Graduating Student Colloquium featuring our Spring 2023 graduating class
Matthew Abruzzo, Brian Lu, Lucy Lu, Aaron Tran
A Unified Survival Criterion for Cloud-Wind Interactions
Cloud-wind interactions play an important role in long-lived multiphase flows in various galaxy-related contexts (e.g. galactic fountains, galactic winds, cool cosmic accretion, or multiphase tails of satellites in a halo and cluster). Clouds can survive the destructive effects of mixing when the ratio between the relevant mixing and cooling timescalesτmix/τcoolis large, via a mechanism we call TRML (turbulent radiative mixing layer) entrainment. Previously proposed survival criteria have been difficult to reconcile and disagree about the size of the smallest surviving cloud by factors of up to ∼100. We present a new criterion that estimates τmix and τcool with the shear timescale and the cooling timescale at intermediate temperatures, and validate it with hydrodynamic ENZO-E simulations of ∼1e4 K, pressure-confined clouds in hot supersonic winds. While all other criteria associate τmix with the cloud destruction timescale, our criterion is more physically consistent with the results of shear layer studies. It also makes accurate predictions about cloud survival consistent with criteria that associate τcool with cooling in the hot-phase. However, our choice of τcool better explains cloud survival when cooling is artificially turned off at those temperatures.
Cosmological constraints from HSC survey first-year data using deep learning
Deep learning is a promising way of extracting information from astronomical data. In this work, we use convolutional neural networks (CNNs) to constrain cosmological and astrophysical parameters with the HSC Y1 data. The pipeline carefully addresses the systematic effects like the baryons, intrinsic alignments, and photo-z errors. We find that the CNNs can yield tighter constraints than the power spectrum and peak counts: S₈=0.793(-0.018,+0.017) without considering baryons and S₈=0.819(-0.024,+0.034) when marginalizing over baryonic physics. With baryons, the S₈ discrepancy between HSC Y1 and Planck 2018 is reduced from ~2.2σ to ~0.4σ.
An abrupt change in the stellar spin-down law at the fully convective boundary
The importance of the existence of a radiative core in generating a solar-like magnetic dynamo is still unclear. Analytic models and magnetohydrodynamic simulations of stars suggest the thin layer between a star's radiative core and its convective zone can produce shearing that reproduces key characteristics of a solar-like dynamo. However, recent studies suggest fully and partially convective stars exhibit very similar period-activity relations, hinting that dynamos generated by stars with and without radiative cores hold similar properties. Here, using kinematic ages, we discover an abrupt change in the stellar spin-down law across the fully convective boundary. We found that fully convective stars exhibit a higher angular momentum loss rate, corresponding to a torque that is ~2.25 times higher for a given angular velocity than partially convective stars around the fully convective boundary. This requires a dipole field strength that is larger by a factor of ~2.5, a mass loss rate that is ~4.2 times larger, or some combination of both of those factors. Since stellar-wind torques depend primarily on large-scale magnetic fields and mass loss rates, both of which derive from magnetic activity, the observed abrupt change in spin-down law suggests that the dynamos of partially and fully convective stars may be fundamentally different.
Electron Energization in Solar Wind Shocks
The electron heating physics in collisionless shocks imprints upon micro-scale dissipation and waves seen in situ by heliospheric spacecraft, as well as X-ray emission from supernova remnants and galaxy clusters. How much do electrons heat, and how do they heat? We model low-beta (magnetic > thermal pressure) solar wind-like shocks with 1D and 2D fully-kinetic particle-in-cell simulations. Fast-mode / oblique-whistler waves accelerate electrons in bulk via proton-scale parallel electric fields; electrons' bulk kinetic energy then converts to heat via magnetic field-aligned electrostatic scattering. We measure the heating for a range of magnetic obliquities and Mach numbers, and we map out how the mechanism operates across different shock parameter regimes.
Followed by wine and cheese.
Host: Lorenzo Sironi