_s$, a proxy for dust size). However, these studies neglected both particle self-gravity and external turbulence. Here, we perform 3D stratified shearing box simulations with both particle self-gravity and turbulent forcing, which we characterize via $α_D$ that measures turbulent diffusion. We find that forced turbulence, at amplitudes plausibly present in some protoplanetary disks, can increase the threshold $Z$ by up to an order of magnitude. For example, for $τ_s = 0.01$, planetesimal formation occurs when $Z \gtrsim 0.06$, $\gtrsim 0.1$, and $\gtrsim 0.2$ at $α_D = 10^{-4}$, $10^{-3.5}$, and $10^{-3}$, respectively. We provide a single fit to the critical $Z$ as a function of $α_D$ and $τ_s$ required for the SI to work (though limited to the range $τ_s = 0.01$--0.1). Our simulations also show that planetesimal formation requires a mid-plane particle-to-gas density ratio that exceeds unity, with the critical value being independent of $α_D$. Finally, we provide the estimation of particle scale height that accounts for both particle feedback and external turbulence. △ Less

Submitted 19 December, 2023; originally announced December 2023.

Comments: 27 pages, 13 figures, submitted to The Astrophysical Journal

More is Better in Modern Machine Learning: when Infinite Overparameterization is Optimal and Overfitting is Obligatory

Authors: James B. Simon, Dhruva Karkada, Nikhil Ghosh, Mikhail Belkin

Abstract: In our era of enormous neural networks, empirical progress has been driven by the philosophy that more is better. Recent deep learning practice has found repeatedly that larger model size, more data, and more computation (resulting in lower training loss) improves performance. In this paper, we give theoretical backing to these empirical observations by showing that these three properties hold in… ▽ More In our era of enormous neural networks, empirical progress has been driven by the philosophy that more is better. Recent deep learning practice has found repeatedly that larger model size, more data, and more computation (resulting in lower training loss) improves performance. In this paper, we give theoretical backing to these empirical observations by showing that these three properties hold in random feature (RF) regression, a class of models equivalent to shallow networks with only the last layer trained. Concretely, we first show that the test risk of RF regression decreases monotonically with both the number of features and the number of samples, provided the ridge penalty is tuned optimally. In particular, this implies that infinite width RF architectures are preferable to those of any finite width. We then proceed to demonstrate that, for a large class of tasks characterized by powerlaw eigenstructure, training to near-zero training loss is obligatory: near-optimal performance can only be achieved when the training error is much smaller than the test error. Grounding our theory in real-world data, we find empirically that standard computer vision tasks with convolutional neural tangent kernels clearly fall into this class. Taken together, our results tell a simple, testable story of the benefits of overparameterization, overfitting, and more data in random feature models. △ Less

Submitted 15 May, 2024; v1 submitted 24 November, 2023; originally announced November 2023.

Comments: Appeared in ICLR 2024

A Spectral Condition for Feature Learning

Authors: Greg Yang, James B. Simon, Jeremy Bernstein

Abstract: The push to train ever larger neural networks has motivated the study of initialization and training at large network width. A key challenge is to scale training so that a network's internal representations evolve nontrivially at all widths, a process known as feature learning. Here, we show that feature learning is achieved by scaling the spectral norm of weight matrices and their updates like… ▽ More The push to train ever larger neural networks has motivated the study of initialization and training at large network width. A key challenge is to scale training so that a network's internal representations evolve nontrivially at all widths, a process known as feature learning. Here, we show that feature learning is achieved by scaling the spectral norm of weight matrices and their updates like $\sqrt{\texttt{fan-out}/\texttt{fan-in}}$, in contrast to widely used but heuristic scalings based on Frobenius norm and entry size. Our spectral scaling analysis also leads to an elementary derivation of \emph{maximal update parametrization}. All in all, we aim to provide the reader with a solid conceptual understanding of feature learning in neural networks. △ Less

Submitted 13 May, 2024; v1 submitted 26 October, 2023; originally announced October 2023.

Les Houches Lectures on Deep Learning at Large & Infinite Width

Authors: Yasaman Bahri, Boris Hanin, Antonin Brossollet, Vittorio Erba, Christian Keup, Rosalba Pacelli, James B. Simon

Abstract: These lectures, presented at the 2022 Les Houches Summer School on Statistical Physics and Machine Learning, focus on the infinite-width limit and large-width regime of deep neural networks. Topics covered include various statistical and dynamical properties of these networks. In particular, the lecturers discuss properties of random deep neural networks; connections between trained deep neural ne… ▽ More These lectures, presented at the 2022 Les Houches Summer School on Statistical Physics and Machine Learning, focus on the infinite-width limit and large-width regime of deep neural networks. Topics covered include various statistical and dynamical properties of these networks. In particular, the lecturers discuss properties of random deep neural networks; connections between trained deep neural networks, linear models, kernels, and Gaussian processes that arise in the infinite-width limit; and perturbative and non-perturbative treatments of large but finite-width networks, at initialization and after training. △ Less

Submitted 12 February, 2024; v1 submitted 4 September, 2023; originally announced September 2023.

Comments: These are notes from lectures delivered by Yasaman Bahri and Boris Hanin at the 2022 Les Houches Summer School on Statistics Physics and Machine Learning and a first version of them were transcribed by Antonin Brossollet, Vittorio Erba, Christian Keup, Rosalba Pacelli, James B. Simon

An Agnostic View on the Cost of Overfitting in (Kernel) Ridge Regression

Authors: Lijia Zhou, James B. Simon, Gal Vardi, Nathan Srebro

Abstract: We study the cost of overfitting in noisy kernel ridge regression (KRR), which we define as the ratio between the test error of the interpolating ridgeless model and the test error of the optimally-tuned model. We take an "agnostic" view in the following sense: we consider the cost as a function of sample size for any target function, even if the sample size is not large enough for consistency or… ▽ More We study the cost of overfitting in noisy kernel ridge regression (KRR), which we define as the ratio between the test error of the interpolating ridgeless model and the test error of the optimally-tuned model. We take an "agnostic" view in the following sense: we consider the cost as a function of sample size for any target function, even if the sample size is not large enough for consistency or the target is outside the RKHS. We analyze the cost of overfitting under a Gaussian universality ansatz using recently derived (non-rigorous) risk estimates in terms of the task eigenstructure. Our analysis provides a more refined characterization of benign, tempered and catastrophic overfitting (cf. Mallinar et al. 2022). △ Less

Submitted 22 March, 2024; v1 submitted 22 June, 2023; originally announced June 2023.

Comments: This is the ICLR CR version

Tune As You Scale: Hyperparameter Optimization For Compute Efficient Training

Authors: Abraham J. Fetterman, Ellie Kitanidis, Joshua Albrecht, Zachary Polizzi, Bryden Fogelman, Maksis Knutins, Bartosz Wróblewski, James B. Simon, Kanjun Qiu

Abstract: Hyperparameter tuning of deep learning models can lead to order-of-magnitude performance gains for the same amount of compute. Despite this, systematic tuning is uncommon, particularly for large models, which are expensive to evaluate and tend to have many hyperparameters, necessitating difficult judgment calls about tradeoffs, budgets, and search bounds. To address these issues and propose a prac… ▽ More Hyperparameter tuning of deep learning models can lead to order-of-magnitude performance gains for the same amount of compute. Despite this, systematic tuning is uncommon, particularly for large models, which are expensive to evaluate and tend to have many hyperparameters, necessitating difficult judgment calls about tradeoffs, budgets, and search bounds. To address these issues and propose a practical method for robustly tuning large models, we present Cost-Aware Pareto Region Bayesian Search (CARBS), a Bayesian optimization algorithm that performs local search around the performance-cost Pareto frontier. CARBS does well even in unbounded search spaces with many hyperparameters, learns scaling relationships so that it can tune models even as they are scaled up, and automates much of the "black magic" of tuning. Among our results, we effectively solve the entire ProcGen benchmark just by tuning a simple baseline (PPO, as provided in the original ProcGen paper). We also reproduce the model size vs. training tokens scaling result from the Chinchilla project (Hoffmann et al. 2022), while simultaneously discovering scaling laws for every other hyperparameter, via an easy automated process that uses significantly less compute and is applicable to any deep learning problem (not just language models). △ Less

Submitted 13 June, 2023; originally announced June 2023.

On the Stepwise Nature of Self-Supervised Learning

Authors: James B. Simon, Maksis Knutins, Liu Ziyin, Daniel Geisz, Abraham J. Fetterman, Joshua Albrecht

Abstract: We present a simple picture of the training process of joint embedding self-supervised learning methods. We find that these methods learn their high-dimensional embeddings one dimension at a time in a sequence of discrete, well-separated steps. We arrive at this conclusion via the study of a linearized model of Barlow Twins applicable to the case in which the trained network is infinitely wide. We… ▽ More We present a simple picture of the training process of joint embedding self-supervised learning methods. We find that these methods learn their high-dimensional embeddings one dimension at a time in a sequence of discrete, well-separated steps. We arrive at this conclusion via the study of a linearized model of Barlow Twins applicable to the case in which the trained network is infinitely wide. We solve the training dynamics of this model from small initialization, finding that the model learns the top eigenmodes of a certain contrastive kernel in a stepwise fashion, and obtain a closed-form expression for the final learned representations. Remarkably, we then see the same stepwise learning phenomenon when training deep ResNets using the Barlow Twins, SimCLR, and VICReg losses. Our theory suggests that, just as kernel regression can be thought of as a model of supervised learning, kernel PCA may serve as a useful model of self-supervised learning. △ Less

Submitted 30 May, 2023; v1 submitted 27 March, 2023; originally announced March 2023.

