arXiv subjects

Kristen Menou

Publications and source records attributed to Kristen Menou.

At least 37 records · Page 2Linked to original sources

arXiv2020

Automated crater shape retrieval using weakly-supervised deep learning

Crater ellipticity determination is a complex and time consuming task that so far has evaded successful automation. We train a state of the art computer vision algorithm to identify craters in Lunar digital elevation maps and retrieve their sizes and 2D shapes. The computational backbone of the model is MaskRCNN, an "instance segmentation" general framework that detects craters in an image while simultaneously producing a mask for each crater that traces its outer rim. Our post-processing pipeline then finds the closest fitting ellipse to these masks, allowing us to retrieve the crater ellipticities. Our model is able to correctly identify 87\% of known craters in the longitude range we hid from the network during training and validation (test set), while predicting thousands of additional craters not present in our training data. Manual validation of a subset of these "new" craters indicates that a majority of them are real, which we take as an indicator of the strength of our model in learning to identify craters, despite incomplete training data. The crater size, ellipticity, and depth distributions predicted by our model are consistent with human-generated results. The model allows us to perform a large scale search for differences in crater diameter and shape distributions between the lunar highlands and maria, and we exclude any such differences with a high statistical significance. The predicted test set catalogue and trained model are available here: https://github.com/malidib/Craters_MaskRCNN/.

astro-ph.EP↗

arXiv2019

Evolution of the radius valley around low mass stars from $Kepler$ and $K2$

We present calculations of the occurrence rate of small close-in planets around low mass dwarf stars using the known planet populations from the $Kepler$ and $K2$ missions. Applying completeness corrections clearly reveals the radius valley in the maximum a-posteriori occurrence rates as a function of orbital separation and planet radius. We measure the slope of the valley to be $r_{p,\text{valley}} \propto F^{-0.060\pm 0.025}$ which bears the opposite sign from that measured around Sun-like stars thus suggesting that thermally driven atmospheric mass loss may not dominate the evolution of planets in the low stellar mass regime or that we are witnessing the emergence of a separate channel of planet formation. The latter notion is supported by the relative occurrence of rocky to non-rocky planets increasing from $0.5\pm 0.1$ around mid-K dwarfs to $8.5\pm 4.6$ around mid-M dwarfs. Furthermore, the center of the radius valley at $1.54\pm 0.16$ R$_{\oplus}$ is shown to shift to smaller sizes with decreasing stellar mass in agreement with physical models of photoevaporation, core-powered mass loss, and gas-poor formation. Although current measurements are insufficient to robustly identify the dominant formation pathway of the radius valley, such inferences may be obtained by $TESS$ with $\mathcal{O}(85,000)$ mid-to-late M dwarfs observed with 2-minute cadence. The measurements presented herein also precisely designate the subset of planetary orbital periods and radii that should be targeted in radial velocity surveys to resolve the rocky to non-rocky transition around low mass stars.

astro-ph.EP↗

arXiv2019

Hot Jupiter Atmospheric Flows at High Resolution

Global Circulation Models (GCMs) of atmospheric flows are now routinely used to interpret observational data on Hot Jupiters. Localized "equatorial $β$-plane" simulations by Fromang et al. (2016) have revealed that a barotropic (horizontal shear) instability of the equatorial jet appears at horizontal resolutions beyond those typically achieved in global models; this instability could limit wind speeds and lead to increased atmospheric variability. To address this possibility, we adapt the computationally efficient, pseudo-spectral PlaSim GCM, originally designed for Earth studies, to model Hot Jupiter atmospheric flows and validate it on the Heng et al. (2011) reference benchmark. We then present high resolution global models of HD209458b, with horizontal resolutions of T85 (128x256) and T127 (192x384). The barotropic instability phenomenology found in $β$-plane simulations is not reproduced in these global models, despite comparably high resolutions. Nevertheless, high resolution models do exhibit additional flow variability on long timescales (of order 100 planet days or more), which is absent from the lower resolution models. It manifests as a breakdown of north-south symmetry of the equatorial wind. From post-processing the atmospheric flows at various resolutions (assuming a cloud-free situation), we show that the stronger flow variability achieved at high resolution does not translate into noticeably stronger dayside infrared flux variability. More generally, our results suggest that high horizontal resolutions are not required to capture the key features of hot Jupiter atmospheric flows.

