Jorit Schmelzle, Aurelien Lucchi, Tomasz Kacprzak, Adam Amara, Raphael Sgier, Alexandre Réfrégier, Thomas Hofmann

We demonstrate the potential of Deep Learning methods for measurements of cosmological parameters from density fields, focusing on the extraction of non-Gaussian information. We consider weak lensing mass maps as our dataset. We aim for our method to be able to distinguish between five models, which were chosen to lie along the $\sigma _8$ - $\Omega _m$ degeneracy, and have nearly the same two-point statistics. We design and implement a Deep Convolutional Neural Network (DCNN) which learns the relation between five cosmological models and the mass maps they generate. We develop a new training strategy which ensures the good performance of the network for high levels of noise. We compare the performance of this approach to commonly used non-Gaussian statistics, namely the skewness and kurtosis of the convergence maps. We find that our implementation of DCNN outperforms the skewness and kurtosis statistics, especially for high noise levels. The network maintains the mean discrimination efficiency greater than $85\%$ even for noise levels corresponding to ground based lensing observations, while the other statistics perform worse in this setting, achieving efficiency less than $70\%$. This demonstrates the ability of CNN-based methods to efficiently break the $\sigma _8$ - $\Omega _m$ degeneracy with weak lensing mass maps alone. We discuss the potential of this method to be applied to the analysis of real weak lensing data and other datasets.

Ana Diaz Rivero, Francis-Yan Cyr-Racine, Cora Dvorkin

Studying the smallest self-bound dark matter structure in our Universe can yield important clues about the fundamental particle nature of dark matter. Galaxy-scale strong gravitational lensing provides a unique way to detect and characterize dark matter substructures at cosmological distances from the Milky Way. Within the cold dark matter (CDM) paradigm, the number of low-mass subhalos within lens galaxies is expected to be large, implying that their contribution to the lensing convergence field is approximately Gaussian and could thus be described by their power spectrum. We develop here a general formalism to compute from first principles the substructure convergence power spectrum for different populations of dark matter subhalos. As an example, we apply our framework to two distinct subhalo populations: a truncated Navarro-Frenk-White subhalo population motivated by standard CDM, and a truncated cored subhalo population motivated by self-interacting dark matter (SIDM). We study in detail how the subhalo abundance, mass function, internal density profile, and concentration affect the amplitude and shape of substructure power spectrum. We determine that the power spectrum is mostly sensitive to a specific combination of the subhalo abundance and moments of the mass function, as well as to the average tidal truncation scale of the largest subhalos included in the analysis. Interestingly, we show that the asymptotic slope of the substructure power spectrum at large wavenumber reflects the internal density profile of the subhalos. In particular, the SIDM power spectrum exhibits a characteristic steepening at large wavenumber absent in the CDM power spectrum, opening the possibility of using this observable, if at all measurable, to discern between these two scenarios.

lxj154

https://arxiv.org/abs/1707.01739

James S. Bullock, Michael Boylan-Kolchin

The dark energy plus cold dark matter ($\Lambda$CDM) cosmological model has been a demonstrably successful framework for predicting and explaining the large-scale structure of Universe and its evolution with time. Yet on length scales smaller than $\sim 1$ Mpc and mass scales smaller than $\sim 10^{11} M_{\odot}$, the theory faces a number of challenges. For example, the observed cores of many dark-matter dominated galaxies are both less dense and less cuspy than naively predicted in $\Lambda$CDM. The number of small galaxies and dwarf satellites in the Local Group is also far below the predicted count of low-mass dark matter halos and subhalos within similar volumes. These issues underlie the most well-documented problems with $\Lambda$CDM: Cusp/Core, Missing Satellites, and Too-Big-to-Fail. The key question is whether a better understanding of baryon physics, dark matter physics, or both will be required to meet these challenges. Other anomalies, including the observed planar and orbital configurations of Local Group satellites and the tight baryonic/dark matter scaling relations obeyed by the galaxy population, have been less thoroughly explored in the context of $\Lambda$CDM theory. Future surveys to discover faint, distant dwarf galaxies and to precisely measure their masses and density structure hold promising avenues for testing possible solutions to the small-scale challenges going forward. Observational programs to constrain or discover and characterize the number of truly dark low-mass halos are among the most important, and achievable, goals in this field over then next decade. These efforts will either further verify the $\Lambda$CDM paradigm or demand a substantial revision in our understanding of the nature of dark matter.

Mahdi Torabian

In this note we propose a topological action for a Poincare times diffeomorphism invariant gauge theory. We show that there is Higgs phase where the gauge symmetry is spontaneous broken to a diagonal Lorentz subgroup and gives the Einstein-Hilbert action plus the dRGT potential terms. In this vacuum, there are five (three from Goldstone modes) propagating degrees of freedom which form polarizations of a massive spin 2 particle, an extra healthy heavy scalar (Higgs) mode and no Boulware-Deser ghost mode. We further show that the action can be derived in a limit from a topological de Sitter invariant gauge theory in 4 dimensions.

Ariel Zhitnitsky

We advocate the idea that the surprising emission of extreme ultra violet (EUV) radiation and soft x-rays from the Sun are powered externally by incident dark matter (DM) particles. The energy and the spectral shape of this otherwise unexpected solar irradiation is estimated within the quark nugget dark matter model. This model was originally invented as a natural explanation of the observed ratio $\Omega_{\rm dark} \sim \Omega_{\rm visible}$ when the DM and visible matter densities assume the same order of magnitude values. This generic consequence of the model is a result of the common origin of both types of matter which are formed during the same QCD transition and both proportional to the same fundamental dimensional parameter $\Lambda_{\rm QCD}$. We also present arguments suggesting that the transient brightening-like "nano-flares" in the Sun may be related to the annihilation events which inevitably occur in the solar atmosphere within this dark matter scenario.

Gregory Gabadadze

Pure massive gravity is strongly coupled at a certain low scale, known as Lambda_3. I show that the theory can be embedded into another one, with new light degrees of freedom, to increase the strong scale to a significantly larger value. Certain universal aspects of the proposed mechanism are discussed, notably that the coupling of the longitudinal mode to a stress-tensor is suppressed, thus making the linear theory consistent with the fifth-force exclusion. An example of the embedding theory studied in detail is 5D AdS massive gravity, with a large cosmological constant. In this example the 4D strong scale can be increased by 19 orders of magnitude. Holographic duality then suggests that the strong scale of the 4D massive gravity can be increased by coupling it to a 4D non-local CFT, endowed with a UV cutoff; however, the 5D classical gravity picture appears to be more tractable.

Eric Cotner, Alexander Kusenko

Scalar condensates with large expectation values can form in the early universe, for example, in theories with supersymmetry. The condensate can undergo fragmentation into Q-balls before decaying. If the Q-balls dominate the energy density for some period of time, statistical fluctuations in their number density can lead to formation of primordial black holes (PBH). In the case of supersymmetry the mass range is limited from above by $10^{23}$g. For a general charged scalar field, this robust mechanism can generate black holes over a much broader mass range, including the black holes with masses of 1-100 solar masses, which is relevant for LIGO observations of gravitational waves. Topological defects can lead to formation of PBH in a similar fashion.