Comments: 9 pages (main text) + 14 pages (refs + appendices). ICML '23

Comets and Planetesimal Formation

Authors: Jacob B. Simon, Jürgen Blum, Til Birnstiel, David Nesvorný

Abstract: In this chapter, we review the processes involved in the formation of planetesimals and comets. We will start with a description of the physics of dust grain growth and how this is mediated by gas-dust interactions in planet-forming disks. We will then delve into the various models of planetesimal formation, describing how these planetesimals form as well as their resulting structure. In doing so,… ▽ More In this chapter, we review the processes involved in the formation of planetesimals and comets. We will start with a description of the physics of dust grain growth and how this is mediated by gas-dust interactions in planet-forming disks. We will then delve into the various models of planetesimal formation, describing how these planetesimals form as well as their resulting structure. In doing so, we focus on and compare two paradigms for planetesimal formation: the gravitational collapse of particle over-densities (which can be produced by a variety of mechanisms) and the growth of particles into planetesimals via collisional and gravitational coagulation. Finally, we compare the predictions from these models with data collected by the Rosetta and New Horizons missions and that obtained via observations of distant Kuiper Belt Objects. △ Less

Submitted 8 December, 2022; originally announced December 2022.

Comments: Planetesimal Formation Review accepted for publication in Comets III

Avalon: A Benchmark for RL Generalization Using Procedurally Generated Worlds

Authors: Joshua Albrecht, Abraham J. Fetterman, Bryden Fogelman, Ellie Kitanidis, Bartosz Wróblewski, Nicole Seo, Michael Rosenthal, Maksis Knutins, Zachary Polizzi, James B. Simon, Kanjun Qiu

Abstract: Despite impressive successes, deep reinforcement learning (RL) systems still fall short of human performance on generalization to new tasks and environments that differ from their training. As a benchmark tailored for studying RL generalization, we introduce Avalon, a set of tasks in which embodied agents in highly diverse procedural 3D worlds must survive by navigating terrain, hunting or gatheri… ▽ More Despite impressive successes, deep reinforcement learning (RL) systems still fall short of human performance on generalization to new tasks and environments that differ from their training. As a benchmark tailored for studying RL generalization, we introduce Avalon, a set of tasks in which embodied agents in highly diverse procedural 3D worlds must survive by navigating terrain, hunting or gathering food, and avoiding hazards. Avalon is unique among existing RL benchmarks in that the reward function, world dynamics, and action space are the same for every task, with tasks differentiated solely by altering the environment; its 20 tasks, ranging in complexity from eat and throw to hunt and navigate, each create worlds in which the agent must perform specific skills in order to survive. This setup enables investigations of generalization within tasks, between tasks, and to compositional tasks that require combining skills learned from previous tasks. Avalon includes a highly efficient simulator, a library of baselines, and a benchmark with scoring metrics evaluated against hundreds of hours of human performance, all of which are open-source and publicly available. We find that standard RL baselines make progress on most tasks but are still far from human performance, suggesting Avalon is challenging enough to advance the quest for generalizable RL. △ Less

Submitted 24 October, 2022; originally announced October 2022.

Comments: Accepted to NeurIPS Datasets and Benchmarks 2022. Video and links to all code, data, etc can be found at https://generallyintelligent.com/avalon/

On Kernel Regression with Data-Dependent Kernels

Authors: James B. Simon

Abstract: The primary hyperparameter in kernel regression (KR) is the choice of kernel. In most theoretical studies of KR, one assumes the kernel is fixed before seeing the training data. Under this assumption, it is known that the optimal kernel is equal to the prior covariance of the target function. In this note, we consider KR in which the kernel may be updated after seeing the training data. We point o… ▽ More The primary hyperparameter in kernel regression (KR) is the choice of kernel. In most theoretical studies of KR, one assumes the kernel is fixed before seeing the training data. Under this assumption, it is known that the optimal kernel is equal to the prior covariance of the target function. In this note, we consider KR in which the kernel may be updated after seeing the training data. We point out that an analogous choice of kernel using the posterior of the target function is optimal in this setting. Connections to the view of deep neural networks as data-dependent kernel learners are discussed. △ Less

Submitted 26 September, 2022; v1 submitted 4 September, 2022; originally announced September 2022.

Comments: 7 pages, 1 figure

Benign, Tempered, or Catastrophic: A Taxonomy of Overfitting

Authors: Neil Mallinar, James B. Simon, Amirhesam Abedsoltan, Parthe Pandit, Mikhail Belkin, Preetum Nakkiran

Abstract: The practical success of overparameterized neural networks has motivated the recent scientific study of interpolating methods, which perfectly fit their training data. Certain interpolating methods, including neural networks, can fit noisy training data without catastrophically bad test performance, in defiance of standard intuitions from statistical learning theory. Aiming to explain this, a body… ▽ More The practical success of overparameterized neural networks has motivated the recent scientific study of interpolating methods, which perfectly fit their training data. Certain interpolating methods, including neural networks, can fit noisy training data without catastrophically bad test performance, in defiance of standard intuitions from statistical learning theory. Aiming to explain this, a body of recent work has studied benign overfitting, a phenomenon where some interpolating methods approach Bayes optimality, even in the presence of noise. In this work we argue that while benign overfitting has been instructive and fruitful to study, many real interpolating methods like neural networks do not fit benignly: modest noise in the training set causes nonzero (but non-infinite) excess risk at test time, implying these models are neither benign nor catastrophic but rather fall in an intermediate regime. We call this intermediate regime tempered overfitting, and we initiate its systematic study. We first explore this phenomenon in the context of kernel (ridge) regression (KR) by obtaining conditions on the ridge parameter and kernel eigenspectrum under which KR exhibits each of the three behaviors. We find that kernels with powerlaw spectra, including Laplace kernels and ReLU neural tangent kernels, exhibit tempered overfitting. We then empirically study deep neural networks through the lens of our taxonomy, and find that those trained to interpolation are tempered, while those stopped early are benign. We hope our work leads to a more refined understanding of overfitting in modern learning. △ Less

Submitted 15 July, 2024; v1 submitted 13 July, 2022; originally announced July 2022.

Comments: NM and JS co-first authors

Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models

Authors: Aarohi Srivastava, Abhinav Rastogi, Abhishek Rao, Abu Awal Md Shoeb, Abubakar Abid, Adam Fisch, Adam R. Brown, Adam Santoro, Aditya Gupta, Adrià Garriga-Alonso, Agnieszka Kluska, Aitor Lewkowycz, Akshat Agarwal, Alethea Power, Alex Ray, Alex Warstadt, Alexander W. Kocurek, Ali Safaya, Ali Tazarv, Alice Xiang, Alicia Parrish, Allen Nie, Aman Hussain, Amanda Askell, Amanda Dsouza , et al. (426 additional authors not shown)

Abstract: Language models demonstrate both quantitative improvement and new qualitative capabilities with increasing scale. Despite their potentially transformative impact, these new capabilities are as yet poorly characterized. In order to inform future research, prepare for disruptive new model capabilities, and ameliorate socially harmful effects, it is vital that we understand the present and near-futur… ▽ More Language models demonstrate both quantitative improvement and new qualitative capabilities with increasing scale. Despite their potentially transformative impact, these new capabilities are as yet poorly characterized. In order to inform future research, prepare for disruptive new model capabilities, and ameliorate socially harmful effects, it is vital that we understand the present and near-future capabilities and limitations of language models. To address this challenge, we introduce the Beyond the Imitation Game benchmark (BIG-bench). BIG-bench currently consists of 204 tasks, contributed by 450 authors across 132 institutions. Task topics are diverse, drawing problems from linguistics, childhood development, math, common-sense reasoning, biology, physics, social bias, software development, and beyond. BIG-bench focuses on tasks that are believed to be beyond the capabilities of current language models. We evaluate the behavior of OpenAI's GPT models, Google-internal dense transformer architectures, and Switch-style sparse transformers on BIG-bench, across model sizes spanning millions to hundreds of billions of parameters. In addition, a team of human expert raters performed all tasks in order to provide a strong baseline. Findings include: model performance and calibration both improve with scale, but are poor in absolute terms (and when compared with rater performance); performance is remarkably similar across model classes, though with benefits from sparsity; tasks that improve gradually and predictably commonly involve a large knowledge or memorization component, whereas tasks that exhibit "breakthrough" behavior at a critical scale often involve multiple steps or components, or brittle metrics; social bias typically increases with scale in settings with ambiguous context, but this can be improved with prompting. △ Less

Submitted 12 June, 2023; v1 submitted 9 June, 2022; originally announced June 2022.

Comments: 27 pages, 17 figures + references and appendices, repo: https://github.com/google/BIG-bench

Journal ref: Transactions on Machine Learning Research, May/2022, https://openreview.net/forum?id=uyTL5Bvosj

The Streaming Instability Cannot Form Planetesimals from mm-size Grains in Pressure Bumps

Authors: Daniel Carrera, Jacob B. Simon

Abstract: We present evidence that it is unlikely that the streaming instability (SI) can form planetesimals from mm grains inside axisymmetric pressure bumps. We conducted the largest simulation of the SI so far (7 million CPU hours), consisting of a large slice of the disk with mm grains, a solar-like dust-to-gas ratio ($Z = 0.01$), and the largest pressure bump that does not cause gravitational instabili… ▽ More We present evidence that it is unlikely that the streaming instability (SI) can form planetesimals from mm grains inside axisymmetric pressure bumps. We conducted the largest simulation of the SI so far (7 million CPU hours), consisting of a large slice of the disk with mm grains, a solar-like dust-to-gas ratio ($Z = 0.01$), and the largest pressure bump that does not cause gravitational instability (GI) in the particle layer. We used a high resolution of $1000/H$ to resolve as many SI unstable modes as possible. The simulation produced a long-lived particle over-density far exceeding the SI criteria (i.e., a critical solid abundance to headwind parameter ratio $Z/Π$) where strong clumping would occur if these conditions were present over an extended region of the disk; yet we observed none. The likely reason is that the time it takes particles to cross the high-$Z/Π$ region ($t_{\rm cross}$) is shorter than the growth timescale of the SI ($t_{\rm grow}$). We propose an added criterion for planetesimal formation by the SI -- that $t_{\rm cross} > t_{\rm grow}$. We show that any bump larger than the one in this run would form planetesimals by the GI instead of the SI. Our results significantly restrict the pathways to planet formation: Either protoplanetary disks regularly form grains larger than 1~mm, or planetesimals do not form by the SI in axisymmetric pressure bumps. Since bumps large enough to induce the GI are likely Rossby-wave unstable, we propose that mm grains may only form planetesimals in vortices. △ Less

Submitted 23 May, 2022; v1 submitted 29 April, 2022; originally announced April 2022.