astro-ph.EP↗

arXiv2019

Morpho-Photometric Redshifts

Machine learning (ML) is a standard approach for estimating the redshifts of galaxies when only photometric information is available. ML photo-z solutions have traditionally ignored the morphological information available in galaxy images or partly included it in the form of hand-crafted features, with mixed results. We train a morphology-aware photometric redshift machine using modern deep learning tools. It uses a custom architecture that jointly trains on galaxy fluxes, colors and images. Galaxy-integrated quantities are fed to a Multi-Layer Perceptron (MLP) branch while images are fed to a convolutional (convnet) branch that can learn relevant morphological features. This split MLP-convnet architecture, which aims to disentangle strong photometric features from comparatively weak morphological ones, proves important for strong performance: a regular convnet-only architecture, while exposed to all available photometric information in images, delivers comparatively poor performance. We present a cross-validated MLP-convnet model trained on 130,000 SDSS-DR12 galaxies that outperforms a hyperoptimized Gradient Boosting solution (hyperopt+XGBoost), as well as the equivalent MLP-only architecture, on the redshift bias metric. The 4-fold cross-validated MLP-convnet model achieves a bias $δz / (1+z) =-0.70 \pm 1 \times 10^{-3} $, approaching the performance of a reference ANNZ2 ensemble of 100 distinct models trained on a comparable dataset. The relative performance of the morphology-aware and morphology-blind models indicates that galaxy morphology does improve ML-based photometric redshift estimation.

astro-ph.IM↗

arXiv2019

Habitable Snowballs: Temperate Land Conditions, Liquid Water, and Implications for CO$_2$ Weathering

Habitable planets are commonly imagined to be temperate planets like Earth, with areas of open ocean and warm land. In contrast, planets in snowball states, where oceans are entirely ice-covered, are believed to be inhospitable. However, we show using a general circulation model that terrestrial planets in the inner habitable zone are able to support large unfrozen areas of land while in a snowball state. Due to their lower albedo, these unfrozen regions reach summer temperatures in excess of 10 $^\circ$Celsius. Such conditions permit CO$_2$ weathering, suggesting that continental weathering can provide a mechanism for trapping planets in stable snowball states. The presence of land areas with warm temperatures and liquid surface water motivates a more-nuanced understanding of habitability during these snowball events.

astro-ph.EP↗

arXiv2019

Turbulent Vertical Mixing in Hot Exoplanet Atmospheres

Turbulent vertical transport driven by double-diffusive shear instabilities is identified as likely important in hot exoplanet atmospheres. In hot Jupiter atmospheres, the resulting vertical mixing appears sufficient to alleviate the nightside cold trap, thus facilitating the maintenance of nocturnal clouds on these planets. The strong level of vertical mixing expected near hot Jupiter thermal photospheres will impact their atmospheric chemistry and even their vertical structures where cloud radiative feedback proves important.

astro-ph.EP↗

arXiv2018

Lunar Crater Identification via Deep Learning

Crater counting on the Moon and other bodies is crucial to constrain the dynamical history of the Solar System. This has traditionally been done by visual inspection of images, thus limiting the scope, efficiency, and/or accuracy of retrieval. In this paper we demonstrate the viability of using convolutional neural networks (CNNs) to determine the positions and sizes of craters from Lunar digital elevation maps (DEMs). We recover 92% of craters from the human-generated test set and almost double the total number of crater detections. Of these new craters, 15% are smaller in diameter than the minimum crater size in the ground-truth dataset. Our median fractional longitude, latitude and radius errors are 11% or less, representing good agreement with the human-generated datasets. From a manual inspection of 361 new craters we estimate the false positive rate of new craters to be 11%. Moreover, our Moon-trained CNN performs well when tested on DEM images of Mercury, detecting a large fraction of craters in each map. Our results suggest that deep learning will be a useful tool for rapidly and automatically extracting craters on various Solar System bodies. We make our code and data publicly available at https://github.com/silburt/DeepMoon.git and https://doi.org/10.5281/zenodo.1133969 .

