Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userA closer look at the spur, blob, wiggle, and gaps in GD-1
58
0
0.0
(
0
)
Ask ChatGPT about the research
No Arabic abstract
The GD-1 stream is one of the longest and coldest stellar streams discovered to date, and one of the best objects for constraining the dark matter properties of the Milky Way. Using data from {it Gaia} DR2 we study the proper motions, distance, morphology and density of the stream to uncover small scale perturbations. The proper motion cleaned data shows a clear distance gradient across the stream, ranging from 7 to 12 kpc. However, unlike earlier studies that found a continuous gradient, we uncover a distance minimum at $varphi_{1}approx$-50 deg, after which the distance increases again. We can reliably trace the stream between -85$<varphi_{1}<$15 deg, showing an even further extent to GD-1 beyond the earlier extension of citet{Price-Whelan18a}. We constrain the stream track and density using a Boolean matched filter approach and find three large under densities and find significant residuals in the stream track lining up with these gaps. In particular, a gap is visible at $varphi_{1}$=-3 deg, surrounded by a clear sinusoidal wiggle. We argue that this wiggle is due to a perturbation since it has the wrong orientation to come from a progenitor. We compute a total initial stellar mass of the stream segment of 1.58$pm$0.07$times$10$^{4}$ M$_{odot}$. With the extended view of the spur in this work, we argue that the spur may be unrelated to the adjacent gap in the stream. Finally, we show that an interaction with the Sagittarius dwarf can create features similar to the spur.
rate research
Read More
Filaments in Herschel molecular cloud images are found to exhibit a characteristic width. This finding is in tension with spatial power spectra of the data, which show no indication of this characteristic scale. We demonstrate that this discrepancy is a result of the methodology adopted for measuring filament widths. First, we perform the previously used analysis technique on artificial scale-free data, and obtain a peaked width distribution of filament-like structures. Next, we repeat the analysis on three Herschel maps and reproduce the narrow distribution of widths found in previous studies $-$ when considering the average width of each filament. However, the distribution of widths measured at all points along a filament spine is broader than the distribution of mean filament widths, indicating that the narrow spread (interpreted as a characteristic width) results from averaging. Furthermore, the width is found to vary significantly from one end of a filament to the other. Therefore, the previously identified peak at 0.1 pc cannot be understood as representing the typical width of filaments. We find an alternative explanation by modelling the observed width distribution as a truncated power-law distribution, sampled with uncertainties. The position of the peak is connected to the lower truncation scale and is likely set by the choice of parameters used in measuring filament widths. We conclude that a characteristic width of filaments is not supported by the available data.
The radio-loud active galactic nucleus in M 87 hosts a powerful jet fueled by a super-massive black hole in its center. A bright feature 80 pc away from the M 87 core has been reported to show superluminal motions, and possibly to be connected with a TeV flare observed around 2005. To complement these studies and to understand the nature of this feature, we analyzed 2 cm VLBI data from 15 observing runs between 2000 and 2009. This feature is successfully detected at the milli-Jansky level from 2003 to 2007. Our detections show that its milli-arcsecond structure appears to be extended with a steep spectrum, and no compact or rapidly moving features are observed. Our results do not favor a blazar scenario for this feature.
Current methods for training robust networks lead to a drop in test accuracy, which has led prior works to posit that a robustness-accuracy tradeoff may be inevitable in deep learning. We take a closer look at this phenomenon and first show that real image datasets are actually separated. With this property in mind, we then prove that robustness and accuracy should both be achievable for benchmark datasets through locally Lipschitz functions, and hence, there should be no inherent tradeoff between robustness and accuracy. Through extensive experiments with robustness methods, we argue that the gap between theory and practice arises from two limitations of current methods: either they fail to impose local Lipschitzness or they are insufficiently generalized. We explore combining dropout with robust training methods and obtain better generalization. We conclude that achieving robustness and accuracy in practice may require using methods that impose local Lipschitzness and augmenting them with deep learning generalization techniques. Code available at https://github.com/yangarbiter/robust-local-lipschitz
We study how the behavior of deep policy gradient algorithms reflects the conceptual framework motivating their development. To this end, we propose a fine-grained analysis of state-of-the-art methods based on key elements of this framework: gradient estimation, value prediction, and optimization landscapes. Our results show that the behavior of deep policy gradient algorithms often deviates from what their motivating framework would predict: the surrogate objective does not match the true reward landscape, learned value estimators fail to fit the true value function, and gradient estimates poorly correlate with the true gradient. The mismatch between predicted and empirical behavior we uncover highlights our poor understanding of current methods, and indicates the need to move beyond current benchmark-centric evaluation methods.
We investigate the bursty star formation histories (SFHs) of dwarf galaxies using the distribution of log($L_{Halpha}/L_{UV}$) of 185 local galaxies. We expand on the work of Weisz et al. 2012 to consider a wider range of SFHs and stellar metallicities, and show that there are large degeneracies in a periodic, top-hat burst model. We argue that all galaxies of a given mass have similar SFHs and we can therefore include the $L_{Halpha}$ distributions (subtracting the median trend with stellar mass, referred to as $Delta text{log}(L_{Halpha})$) in our analyses. $Delta text{log}(L_{Halpha})$ traces the amplitude of the bursts, and log($L_{Halpha}/L_{UV}$) is a function of timescale, amplitude, and shape of the bursts. We examine the 2-dimensional distribution of these two indicators constrain the SFHs. We use exponentially rising/falling bursts to determine timescales ($e$-folding time, $tau$). We find that galaxies below $10^{7.5}$ M$_{odot}$ undergo large (amplitudes of $sim 100$) and rapid ($tau < 30$ Myr) bursts, while galaxies above $10^{8.5}$ M$_{odot}$ experience smaller (maximum amplitudes $sim 10$), slower ($tau gtrsim 300$ Myr) bursts. We compare to the FIRE-2 hydrodynamical simulations and find that the burst amplitudes agree with observations, but they are too rapid in more massive galaxies ($M_* > 10^8$ M$_{odot}$). Finally, we confirm that stochastic sampling of the stellar mass function can not reproduce the observed distributions unless the standard assumptions of cluster and stellar mass functions are changed. With the next generation of telescopes, measurements of $L_{UV}$ and $L_{Halpha}$ will become available for dwarf galaxies at high-redshift, enabling similar analyses of galaxies in the early universe.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