Comments: 13 pages, 8 figures, accepted to ApJL. Updated with minor changes to improve clarity with respect to prior works

Hydro-, Magnetohydro-, and Dust-Gas Dynamics of Protoplanetary Disks

Authors: G. Lesur, B. Ercolano, M. Flock, M. -K. Lin, C. -C. Yang, J. A. Barranco, P. Benitez-Llambay, J. Goodman, A. Johansen, H. Klahr, G. Laibe, W. Lyra, P. Marcus, R. P. Nelson, J. Squire, J. B. Simon, N. Turner, O. M. Umurhan, A. N. Youdin

Abstract: The building of planetary systems is controlled by the gas and dust dynamics of protoplanetary disks. While the gas is simultaneously accreted onto the central star and dissipated away by winds, dust grains aggregate and collapse to form planetesimals and eventually planets. This dust and gas dynamics involves instabilities, turbulence and complex non-linear interactions which ultimately control t… ▽ More The building of planetary systems is controlled by the gas and dust dynamics of protoplanetary disks. While the gas is simultaneously accreted onto the central star and dissipated away by winds, dust grains aggregate and collapse to form planetesimals and eventually planets. This dust and gas dynamics involves instabilities, turbulence and complex non-linear interactions which ultimately control the observational appearance and the secular evolution of these disks. This chapter is dedicated to the most recent developments in our understanding of the dynamics of gaseous and dusty disks, covering hydrodynamic and magnetohydrodynamic turbulence, gas-dust instabilities, dust clumping and disk winds. We show how these physical processes have been tested from observations and highlight standing questions that should be addressed in the future. △ Less

Submitted 18 March, 2022; originally announced March 2022.

Comments: 32 pages, 7 figures, to appear in Protostars and Planets VII, eds: Shu-ichiro Inutsuka, Yuri Aikawa, Takayuki Muto, Kengo Tomida, and Motohide Tamura

Using Bayesian Deep Learning to infer Planet Mass from Gaps in Protoplanetary Disks

Authors: Sayantan Auddy, Ramit Dey, Min-Kai Lin, Daniel Carrera, Jacob B. Simon

Abstract: Planet induced sub-structures, like annular gaps, observed in dust emission from protoplanetary disks provide a unique probe to characterize unseen young planets. While deep learning based model has an edge in characterizing the planet's properties over traditional methods, like customized simulations and empirical relations, it lacks in its ability to quantify the uncertainty associated with its… ▽ More Planet induced sub-structures, like annular gaps, observed in dust emission from protoplanetary disks provide a unique probe to characterize unseen young planets. While deep learning based model has an edge in characterizing the planet's properties over traditional methods, like customized simulations and empirical relations, it lacks in its ability to quantify the uncertainty associated with its predictions. In this paper, we introduce a Bayesian deep learning network "DPNNet-Bayesian" that can predict planet mass from disk gaps and provides uncertainties associated with the prediction. A unique feature of our approach is that it can distinguish between the uncertainty associated with the deep learning architecture and uncertainty inherent in the input data due to measurement noise. The model is trained on a data set generated from disk-planet simulations using the \textsc{fargo3d} hydrodynamics code with a newly implemented fixed grain size module and improved initial conditions. The Bayesian framework enables estimating a gauge/confidence interval over the validity of the prediction when applied to unknown observations. As a proof-of-concept, we apply DPNNet-Bayesian to dust gaps observed in HL Tau. The network predicts masses of $ 86.0 \pm 5.5 M_{\Earth} $, $ 43.8 \pm 3.3 M_{\Earth} $, and $ 92.2 \pm 5.1 M_{\Earth} $ respectively, which are comparable to other studies based on specialized simulations. △ Less

Submitted 23 February, 2022; originally announced February 2022.

Comments: 14 pages, 6 figures, submitted to ApJ

The Eigenlearning Framework: A Conservation Law Perspective on Kernel Regression and Wide Neural Networks

Authors: James B. Simon, Madeline Dickens, Dhruva Karkada, Michael R. DeWeese

Abstract: We derive simple closed-form estimates for the test risk and other generalization metrics of kernel ridge regression (KRR). Relative to prior work, our derivations are greatly simplified and our final expressions are more readily interpreted. These improvements are enabled by our identification of a sharp conservation law which limits the ability of KRR to learn any orthonormal basis of functions.… ▽ More We derive simple closed-form estimates for the test risk and other generalization metrics of kernel ridge regression (KRR). Relative to prior work, our derivations are greatly simplified and our final expressions are more readily interpreted. These improvements are enabled by our identification of a sharp conservation law which limits the ability of KRR to learn any orthonormal basis of functions. Test risk and other objects of interest are expressed transparently in terms of our conserved quantity evaluated in the kernel eigenbasis. We use our improved framework to: i) provide a theoretical explanation for the "deep bootstrap" of Nakkiran et al (2020), ii) generalize a previous result regarding the hardness of the classic parity problem, iii) fashion a theoretical tool for the study of adversarial robustness, and iv) draw a tight analogy between KRR and a well-studied system in statistical physics. △ Less

Submitted 26 October, 2023; v1 submitted 8 October, 2021; originally announced October 2021.

Comments: 12 pages (main text) + 25 pages (refs + appendices). A previous version of this manuscript was entitled "Neural Tangent Kernel Eigenvalues Accurately Predict Generalization."

Resilience of Planetesimal Formation in Weakly-Reinforced Pressure Bumps

Authors: Daniel Carrera, Andrew Thomas, Jacob B. Simon, Matthew A. Small, Katherine A. Kretke, Hubert Klahr

Abstract: The discovery that axisymmetric dust rings are ubiquitous in protoplanetary disks has provoked a flurry of research on the role of pressure bumps in planet formation. High-resolution simulations by our group have shown that even a modest bump can collect enough dust to trigger planetesimal formation by the streaming instability. In this work, we probe the limits of planetesimal formation when the… ▽ More The discovery that axisymmetric dust rings are ubiquitous in protoplanetary disks has provoked a flurry of research on the role of pressure bumps in planet formation. High-resolution simulations by our group have shown that even a modest bump can collect enough dust to trigger planetesimal formation by the streaming instability. In this work, we probe the limits of planetesimal formation when the external source of pressure bump reinforcement is extremely weak. We conduct simulations of radially elongated shearing boxes to capture the entire bump, which is generated and maintained over some timescale $t_{\rm reinf}$ by a Newtonian relaxation scheme. We find that planetesimal formation is extremely resilient for cm-sized grains. We reduced the strength of reinforcement by up to a factor of 100 and the location and initial masses of planetesimals were essentially unaffected. However, we do find that strong reinforcement causes much faster pebble drift compared to the the standard pebble drift rates. The resulting larger pebble flux enhances the planetesimal growth rate by pebble accretion. We hypothesize that to sustain the bump, our code has to extract angular momentum (the strength of this negative torque depends on $t_{\rm reinf}$), and some of this torque is transferred to the particles, causing them to drift faster for a stronger torque (i.e., smaller $t_{\rm reinf}$). Since any physical process that sustains a pressure bump must do so by torquing the gas, we conjecture that the effect on pebble drift is a real phenomenon, motivating further work with physically realistic sources to generate the bump. △ Less

Submitted 1 February, 2022; v1 submitted 18 August, 2021; originally announced August 2021.

Comments: 12 pages, 8 figures, accepted to ApJ

SGD with a Constant Large Learning Rate Can Converge to Local Maxima

Authors: Liu Ziyin, Botao Li, James B. Simon, Masahito Ueda

Abstract: Previous works on stochastic gradient descent (SGD) often focus on its success. In this work, we construct worst-case optimization problems illustrating that, when not in the regimes that the previous works often assume, SGD can exhibit many strange and potentially undesirable behaviors. Specifically, we construct landscapes and data distributions such that (1) SGD converges to local maxima, (2) S… ▽ More Previous works on stochastic gradient descent (SGD) often focus on its success. In this work, we construct worst-case optimization problems illustrating that, when not in the regimes that the previous works often assume, SGD can exhibit many strange and potentially undesirable behaviors. Specifically, we construct landscapes and data distributions such that (1) SGD converges to local maxima, (2) SGD escapes saddle points arbitrarily slowly, (3) SGD prefers sharp minima over flat ones, and (4) AMSGrad converges to local maxima. We also realize results in a minimal neural network-like example. Our results highlight the importance of simultaneously analyzing the minibatch sampling, discrete-time updates rules, and realistic landscapes to understand the role of SGD in deep learning. △ Less

Submitted 27 May, 2023; v1 submitted 25 July, 2021; originally announced July 2021.