astro-ph.EP↗

arXiv2017

Predictions of planet detections with near infrared radial velocities in the up-coming SPIRou Legacy Survey-Planet Search

The SPIRou near infrared spectro-polarimeter is destined to begin science operations at the Canada-France-Hawaii Telescope in mid-2018. One of the instrument's primary science goals is to discover the closest exoplanets to the Solar System by conducting a 3-5 year long radial velocity survey of nearby M dwarfs at an expected precision of $\sim 1$ m s$^{-1}$; the SPIRou Legacy Survey-Planet Search (SLS-PS). In this study we conduct a detailed Monte-Carlo simulation of the SLS-PS using our current understanding of the occurrence rate of M dwarf planetary systems and physical models of stellar activity. From simultaneous modelling of planetary signals and activity, we predict the population of planets detected in the SLS-PS. With our fiducial survey strategy and expected instrument performance over a nominal survey length of $\sim 3$ years, we expect SPIRou to detect $85.3^{+29.3}_{-12.4}$ planets including $20.0^{+16.8}_{-7.2}$ habitable zone planets and $8.1^{+7.6}_{-3.2}$ Earth-like planets from a sample of 100 M1-M8.5 dwarfs out to 11 pc. By studying mid-to-late M dwarfs previously inaccessible to existing optical velocimeters, SPIRou will put meaningful constraints on the occurrence rate of planets around those stars including the value of $η_{\oplus}$ at an expected level of precision of $\lesssim 45$%. We also predict a subset of $46.7^{+16.0}_{-6.0}$ planets may be accessible with dedicated high-contrast imagers on the next generation of ELTs including $4.9^{+4.7}_{-2.0}$ potentially imagable Earth-like planets. Lastly, we compare the results of our fiducial survey strategy to other foreseeable survey versions to quantify which strategy is optimized to reach the SLS-PS science goals. The results of our simulations are made available to the community on github.

astro-ph.EP↗

arXiv2017

GCM Simulations of Unstable Climates in the Habitable Zone

It has recently been proposed that Earth-like planets in the outer regions of the habitable zone experience unstable climates, repeatedly cycling between glaciated and deglaciated climatic states (Menou 2015). While this result has been confirmed and also extended to explain early Mars climate records (Haqq-Misra et al. 2016; Batalha et al. 2016), all existing work relies on highly idealized low-dimensional climate models. Here, we confirm that the phenomenology of climate cycles remains in 3D Earth climate models with considerably more degrees of freedom. To circumvent the computational barrier of integrating climate on Gyr timescales, we design a hybrid 0D-3D integrator which uses a general circulation model (GCM) as a short relaxation step along a long evolutionary climate sequence. We find that GCM climate cycles are qualitatively consistent with reported low-dimensional results. This establishes on a firmer ground the notion that outer habitable zone planets may be preferentially found in transiently glaciated states.

astro-ph.EP↗

arXiv2017

No snowball on habitable tidally locked planets

The TRAPPIST-1, Proxima Centauri, and LHS 1140 systems are the most exciting prospects for future follow-up observations of potentially inhabited planets. All orbit nearby M-stars and are likely tidally locked in 1:1 spin-orbit states, which motivates the consideration of the effects that tidal locking might have on planetary habitability. On Earth, periods of global glaciation (snowballs) may have been essential for habitability and remote signs of life (biosignatures) because they are correlated with increases in the complexity of life and in the atmospheric oxygen concentration. In this paper we investigate the snowball bifurcation (sudden onset of global glaciation) on tidally locked planets using both an energy balance model and an intermediate-complexity global climate model. We show that tidally locked planets are unlikely to exhibit a snowball bifurcation as a direct result of the spatial pattern of insolation they receive. Instead they will smoothly transition from partial to complete ice coverage and back. A major implication of this work is that tidally locked planets with an active carbon cycle should not be found in a snowball state. Moreover, this work implies that tidally locked planets near the outer edge of the habitable zone with low CO2 outgassing fluxes will equilibrate with a small unglaciated substellar region rather than cycling between warm and snowball states. More work is needed to determine how the lack of a snowball bifurcation might affect the development of life on a tidally locked planet.