Comments: Fixed typos

Interleaved Electro-Optic Dual Comb Generation to Expand Bandwidth and Scan Rate for Molecular Spectroscopy and Dynamics Studies near 1.6 μm

Authors: Jasper R. Stroud, James B. Simon, Gerd A. Wagner, David F. Plusquellic

Abstract: A chirped-pulse interleaving method is reported for generation of dual optical frequency combs based on electro-optic phase modulators (EOM) in a free-running all-fiber based system. Methods are discussed to easily modify the linear chirp rate and comb resolution by more than three orders of magnitude and to significantly increase the spectral bandwidth coverage. The agility of the technique is sh… ▽ More A chirped-pulse interleaving method is reported for generation of dual optical frequency combs based on electro-optic phase modulators (EOM) in a free-running all-fiber based system. Methods are discussed to easily modify the linear chirp rate and comb resolution by more than three orders of magnitude and to significantly increase the spectral bandwidth coverage. The agility of the technique is shown to both capture complex line shapes and to magnify rapid passage effects in spectroscopic and molecular dynamics studies of CO2. These methods are well-suited for applications in the areas of remote sensing, reaction dynamics, and sub-Doppler studies across the wide spectral regions accessible to EOMs. △ Less

Submitted 21 June, 2021; originally announced June 2021.

Reverse Engineering the Neural Tangent Kernel

Authors: James B. Simon, Sajant Anand, Michael R. DeWeese

Abstract: The development of methods to guide the design of neural networks is an important open challenge for deep learning theory. As a paradigm for principled neural architecture design, we propose the translation of high-performing kernels, which are better-understood and amenable to first-principles design, into equivalent network architectures, which have superior efficiency, flexibility, and feature… ▽ More The development of methods to guide the design of neural networks is an important open challenge for deep learning theory. As a paradigm for principled neural architecture design, we propose the translation of high-performing kernels, which are better-understood and amenable to first-principles design, into equivalent network architectures, which have superior efficiency, flexibility, and feature learning. To this end, we constructively prove that, with just an appropriate choice of activation function, any positive-semidefinite dot-product kernel can be realized as either the NNGP or neural tangent kernel of a fully-connected neural network with only one hidden layer. We verify our construction numerically and demonstrate its utility as a design tool for finite fully-connected networks in several experiments. △ Less

Submitted 13 August, 2022; v1 submitted 6 June, 2021; originally announced June 2021.

Comments: 15 pages, 5 figures

Binary Planetesimal Formation from Gravitationally Collapsing Pebble Clouds

Authors: David Nesvorny, Rixin Li, Jacob B. Simon, Andrew N. Youdin, Derek C. Richardson, Raphael Marschall, William M. Grundy

Abstract: Planetesimals are compact astrophysical objects roughly 1-1000 km in size, massive enough to be held together by gravity. They can grow by accreting material to become full-size planets. Planetesimals themselves are thought to form by complex physical processes from small grains in protoplanetary disks. The streaming instability (SI) model states that mm/cm-size particles (pebbles) are aerodynamic… ▽ More Planetesimals are compact astrophysical objects roughly 1-1000 km in size, massive enough to be held together by gravity. They can grow by accreting material to become full-size planets. Planetesimals themselves are thought to form by complex physical processes from small grains in protoplanetary disks. The streaming instability (SI) model states that mm/cm-size particles (pebbles) are aerodynamically collected into self-gravitating clouds which then directly collapse into planetesimals. Here we analyze ATHENA simulations of the SI to characterize the initial properties (e.g., rotation) of pebble clouds. Their gravitational collapse is followed with the PKDGRAV N-body code, which has been modified to realistically account for pebble collisions. We find that pebble clouds rapidly collapse into short-lived disk structures from which planetesimals form. The planetesimal properties depend on the cloud's scaled angular momentum, l=L/(M R_H^2 Omega, where L and M are the angular momentum and mass, R_H is the Hill radius, and Omega is the orbital frequency. Low-l pebble clouds produce tight (or contact) binaries and single planetesimals. Compact high-l clouds give birth to binary planetesimals with attributes that closely resemble the equal-size binaries found in the Kuiper belt. Significantly, the SI-triggered gravitational collapse can explain the angular momentum distribution of known equal-size binaries -- a result pending verification from studies with improved resolution. About 10% of collapse simulations produce hierarchical systems with two or more large moons. These systems should be found in the Kuiper belt when observations reach the threshold sensitivity. △ Less

Submitted 13 November, 2020; originally announced November 2020.

Comments: Planetary Space Journal

Visualizing the Kinematics of Planet Formation

Authors: Disk Dynamics Collaboration, Philip J. Armitage, Jaehan Bae, Myriam Benisty, Edwin A. Bergin, Simon Casassus, Ian Czekala, Stefano Facchini, Jeffrey Fung, Cassandra Hall, John D. Ilee, Miriam Keppler, Aleksandra Kuznetsova, Romane Le Gal, Ryan A. Loomis, Wladimir Lyra, Natascha Manger, Sebastian Perez, Christophe Pinte, Daniel J. Price, Giovanni Rosotti, Judit Szulagyi, Kamber Schwarz, Jacob B. Simon, Richard Teague , et al. (1 additional authors not shown)

Abstract: A stunning range of substructures in the dust of protoplanetary disks is routinely observed across a range of wavelengths. These gaps, rings and spirals are highly indicative of a population of unseen planets, hinting at the possibility of current observational facilities being able to capture planet-formation in action. Over the last decade, our understanding of the influence of a young planet on… ▽ More A stunning range of substructures in the dust of protoplanetary disks is routinely observed across a range of wavelengths. These gaps, rings and spirals are highly indicative of a population of unseen planets, hinting at the possibility of current observational facilities being able to capture planet-formation in action. Over the last decade, our understanding of the influence of a young planet on the dynamical structure of its parental disk has progressed significantly, revealing a host of potentially observable features which would betray the presence of a deeply embedded planet. In concert, recent observations have shown that subtle perturbations in the kinematic structure of protoplanetary disks are found in multiple sources, potentially the characteristic disturbances associated with embedded planets. In this work, we review the theoretical background of planet-disk interactions, focusing on the kinematical features, and the current methodologies used to observe these interactions in spatially and spectrally resolved observations. We discuss the potential pit falls of such kinematical detections of planets, providing best-practices for imaging and analysing interferometric data, along with a set of criteria to use as a benchmark for any claimed detection of embedded planets. We finish with a discussion on the current state of simulations in regard to planet-disk interactions, highlighting areas of particular interest and future directions which will provide the most significant impact in our search for embedded planets. This work is the culmination of the 'Visualizing the Kinematics of Planet Formation' workshop, held in October 2019 at the Center for Computational Astrophysics at the Flatiron Institute in New York City. △ Less

Submitted 15 September, 2020; v1 submitted 9 September, 2020; originally announced September 2020.

Comments: To be submitted to PASA. Comments welcome

Protoplanetary Disk Rings as Sites for Planetesimal Formation

Authors: Daniel Carrera, Jacob B. Simon, Rixin Li, Katherine A. Kretke, Hubert Klahr

Abstract: Axisymmetric dust rings are a ubiquitous feature of young protoplanetary disks. These rings are likely caused by pressure bumps in the gas profile; a small bump can induce a traffic jam-like pattern in the dust density, while a large bump may halt radial dust drift entirely. The resulting increase in dust concentration may trigger planetesimal formation by the streaming instability (SI), as the SI… ▽ More Axisymmetric dust rings are a ubiquitous feature of young protoplanetary disks. These rings are likely caused by pressure bumps in the gas profile; a small bump can induce a traffic jam-like pattern in the dust density, while a large bump may halt radial dust drift entirely. The resulting increase in dust concentration may trigger planetesimal formation by the streaming instability (SI), as the SI itself requires some initial concentration. Here we present the first 3D simulations of planetesimal formation in the presence of a pressure bump modeled specifically after those observed by ALMA. In particular, we place a pressure bump at the center of a large 3D shearing box, along with an initial solid-to-gas ratio of $Z = 0.01$, and we include both particle back-reaction and particle self-gravity. We consider both mm-sized and cm-sized particles separately. For simulations with cm-sized particles, we find that even a small pressure bump leads to the formation of planetesimals via the streaming instability; a pressure bump does {\it not} need to fully halt radial particle drift for the SI to become efficient. Furthermore, pure gravitational collapse via concentration in pressure bumps (such as would occur at sufficiently high concentrations and without the streaming instability) is not responsible for planetesimal formation. For mm-sized particles, we find tentative evidence that planetesimal formation does not occur. This result, if it holds up at higher resolution and for a broader range of parameters, could put strong constraints on where in protoplanetary disks planetesimals can form. Ultimately, however, our results suggest that for cm-sized particles, planetesimal formation in pressure bumps is an extremely robust process. △ Less

Submitted 12 January, 2021; v1 submitted 4 August, 2020; originally announced August 2020.

Comments: accepted to ApJ; 22 pages, 16 figures

Measuring turbulent motion in planet-forming disks with ALMA: A detection around DM Tau and non-detections around MWC 480 and V4046 Sgr

Authors: Kevin Flaherty, A. Meredith Hughes, Jacob B. Simon, Chunhua Qi, Xue-ning Bai, Alyssa Bulatek, Sean M. Andrews, David J. Wilner, Agnes Kospal

Abstract: Turbulence is a crucial factor in many models of planet formation, but it has only been directly constrained among a small number of planet forming disks. Building on the upper limits on turbulence placed in disks around HD 163296 and TW Hya, we present ALMA CO J=2-1 line observations at $\sim0.3"$ (20-50 au) resolution and 80 ms$^{-1}$ channel spacing of the disks around DM Tau, MWC 480, and V404… ▽ More Turbulence is a crucial factor in many models of planet formation, but it has only been directly constrained among a small number of planet forming disks. Building on the upper limits on turbulence placed in disks around HD 163296 and TW Hya, we present ALMA CO J=2-1 line observations at $\sim0.3"$ (20-50 au) resolution and 80 ms$^{-1}$ channel spacing of the disks around DM Tau, MWC 480, and V4046 Sgr. Using parametric models of disk structure, we robustly detect non-thermal gas motions around DM Tau of between 0.25 c$_s$ and 0.33 c$_s$, with the range dominated by systematic effects, making this one of the only systems with directly measured non-zero turbulence. Using the same methodology, we place stringent upper limits on the non-thermal gas motion around MWC 480 ($<$0.08c$_s$) and V4046 Sgr ($<$0.12 c$_s$). The preponderance of upper limits in this small sample, and the modest turbulence levels consistent with dust studies, suggest that weak turbulence ($α\lesssim10^{-3}$) may be a common, albeit not universal, feature of planet-forming disks. We explore the particular physical conditions around DM Tau that could lead this system to be more turbulent than the others. △ Less

Submitted 28 April, 2020; v1 submitted 25 April, 2020; originally announced April 2020.