astro-ph.EP↗

arXiv2016

A Machine Learns to Predict the Stability of Tightly Packed Planetary Systems

The requirement that planetary systems be dynamically stable is often used to vet new discoveries or set limits on unconstrained masses or orbital elements. This is typically carried out via computationally expensive N-body simulations. We show that characterizing the complicated and multi-dimensional stability boundary of tightly packed systems is amenable to machine learning methods. We find that training an XGBoost machine learning algorithm on physically motivated features yields an accurate classifier of stability in packed systems. On the stability timescale investigated ($10^7$ orbits), it is 3 orders of magnitude faster than direct N-body simulations. Optimized machine learning classifiers for dynamical stability may thus prove useful across the discipline, e.g., to characterize the exoplanet sample discovered by the upcoming Transiting Exoplanet Survey Satellite (TESS). This proof of concept motivates investing computational resources to train algorithms capable of predicting stability over longer timescales and over broader regions of phase space.

astro-ph.EP↗

arXiv2016

On the radial velocity detection of additional planets in transiting, slowly rotating M-dwarf systems: the case of GJ 1132

M-dwarfs are known to commonly host high-multiplicity planetary systems. Therefore M-dwarf planetary systems with a known transiting planet are expected to contain additional small planets ($r_p \le 4$ R$_{\oplus}$, $m_p \lesssim 20$ M$_{\oplus}$) that are not seen in transit. In this study we investigate the effort required to detect such planets using precision velocimetry around the sizable subset of M-dwarfs which are slowly rotating ($P_{\mathrm{rot}} \gtrsim 40$ days) and hence more likely to be inactive. We focus on the test case of GJ 1132. Specifically, we perform a suite of Monte-Carlo simulations of the star's radial velocity signal featuring astrophysical contributions from stellar jitter due to rotationally modulated active regions and keplarian signals from the known transiting planet and hypothetical additional planets not seen in transit. We then compute the detection completeness of non-transiting planets around GJ 1132 and consequently estimate the number of RV measurements required to detect those planets. We show that with 1 m s$^{-1}$ RV precision per measurement, only $\sim 50$ measurements are required to achieve a 50\% detection completeness to all non-transiting planets in the system and to planets which are potentially habitable. Throughout we advocate the use of Gaussian process regression as an effective tool for mitigating the effects of stellar jitter including stars with high activity. Given that GJ 1132 is representative of a large population of slowly rotating M-dwarfs, we conclude with a discussion of how our results may be extended to other systems with known transiting planets such as those which will be discovered with TESS.

astro-ph.EP↗

arXiv2016

Disk-fed giant planet formation

Massive giant planets, such as the ones being discovered by direct imaging surveys, likely experience the majority of their growth through a circumplanetary disc. We argue that the entropy of accreted material is determined by boundary layer processes, unlike the "cold-" or "hot-start" hypotheses usually invoked in the core accretion and direct collapse scenarios. A simple planetary evolution model illustrates how a wide range of radius and luminosity tracks become possible, depending on details of the accretion process. Specifically, the proto-planet evolves towards "hot-start" tracks if the scale-height of the boundary layer is $\gtrsim0.24$, a value not much larger than the scale-height of the circumplanetary disc. Understanding the luminosity and radii of young giant planets will thus require detailed models of circumplanetary accretion.

astro-ph.EP↗

arXiv2015

Dynamical Stability of Imaged Planetary Systems in Formation: Application to HL Tau

A recent ALMA image revealed several concentric gaps in the protoplanetary disk surrounding the young star HL Tau. We consider the hypothesis that these gaps are carved by planets, and present a general framework for understanding the dynamical stability of such systems over typical disk lifetimes, providing estimates for the maximum planetary masses. We collect these easily evaluated constraints into a workflow that can help guide the design and interpretation of new observational campaigns and numerical simulations of gap opening in such systems. We argue that the locations of resonances should be significantly shifted in massive disks like HL Tau, and that theoretical uncertainties in the exact offset, together with observational errors, imply a large uncertainty in the dynamical state and stability in such disks. This presents an important barrier to using systems like HL Tau as a proxy for the initial conditions following planet formation. An important observational avenue to breaking this degeneracy is to search for eccentric gaps, which could implicate resonantly interacting planets. Unfortunately, massive disks like HL Tau should induce swift pericenter precession that would smear out any such eccentric features of planetary origin. This motivates pushing toward more typical, less massive disks. For a nominal non-resonant model of the HL Tau system with five planets, we find a maximum mass for the outer three bodies of approximately 2 Neptune masses. In a resonant configuration, these planets can reach at least the mass of Saturn. The inner two planets' masses are unconstrained by dynamical stability arguments.

astro-ph.EP↗

arXiv2015

Asynchronous rotation of Earth-mass planets in the habitable zone of lower-mass stars

Planets in the habitable zone of lower-mass stars are often assumed to be in a state of tidally synchronized rotation, which would considerably affect their putative habitability. Although thermal tides cause Venus to rotate retrogradely, simple scaling arguments tend to attribute this peculiarity to the massive Venusian atmosphere. Using a global climate model, we show that even a relatively thin atmosphere can drive terrestrial planets' rotation away from synchronicity. We derive a more realistic atmospheric tide model that predicts four asynchronous equilibrium spin states, two being stable, when the amplitude of the thermal tide exceeds a threshold that is met for habitable Earth-like planets with a 1-bar atmosphere around stars more massive than 0.5-0.7Msun. Thus, many recently discovered terrestrial planets could exhibit asynchronous spin-orbit rotation, even with a thin atmosphere.

astro-ph.EP↗

arXiv2014

Climate Stability of Habitable Earth-like Planets

The carbon-silicate cycle regulates the atmospheric $CO_2$ content of terrestrial planets on geological timescales through a balance between the rates of $CO_2$ volcanic outgassing and planetary intake from rock weathering. It is thought to act as an efficient climatic thermostat on Earth and, by extension, on other habitable planets. If, however, the weathering rate increases with the atmospheric $CO_2$ content, as expected on planets lacking land vascular plants, the carbon-silicate cycle feedback can become severely limited. Here we show that Earth-like planets receiving less sunlight than current Earth may no longer possess a stable warm climate but instead repeatedly cycle between unstable glaciated and deglaciated climatic states. This has implications for the search for life on exoplanets in the habitable zone of nearby stars.

astro-ph.EP↗

arXiv2013

Water-Trapped Worlds

Although tidally-locked habitable planets orbiting nearby M-dwarf stars are among the best astronomical targets to search for extrasolar life, they may also be deficient in volatiles and water. Climate models for this class of planets show atmospheric transport of water from the dayside to the nightside, where it is precipitated as snow and trapped as ice. Since ice only slowly flows back to the dayside upon accumulation, the resulting hydrological cycle can trap a large amount of water in the form of nightside ice. Using ice sheet dynamical and thermodynamical constraints, I illustrate how planets with less than about a quarter the Earth's oceans could trap most of their surface water on the nightside. This would leave their dayside, where habitable conditions are met, potentially dry. The amount and distribution of residual liquid water on the dayside depend on a variety of geophysical factors, including the efficiency of rock weathering at regulating atmospheric CO2 as dayside ocean basins dry-up. Water-trapped worlds with dry daysides may offer similar advantages as land planets for habitability, by contrast with worlds where more abundant water freely flows around the globe.

astro-ph.EP↗

arXiv2012

Thermo-Resistive Instability of Hot Planetary Atmospheres

The atmospheres of hot Jupiters and other strongly-forced exoplanets are susceptible to a thermal instability in the presence of ohmic dissipation, weak magnetic drag and strong winds. The instability occurs in radiatively-dominated atmospheric regions when the ohmic dissipation rate increases with temperature faster than the radiative (cooling) rate. The instability domain covers a specific range of atmospheric pressures and temperatures, typically P ~ 3-300 mbar and T ~ 1500-2500K for hot Jupiters, which makes it a candidate mechanism to explain the dayside thermal "inversions" inferred for a number of such exoplanets. The instability is suppressed by high levels of non-thermal photoionization, in possible agreement with a recently established observational trend. We highlight several shortcomings of the instability treatment presented here. Understanding the emergence and outcome of the instability, which should result in locally hotter atmospheres with stronger levels of drag, will require global non-linear atmospheric models with adequate MHD prescriptions.

astro-ph.EP↗