Comments: 21 pages, 10 figures, accepted to ApJ

Turbulence Regulates the Rate of Planetesimal Formation via Gravitational Collapse

Authors: Daniel A. Gole, Jacob B. Simon, Rixin Li, Andrew N. Youdin, Philip J. Armitage

Abstract: We study how the interaction between the streaming instability and intrinsic gas-phase turbulence affects planetesimal formation via gravitational collapse in protoplanetary disks. Turbulence impedes the formation of particle clumps by acting as an effective turbulent diffusivity, but it can also promote planetesimal formation by concentrating solids, for example in zonal flows. We quantify the ef… ▽ More We study how the interaction between the streaming instability and intrinsic gas-phase turbulence affects planetesimal formation via gravitational collapse in protoplanetary disks. Turbulence impedes the formation of particle clumps by acting as an effective turbulent diffusivity, but it can also promote planetesimal formation by concentrating solids, for example in zonal flows. We quantify the effect of turbulent diffusivity using numerical simulations of the streaming instability in small local domains, forced with velocity perturbations that establish approximately Kolmogorov-like turbulence. We find that planetesimal formation is suppressed by turbulence once velocity fluctuations exceed $\langle δv^2 \rangle \simeq 10^{-3.5} - 10^{-3} c_s^2$. Turbulence whose strength is just below the threshold reduces the rate at which solids are bound into clumps. Our results suggest that the well-established turbulent thickening of the mid-plane solid layer is the primary mechanism by which turbulence influences planetesimal formation and that planetesimal formation requires a mid-plane solid-to-gas ratio $ε\gtrsim 0.5$. We also quantify the initial planetesimal mass function using a new clump-tracking method to determine each planetesimal mass shortly after collapse. For models in which planetesimals form, we show that the mass function is well-described by a broken power law, whose parameters are robust to the inclusion and strength of imposed turbulence. Turbulence in protoplanetary disks is likely to substantially exceed the threshold for planetesimal formation at radii where temperatures $T \gtrsim 10^3 \ {\rm K}$ lead to thermal ionization. Planetesimal formation may therefore be unviable in the inner disk out to 2-3 times the dust sublimation radius. △ Less

Submitted 21 October, 2020; v1 submitted 27 January, 2020; originally announced January 2020.

Comments: Accepted to ApJ after revisions

Strongly magnetized accretion disks: structure and accretion from global magnetohydrodynamic simulations

Authors: Bhupendra Mishra, Mitchell C. Begelman, Philip J. Armitage, Jacob B. Simon

Abstract: We use global magnetohydrodynamic simulations to study the influence of net vertical magnetic fields on the structure of geometrically thin ($H/r \approx 0.05$) accretion disks in the Newtonian limit. We consider initial mid-plane gas to magnetic pressure ratios $β_0 = 1000,\, 300$ and $100$, spanning the transition between weakly and strongly magnetized accretion regimes. We find that magnetic pr… ▽ More We use global magnetohydrodynamic simulations to study the influence of net vertical magnetic fields on the structure of geometrically thin ($H/r \approx 0.05$) accretion disks in the Newtonian limit. We consider initial mid-plane gas to magnetic pressure ratios $β_0 = 1000,\, 300$ and $100$, spanning the transition between weakly and strongly magnetized accretion regimes. We find that magnetic pressure is important for the disks' vertical structure in all three cases, with accretion occurring at $z/R\approx 0.2$ in the two most strongly magnetized models. The disk midplane shows outflow rather than accretion. Accretion through the surface layers is driven mainly by stress due to coherent large scale magnetic field rather than by turbulent stress. Equivalent viscosity parameters measured from our simulations show similar dependencies on initial $β_0$ to those seen in shearing box simulations, though the disk midplane is not magnetic pressure dominated even for the strongest magnetic field case. Winds are present but are not the dominant driver of disk evolution. Over the (limited) duration of our simulations, we find evidence that the net flux attains a quasi-steady state at levels that can stably maintain a strongly magnetized disk. We suggest that geometrically thin accretion disks in observed systems may commonly exist in a magnetically "elevated" state, characterized by non-zero but modest vertical magnetic fluxes, with potentially important implications for disk phenomenology in X-ray binaries (XRBs) and active galactic nuclei (AGN). △ Less

Submitted 21 January, 2020; v1 submitted 21 July, 2019; originally announced July 2019.

Trans-Neptunian Binaries as Evidence for Planetesimal Formation by the Streaming Instability

Authors: David Nesvorny, Rixin Li, Andrew N. Youdin, Jacob B. Simon, William M. Grundy

Abstract: A critical step toward the emergence of planets in a protoplanetary disk consists in accretion of planetesimals, bodies 1-1000 km in size, from smaller disk constituents. This process is poorly understood partly because we lack good observational constraints on the complex physical processes that contribute to planetesimal formation. In the outer solar system, the best place to look for clues is t… ▽ More A critical step toward the emergence of planets in a protoplanetary disk consists in accretion of planetesimals, bodies 1-1000 km in size, from smaller disk constituents. This process is poorly understood partly because we lack good observational constraints on the complex physical processes that contribute to planetesimal formation. In the outer solar system, the best place to look for clues is the Kuiper belt, where icy planetesimals survived to this day. Here we report evidence that Kuiper belt planetesimals formed by the streaming instability, a process in which aerodynamically concentrated clumps of pebbles gravitationally collapse into 100-km-class bodies. Gravitational collapse was previously suggested to explain the ubiquity of equal-size binaries in the Kuiper belt. We analyze new hydrodynamical simulations of the streaming instability to determine the model expectations for the spatial orientation of binary orbits. The predicted broad inclination distribution with 80% of prograde binary orbits matches the observations of trans-Neptunian binaries. The formation models which imply predominantly retrograde binary orbits can be ruled out. Given its applicability over a broad range of protoplanetary disk conditions, it is expected that the streaming instability seeded planetesimal formation also elsewhere in the solar system, and beyond. △ Less

Submitted 26 June, 2019; originally announced June 2019.

Comments: Nature Astronomy

The Nature of Turbulence in the Outer Regions of Protoplanetary Disks

Authors: Daniel A. Gole, Jacob B. Simon

Abstract: We carry out a series of local, shearing box simulations of the outer regions of protoplanetary disks, where ambipolar diffusion is important due to low ionization levels, to better characterize the nature of turbulence and angular momentum transport in these disks. These simulations are divided into two groups, one with far ultraviolet (FUV) ionization included, and one without FUV. In both cases… ▽ More We carry out a series of local, shearing box simulations of the outer regions of protoplanetary disks, where ambipolar diffusion is important due to low ionization levels, to better characterize the nature of turbulence and angular momentum transport in these disks. These simulations are divided into two groups, one with far ultraviolet (FUV) ionization included, and one without FUV. In both cases, we explore a large range in diffusivity values. We find that in the simulations without FUV, the properties of the turbulence are similar to the unstratified simulations of Bai & Stone (2011); for a given diffusivity, the MRI can still be present so long as the magnetic field is sufficiently weak. Furthermore, the dynamics of the mid-plane in these simulations are primarily controlled by the MRI. In the FUV simulations on the other hand, the MRI-active FUV layers transport strong toroidal magnetic flux to the mid-plane, which shuts off the MRI. Instead, angular momentum transport at the mid-plane is dominated by laminar magnetic fields, resulting in lower levels of turbulent Maxwell stress compared to the no-FUV simulations. Finally, we perform a temporal correlation analysis on the FUV simulations, confirming our result that the dynamics in the mid-plane region is strongly controlled by the FUV-ionized layers. △ Less

Submitted 24 October, 2018; originally announced October 2018.

Comments: 12 pages, 13 figures

The Mass and Size Distribution of Planetesimals Formed by the Streaming Instability. II. The Effect of the Radial Gas Pressure Gradient

Authors: Charles P. Abod, Jacob B. Simon, Rixin Li, Philip J. Armitage, Andrew N. Youdin, Katherine A. Kretke

Abstract: The streaming instability concentrates solid particles in protoplanetary disks, leading to gravitational collapse into planetesimals. Despite its key role in producing particle clumping and determining critical length scales in the instability's linear regime, the influence of the disk's radial pressure gradient on planetesimal properties has not been examined in detail. Here, we use streaming ins… ▽ More The streaming instability concentrates solid particles in protoplanetary disks, leading to gravitational collapse into planetesimals. Despite its key role in producing particle clumping and determining critical length scales in the instability's linear regime, the influence of the disk's radial pressure gradient on planetesimal properties has not been examined in detail. Here, we use streaming instability simulations that include particle self-gravity to study how the planetesimal initial mass function depends on the radial pressure gradient. Fitting our results to a power-law, ${\rm d}N / {\rm d}M_p \propto M_p^{-p}$, we find a single value $p \approx 1.6$ describes simulations in which the pressure gradient varies by $\gtrsim 2$. An exponentially truncated power-law provides a significantly better fit, with a low mass slope of $p^\prime \approx 1.3$ that weakly depends on the pressure gradient. The characteristic truncation mass is found to be $\sim M_G = 4 π^5 G^2 Σ_p^3 / Ω^4$. We exclude the cubic dependence of the characteristic mass with pressure gradient suggested by linear considerations, finding instead a linear scaling. These results strengthen the case for a streaming-derived initial mass function that depends at most weakly on the aerodynamic properties of the disk and participating solids. A simulation initialized with zero pressure gradient---which is {\em not} subject to the streaming instability---also yields a top-heavy mass function but with modest evidence for a different shape. We discuss the consistency of the theoretically predicted mass function with observations of Kuiper Belt planetesimals, and describe implications for models of early stage planet formation.. △ Less

Submitted 29 August, 2019; v1 submitted 23 October, 2018; originally announced October 2018.

Comments: 18 pages, 10 figures, 3 tables, accepted to ApJ

Turbulence in the TW Hya Disk

Authors: Kevin M. Flaherty, A. Meredith Hughes, Richard Teague, Jacob B. Simon, Sean M. Andrews, David J. Wilner

Abstract: Turbulence is a fundamental parameter in models of grain growth during the early stages of planet formation. As such, observational constraints on its magnitude are crucial. Here we self-consistently analyze ALMA CO(2-1), SMA CO(3-2), and SMA CO(6-5) observations of the disk around TW Hya and find an upper limit on the turbulent broadening of $<$0.08c$_s$ ($α<$0.007 for $α$ defined only within 2-3… ▽ More Turbulence is a fundamental parameter in models of grain growth during the early stages of planet formation. As such, observational constraints on its magnitude are crucial. Here we self-consistently analyze ALMA CO(2-1), SMA CO(3-2), and SMA CO(6-5) observations of the disk around TW Hya and find an upper limit on the turbulent broadening of $<$0.08c$_s$ ($α<$0.007 for $α$ defined only within 2-3 pressure scale heights above the midplane), lower than the tentative detection previously found from an analysis of the CO(2-1) data. We examine in detail the challenges of image plane fitting vs directly fitting the visibilities, while also considering the role of the vertical temperature gradient, systematic uncertainty in the amplitude calibration, and assumptions about the CO abundance, as potential sources of the discrepancy in the turbulence measurements. These tests result in variations of the turbulence limit between $<$0.04c$_s$ and $<$0.13c$_s$, consistently lower than the 0.2-0.4c$_s$ found previously. Having ruled out numerous factors, we restrict the source of the discrepancy to our assumed coupling between temperature and density through hydrostatic equilibrium in the presence of a vertical temperature gradient and/or the confinement of CO to a thin molecular layer above the midplane, although further work is needed to quantify the influence of these prescriptions. Assumptions about hydrostatic equilibrium and the CO distribution are physically motivated, and may have a small influence on measuring the kinematics of the gas, but they become important when constraining small effects such as the strength of the turbulence within a protoplanetary disk. △ Less

Submitted 10 March, 2018; originally announced March 2018.

Comments: 14 pages, 4 figures, accepted to ApJ

On the Numerical Robustness of the Streaming Instability: Particle Concentration and Gas Dynamics in Protoplanetary Disks

Authors: Rixin Li, Andrew N. Youdin, Jacob B. Simon

Abstract: The Streaming Instability (SI) is a mechanism to concentrate solids in protoplanetary disks. Nonlinear particle clumping from the SI can trigger gravitational collapse into planetesimals. To better understand the numerical robustness of the SI, we perform a suite of vertically-stratified 3D simulations with fixed physical parameters known to produce strong clumping. We vary the numerical implement… ▽ More The Streaming Instability (SI) is a mechanism to concentrate solids in protoplanetary disks. Nonlinear particle clumping from the SI can trigger gravitational collapse into planetesimals. To better understand the numerical robustness of the SI, we perform a suite of vertically-stratified 3D simulations with fixed physical parameters known to produce strong clumping. We vary the numerical implementation, namely the computational domain size and the vertical boundary conditions (vBCs), comparing newly-implemented outflow vBCs to the previously-used periodic and reflecting vBCs. We find strong particle clumping by the SI is mostly independent of the vBCs. However, peak particle densities are higher in larger simulation domains due to a larger particle mass reservoir. We report SI-triggered zonal flows, i.e., azimuthally-banded radial variations of gas pressure. These structures have low amplitudes, insufficient to halt particle radial drift, confirming that particle trapping in gas pressure maxima is not the mechanism of the SI. We find that outflow vBCs produce artificially large gas outflow rates at vertical boundaries. However, the outflow vBCs reduce artificial reflections at vertical boundaries, allowing more particle sedimentation, and showing less temporal variation and better convergence with box size. The radial spacing of dense particle filaments is $\sim0.15$ gas scale heights ($H$) for all vBCs, which sets the feeding zone for planetesimal growth in self-gravitating simulations. Our results validate the use of the outflow vBCs in SI simulations, even with vertical boundaries close ($\leq 0.4H$) to the disk midplane. Overall, our study demonstrates the numerical robustness of nonlinear particle clumping by the SI. △ Less

Submitted 12 June, 2018; v1 submitted 9 March, 2018; originally announced March 2018.

Comments: 20 pages, 14 figures, accepted for publication in the ApJ

Origin of Weak Turbulence in the Outer Regions of Protoplanetary Disks

Authors: Jacob B. Simon, Xue-Ning Bai, Kevin M. Flaherty, A. Meredith Hughes

Abstract: The mechanism behind angular momentum transport in protoplanetary disks, and whether this transport is turbulent in nature, is a fundamental issue in planet formation studies. Recent ALMA observations have suggested that turbulent velocities in the outer regions of these disks are less than ~5-10% of the sound speed, contradicting theoretical predictions of turbulence driven by the magnetorotation… ▽ More The mechanism behind angular momentum transport in protoplanetary disks, and whether this transport is turbulent in nature, is a fundamental issue in planet formation studies. Recent ALMA observations have suggested that turbulent velocities in the outer regions of these disks are less than ~5-10% of the sound speed, contradicting theoretical predictions of turbulence driven by the magnetorotational instability (MRI). These observations have generally been interpreted to be consistent with a large-scale laminar magnetic wind driving accretion. Here, we carry out local, shearing box simulations with varying ionization levels and background magnetic field strengths in order to determine which parameters produce results consistent with observations. We find that even when the background magnetic field launches a strong largely laminar wind, significant turbulence persists and is driven by localized regions of vertical magnetic field (the result of zonal flows) that are unstable to the MRI. The only conditions for which we find turbulent velocities below the observational limits are weak background magnetic fields and ionization levels well below that usually assumed in theoretical studies. We interpret these findings within the context of a preliminary model in which a large scale magnetic field, confined to the inner disk, hinders ionizing sources from reaching large radial distances, e.g., through a sufficiently dense wind. Thus, in addition to such a wind, this model predicts that for disks with weakly turbulent outer regions, the outer disk will have significantly reduced ionization levels compared to standard models and will harbor only a weak vertical magnetic field. △ Less

Submitted 6 August, 2018; v1 submitted 13 November, 2017; originally announced November 2017.

Comments: Accepted to ApJ, 13 pages, 6 figures, modified from original version to include more extensive discussion and analysis and a title change

A Three-Dimensional View of Turbulence: Constraints on Turbulent Motions in the HD 163296 Protoplanetary Disk using DCO$^+$

Authors: Kevin M. Flaherty, A. Meredith Hughes, Sanaea C. Rose, Jacob B. Simon, Chunhua Qi, Sean M. Andrews, Agnes Kospal, David J. Wilner, Eugene Chiang, Philip J. Armitage, Xue-ning Bai

Abstract: Gas kinematics are an important part of the planet formation process. Turbulence influences planetesimal growth and migration from the scale of sub-micron dust grains through gas-giant planets. Radio observations of resolved molecular line emission can directly measure this non-thermal motion and, taking advantage of the layered chemical structure of disks, different molecular lines can be combine… ▽ More Gas kinematics are an important part of the planet formation process. Turbulence influences planetesimal growth and migration from the scale of sub-micron dust grains through gas-giant planets. Radio observations of resolved molecular line emission can directly measure this non-thermal motion and, taking advantage of the layered chemical structure of disks, different molecular lines can be combined to map the turbulence throughout the vertical extent of a protoplanetary disk. Here we present ALMA observations of three molecules (DCO$^+$(3-2), C$^{18}$O(2-1) and CO(2-1)) from the disk around HD 163296. We are able to place stringent upper limits ($v_{\rm turb}<$0.06c$_s$, $<$0.05c$_s$ and $<$0.04c$_s$ for CO(2-1), C$^{18}$O(2-1) and DCO$^+$(3-2) respectively), corresponding to $α\lesssim$3$\times$10$^{-3}$, similar to our prior limit derived from CO(3-2). This indicates that there is little turbulence throughout the vertical extent of the disk, contrary to theoretical predictions based on the magneto-rotational instability and gravito-turbulence. In modeling the DCO$^+$ emission we also find that it is confined to three concentric rings at 65.7$\pm$0.9 au, 149.9$^{+0.5}_{-0.7}$ au and 259$\pm$1 au, indicative of a complex chemical environment. △ Less

Submitted 14 June, 2017; originally announced June 2017.

Comments: Accepted to ApJ

Evidence for universality in the initial planetesimal mass function

Authors: Jacob B. Simon, Philip J. Armitage, Andrew N. Youdin, Rixin Li

Abstract: Planetesimals may form from the gravitational collapse of dense particle clumps initiated by the streaming instability. We use simulations of aerodynamically coupled gas-particle mixtures to investigate whether the properties of planetesimals formed in this way depend upon the sizes of the particles that participate in the instability. Based on three high resolution simulations that span a range o… ▽ More Planetesimals may form from the gravitational collapse of dense particle clumps initiated by the streaming instability. We use simulations of aerodynamically coupled gas-particle mixtures to investigate whether the properties of planetesimals formed in this way depend upon the sizes of the particles that participate in the instability. Based on three high resolution simulations that span a range of dimensionless stopping time $6 \times 10^{-3} \leq τ\leq 2$ no statistically significant differences in the initial planetesimal mass function are found. The mass functions are fit by a power-law, ${\rm d}N / {\rm d}M_p \propto M_p^{-p}$, with $p=1.5-1.7$ and errors of $Δp \approx 0.1$. Comparing the particle density fields prior to collapse, we find that the high wavenumber power spectra are similarly indistinguishable, though the large-scale geometry of structures induced via the streaming instability is significantly different between all three cases. We interpret the results as evidence for a near-universal slope to the mass function, arising from the small-scale structure of streaming-induced turbulence. △ Less

Submitted 13 September, 2017; v1 submitted 10 May, 2017; originally announced May 2017.

Comments: 7 pages, 4 figures, accepted to ApJ Letters after minor modifications, including two new figures and some new text that better clarify our results

Prompt planetesimal formation beyond the snow line

Authors: Philip J. Armitage, Josh A. Eisner, Jacob B. Simon

Abstract: We develop a simple model to predict the radial distribution of planetesimal formation. The model is based on the observed growth of dust to mm-sized particles, which drift radially, pile-up, and form planetesimals where the stopping time and dust-to-gas ratio intersect the allowed region for streaming instability-induced gravitational collapse. Using an approximate analytic treatment, we first sh… ▽ More We develop a simple model to predict the radial distribution of planetesimal formation. The model is based on the observed growth of dust to mm-sized particles, which drift radially, pile-up, and form planetesimals where the stopping time and dust-to-gas ratio intersect the allowed region for streaming instability-induced gravitational collapse. Using an approximate analytic treatment, we first show that drifting particles define a track in metallicity--stopping time space whose only substantial dependence is on the disk's angular momentum transport efficiency. Prompt planetesimal formation is feasible for high particle accretion rates (relative to the gas, $\dot{M}_p / \dot{M} > 3 \times 10^{-2}$ for $α= 10^{-2}$), that could only be sustained for a limited period of time. If it is possible, it would lead to the deposition of a broad and massive belt of planetesimals with a sharp outer edge. Including turbulent diffusion and vapor condensation processes numerically, we find that a modest enhancement of solids near the snow line occurs for cm-sized particles, but that this is largely immaterial for planetesimal formation. We note that radial drift couples planetesimal formation across radii in the disk, and suggest that considerations of planetesimal formation favor a model in which the initial deposition of material for giant planet cores occurs well beyond the snow line. △ Less

Submitted 11 August, 2016; originally announced August 2016.

Comments: ApJ Letters, in press

The Influence of Magnetic Field Geometry on the Formation of Close-in Exoplanets

Authors: Jacob B. Simon

Abstract: Approximately half of Sun-like stars harbor exoplanets packed within a radius of ~0.3 AU, but the formation of these planets and why they form in only half of known systems are still not well understood. We employ a one-dimensional steady state model to gain physical insight into the origin of these close-in exoplanets. We use Shakura & Sunyaev alpha values extracted from recent numerical simulati… ▽ More Approximately half of Sun-like stars harbor exoplanets packed within a radius of ~0.3 AU, but the formation of these planets and why they form in only half of known systems are still not well understood. We employ a one-dimensional steady state model to gain physical insight into the origin of these close-in exoplanets. We use Shakura & Sunyaev alpha values extracted from recent numerical simulations of protoplanetary disk accretion processes in which the magnitude of alpha, and thus the steady-state gas surface density, depends on the orientation of large scale magnetic fields with respect to the disk's rotation axis. Solving for the metallicity as a function of radius, we find that for fields anti-aligned with the rotation axis, the inner regions of our model disk often falls within a region of parameter space not suitable for planetesimal formation, whereas in the aligned case, the inner disk regions are likely to produce planetesimals through some combination of streaming instability and gravitational collapse, though the degree to which this is true depends on the assumed parameters of our model. More robustly, the aligned field case always produces higher concentrations of solids at small radii compared to the anti-aligned case. In the in situ formation model, this bimodal distribution of solid enhancement leads directly to the observed dichotomy in exoplanet orbital distances. △ Less

Submitted 1 August, 2016; originally announced August 2016.

Comments: 6 pages, 4 figures, accepted for publication in ApJ Letters

Turbulence, Transport and Waves in Ohmic Dead Zones

Authors: Daniel Gole, Jacob B. Simon, Stephen H. Lubow, Philip J. Armitage

Abstract: We use local numerical simulations to study a vertically stratified accretion disk with a resistive mid-plane that damps magnetohydrodynamic (MHD) turbulence. This is an idealized model for the dead zones that may be present at some radii in protoplanetary and dwarf novae disks. We vary the relative thickness of the dead and active zones to quantify how forced fluid motions in the dead zone change… ▽ More We use local numerical simulations to study a vertically stratified accretion disk with a resistive mid-plane that damps magnetohydrodynamic (MHD) turbulence. This is an idealized model for the dead zones that may be present at some radii in protoplanetary and dwarf novae disks. We vary the relative thickness of the dead and active zones to quantify how forced fluid motions in the dead zone change. We find that the residual Reynolds stress near the mid-plane decreases with increasing dead zone thickness, becoming negligible in cases where the active to dead mass ratio is less than a few percent. This implies that purely Ohmic dead zones would be vulnerable to episodic accretion outbursts via the mechanism of Martin & Lubow (2011). We show that even thick dead zones support a large amount of kinetic energy, but this energy is largely in fluid motions that are inefficient at angular momentum transport. Confirming results from Oishi & Mac Low (2009), the perturbed velocity field in the dead zone is dominated by an oscillatory, vertically extended circulation pattern with a low frequency compared to the orbital frequency. This disturbance has the properties predicted for the lowest order r mode in a hydrodynamic disk. We suggest that in a global disk similar excitations would lead to propagating waves, whose properties would vary with the thickness of the dead zone and the nature of the perturbations (isothermal or adiabatic). Flows with similar amplitudes would buckle settled particle layers and could reduce the efficiency of pebble accretion. △ Less

Submitted 9 June, 2016; originally announced June 2016.

Comments: 14 pages, 17 figures

Strongly magnetized accretion discs require poloidal flux

Authors: Greg Salvesen, Philip J. Armitage, Jacob B. Simon, Mitchell C. Begelman

Abstract: Motivated by indirect observational evidence for strongly magnetized accretion discs around black holes, and the novel theoretical properties of such solutions, we investigate how a strong magnetization state can develop and persist. To this end, we perform local simulations of accretion discs with an initially purely toroidal magnetic field of equipartition strength. We demonstrate that discs wit… ▽ More Motivated by indirect observational evidence for strongly magnetized accretion discs around black holes, and the novel theoretical properties of such solutions, we investigate how a strong magnetization state can develop and persist. To this end, we perform local simulations of accretion discs with an initially purely toroidal magnetic field of equipartition strength. We demonstrate that discs with zero net vertical magnetic flux and realistic boundary conditions cannot sustain a strong toroidal field. However, a magnetic pressure-dominated disc can form from an initial configuration with a sufficient amount of net vertical flux and realistic boundary conditions. Our results suggest that poloidal flux is a necessary prerequisite for the sustainability of strongly magnetized accretion discs. △ Less

Submitted 18 May, 2016; v1 submitted 15 February, 2016; originally announced February 2016.

Comments: 7 pages, 3 figures, 1 table; Accepted to MNRAS

The Mass and Size Distribution of Planetesimals Formed by the Streaming Instability. I. The Role of Self-Gravity

Authors: Jacob B. Simon, Philip J. Armitage, Rixin Li, Andrew N. Youdin

Abstract: We study the formation of planetesimals in protoplanetary disks from the gravitational collapse of solid over-densities generated via the streaming instability. To carry out these studies, we implement and test a particle-mesh self-gravity module for the Athena code that enables the simulation of aerodynamically coupled systems of gas and collisionless self-gravitating solid particles. Upon employ… ▽ More We study the formation of planetesimals in protoplanetary disks from the gravitational collapse of solid over-densities generated via the streaming instability. To carry out these studies, we implement and test a particle-mesh self-gravity module for the Athena code that enables the simulation of aerodynamically coupled systems of gas and collisionless self-gravitating solid particles. Upon employment of our algorithm to planetesimal formation simulations, we find that (when a direct comparison is possible) the Athena simulations yield predicted planetesimal properties that agree well with those found in prior work using different numerical techniques. In particular, the gravitational collapse of streaming-initiated clumps leads to an initial planetesimal mass function that is well-represented by a power-law, dN/dM ~ M^{-p}, with p = 1.6 +/- 0.1, which equates to a differential size distribution dN/dR ~ R^{-q}, with q = 2.8 +/- 0.1. We find no significant trends with resolution from a convergence study of up to 512^3 grid zones and Npar = 1.5x10^8 particles. Likewise, the power-law slope appears indifferent to changes in the relative strength of self-gravity and tidal shear, and to the time when (for reasons of numerical economy) self-gravity is turned on, though the strength of these claims is limited by small number statistics. For a typically assumed radial distribution of minimum mass solar nebula solids (assumed here to have dimensionless stopping time τ= 0.3), our results support the hypothesis that bodies on the scale of large asteroids or Kuiper Belt Objects could have formed as the high-mass tail of a primordial planetesimal population. △ Less

Submitted 31 March, 2016; v1 submitted 30 November, 2015; originally announced December 2015.

Comments: 18 pages, 16 figures, accepted to ApJ after minor revisions and title change

Accretion disc dynamo activity in local simulations spanning weak-to-strong net vertical magnetic flux regimes

Authors: Greg Salvesen, Jacob B. Simon, Philip J. Armitage, Mitchell C. Begelman

Abstract: Strongly magnetized accretion discs around black holes have attractive features that may explain enigmatic aspects of X-ray binary behaviour. The structure and evolution of these discs are governed by a dynamo-like mechanism, which channels part of the accretion power liberated by the magnetorotational instability (MRI) into an ordered toroidal magnetic field. To study dynamo activity, we performe… ▽ More Strongly magnetized accretion discs around black holes have attractive features that may explain enigmatic aspects of X-ray binary behaviour. The structure and evolution of these discs are governed by a dynamo-like mechanism, which channels part of the accretion power liberated by the magnetorotational instability (MRI) into an ordered toroidal magnetic field. To study dynamo activity, we performed three-dimensional, stratified, isothermal, ideal magnetohydrodynamic shearing box simulations. The strength of the self-sustained toroidal magnetic field depends on the net vertical magnetic flux, which we vary across almost the entire range over which the MRI is linearly unstable. We quantify disc structure and dynamo properties as a function of the initial ratio of mid-plane gas pressure to vertical magnetic field pressure, $β_0^{\rm mid} = p_{\rm gas} / p_B$. For $10^5 \geq β_0^{\rm mid} \geq 10$ the effective $α$-viscosity parameter scales as a power-law. Dynamo activity persists up to and including $β_0^{\rm mid} = 10^2$, at which point the entire vertical column of the disc is magnetic pressure-dominated. Still stronger fields result in a highly inhomogeneous disc structure, with large density fluctuations. We show that the turbulent steady state $β^{\rm mid}$ in our simulations is well-matched by the analytic model of Begelman et al. (2015) describing the creation and buoyant escape of toroidal field, while the vertical structure of the disc can be broadly reproduced using this model. Finally, we discuss the implications of our results for observed properties of X-ray binaries. △ Less

Submitted 12 January, 2016; v1 submitted 19 November, 2015; originally announced November 2015.

Comments: 19 pages, 13 figures, 3 tables; Accepted to MNRAS

Weak Turbulence in the HD 163296 Protoplanetary Disk Revealed by ALMA CO Observations

Authors: Kevin M. Flaherty, A. Meredith Hughes, Katherine A. Rosenfeld, Sean M. Andrews, Eugene Chiang, Jacob B. Simon, Skylar Kerzner, David J. Wilner

Abstract: Turbulence can transport angular momentum in protoplanetary disks and influence the growth and evolution of planets. With spatially and spectrally resolved molecular emission line measurements provided by (sub)millimeter interferometric observations, it is possible to directly measure non-thermal motions in the disk gas that can be attributed to this turbulence. We report a new constraint on the t… ▽ More Turbulence can transport angular momentum in protoplanetary disks and influence the growth and evolution of planets. With spatially and spectrally resolved molecular emission line measurements provided by (sub)millimeter interferometric observations, it is possible to directly measure non-thermal motions in the disk gas that can be attributed to this turbulence. We report a new constraint on the turbulence in the disk around HD 163296, a nearby young A star, determined from ALMA Science Verification observations of four CO emission lines (the CO(3-2), CO(2-1), 13CO(2-1), and C18O(2-1) transitions). The different optical depths for these lines permit probes of non-thermal line-widths at a range of physical conditions (temperature and density) and depths into the disk interior. We derive stringent limits on the non-thermal motions in the upper layers of the outer disk such that any contribution to the line-widths from turbulence is <3% of the local sound speed. These limits are approximately an order of magnitude lower than theoretical predictions for full-blown MHD turbulence driven by the magneto-rotational instability, potentially suggesting that this mechanism is less efficient in the outer (R>30AU) disk than has been previously considered. △ Less

Submitted 5 October, 2015; originally announced October 2015.

Comments: Accepted to ApJ

Magnetically driven accretion in protoplanetary discs

Authors: Jacob B. Simon, Geoffroy Lesur, Matthew W. Kunz, Philip J. Armitage

Abstract: We characterize magnetically driven accretion at radii between 1 au and 100 au in protoplanetary discs, using a series of local non-ideal magnetohydrodynamic (MHD) simulations. The simulations assume a Minimum Mass Solar Nebula (MMSN) disc that is threaded by a net vertical magnetic field of specified strength. Confirming previous results, we find that the Hall effect has only a modest impact on a… ▽ More We characterize magnetically driven accretion at radii between 1 au and 100 au in protoplanetary discs, using a series of local non-ideal magnetohydrodynamic (MHD) simulations. The simulations assume a Minimum Mass Solar Nebula (MMSN) disc that is threaded by a net vertical magnetic field of specified strength. Confirming previous results, we find that the Hall effect has only a modest impact on accretion at 30 au, and essentially none at 100 au. At 1-10 au the Hall effect introduces a pronounced bi-modality in the accretion process, with vertical magnetic fields aligned to the disc rotation supporting a strong laminar Maxwell stress that is absent if the field is anti-aligned. In the anti-aligned case, we instead find evidence for bursts of turbulent stress at 5-10 au, which we tentatively identify with the non-axisymmetric Hall-shear instability. The presence or absence of these bursts depends upon the details of the adopted chemical model, which suggests that appreciable regions of actual protoplanetary discs might lie close to the borderline between laminar and turbulent behaviour. Given the number of important control parameters that have already been identified in MHD models, quantitative predictions for disc structure in terms of only radius and accretion rate appear to be difficult. Instead, we identify robust qualitative tests of magnetically driven accretion. These include the presence of turbulence in the outer disc, independent of the orientation of the vertical magnetic fields, and a Hall-mediated bi-modality in turbulent properties extending from the region of thermal ionization to 10 au. △ Less

Submitted 4 September, 2015; v1 submitted 4 August, 2015; originally announced August 2015.

Comments: accepted to MNRAS after very minor revisions

Signatures of MRI-Driven Turbulence in Protoplanetary Disks: Predictions for ALMA Observations

Authors: Jacob B. Simon, A. Meredith Hughes, Kevin M. Flaherty, Xue-Ning Bai, Philip J. Armitage

Abstract: Spatially resolved observations of molecular line emission have the potential to yield unique constraints on the nature of turbulence within protoplanetary disks. Using a combination of local non-ideal magnetohydrodynamic simulations and radiative transfer calculations, tailored to properties of the disk around HD 163296, we assess the ability of ALMA to detect turbulence driven by the magnetorota… ▽ More Spatially resolved observations of molecular line emission have the potential to yield unique constraints on the nature of turbulence within protoplanetary disks. Using a combination of local non-ideal magnetohydrodynamic simulations and radiative transfer calculations, tailored to properties of the disk around HD 163296, we assess the ability of ALMA to detect turbulence driven by the magnetorotational instability (MRI). Our local simulations show that the MRI produces small-scale turbulent velocity fluctuations that increase in strength with height above the mid-plane. For a set of simulations at different disk radii, we fit a Maxell-Boltzmann distribution to the turbulent velocity and construct a turbulent broadening parameter as a function of radius and height. We input this broadening into radiative transfer calculations to quantify observational signatures of MRI-driven disk turbulence. We find that the ratio of the peak line flux to the flux at line center is a robust diagnostic of turbulence that is only mildly degenerate with systematic uncertainties in disk temperature. For the CO(3-2) line, which we expect to probe the most magnetically active slice of the disk column, variations in the predicted peak-to-trough ratio between our most and least turbulent models span a range of approximately 15%. Additional independent constraints can be derived from the morphology of spatially resolved line profiles, and we estimate the resolution required to detect turbulence on different spatial scales. We discuss the role of lower optical depth molecular tracers, which trace regions closer to the disk mid-plane where velocities in MRI-driven models are systematically lower. △ Less

Submitted 22 June, 2015; v1 submitted 12 January, 2015; originally announced January 2015.

Comments: accepted to ApJ after minor revisions

Efficiency of Particle Trapping in the Outer Regions of Protoplanetary Disks

Authors: Jacob B. Simon, Philip J. Armitage

Abstract: We investigate the strength of axisymmetric local pressure maxima (zonal flows) in the outer regions of protoplanetary disks, where ambipolar diffusion reduces turbulent stresses driven by the magnetorotational instability. Using local numerical simulations we show that in the absence of net vertical magnetic fields, the strength of turbulence in the ambipolar dominated region of the disk is low a… ▽ More We investigate the strength of axisymmetric local pressure maxima (zonal flows) in the outer regions of protoplanetary disks, where ambipolar diffusion reduces turbulent stresses driven by the magnetorotational instability. Using local numerical simulations we show that in the absence of net vertical magnetic fields, the strength of turbulence in the ambipolar dominated region of the disk is low and any zonal flows that are present are weak. For net fields strong enough to yield observed protostellar accretion rates, however, zonal flows with a density amplitude of 10-20% are formed. These strengths are comparable to those seen in simulations of ideal MHD disk turbulence. We investigate whether these zonal flows are able to reverse the inward radial drift of solids, leading to prolonged and enhanced concentration as a prelude to planetesimal formation. For commonly assumed mean surface density profiles (surface density proportional to radius to the -1/2 power or steeper) we find that the predicted perturbations to the background disk profile do not correspond to local pressure maxima. This is a consequence of radial width of the simulated zonal flows, which is larger than was assumed in prior analytic models of particle trapping. These larger scale flows would only trap particles for higher amplitude fluctuations than observed. We conclude that zonal flows are likely to be present in the outer regions of protoplanetary disks and are potentially large enough to be observable, but are unlikely to lead to strong particle trapping. △ Less

Submitted 6 February, 2014; originally announced February 2014.

Comments: accepted for publication in ApJ, 8 pages, 5 figures