Date: 28 de junio de 2022
Location: UNSAM y zoom (plase fill this form to get the link: https://forms.gle/6Dwth5UaGSywwMf58)
PROGRAM
Morning: INTEX/IAI/ICAS/ICIFI
10h: Juan Manuel Armaleo
Departamento de Física - UBA, Argentina
Title: Probing spin-2 Ultra-Light Dark Matter (ULDM) with pulsars and gravitational waves detectors
Abstract: The energy content of the Universe is made up of approximately 95% by dark components: in particular, ~25% corresponds to dark matter. Several models have emerged through the years trying to explain the nature of such matter; recently, the so-called "Ultra-Light Dark Matter" (ULDM) model - named like this given its extremely small mass - has gained great interest in the community. In this talk I will present this model, highlighting its main characteristics. In particular I will focus on the spin-2 ULDM model and how we can use pulsars and gravitational waves interferometers to constrain the model.
11:00h: coffee
11:30h: Lucas Bignone
Instituto de Astronomía y Física del Espacio, CONICET-UBA
Title: Métodos generativos para la clasificación no supervisada de imágenes astronómicas y satelitales
Abstract: Las redes neuronales están siendo utilizadas cada vez más frecuentemente para el análisis e interpretación de datos astronómicos, especialmente en tareas de clasificación. Los métodos denominados como no supervisados tienen la ventaja de no requerir de una clasificación humana previa, con lo cual resultan mucho más económicos y rápidos de aplicar. Además, son muy flexibles, ya que al no asumir una taxonomía a priori para la clasificación, se pueden adaptar fácilmente a distintos tipos de datos. En esta charla haré un breve repaso de la metodología, mostraré resultados aplicados a la clasificación morfológica de galaxias en el dominio óptico y cinemático. Finalmente, introduciré nuevas técnicas basadas en redes generativas adversarias (GANs) específicamente diseñadas para ser invariantes ante rotaciones y translaciones de la imagen y describiré la utilización de estas redes en un proyecto multidisciplinario para (1) la clasificación morfológica de galaxias y (2) la clasificación de nubes en la atmósfera terrestre, registradas en imágenes satelitales.
12:30h: Lunch
Afternoon: LaboCluster/ICAS
14hs: Elizabeth Gonzalez
Instituto de Astronomía Teórica y Experimental y OAC, UNC
Title: Aplicación de técnicas de apilamiento en el análisis de lentes débiles de sistemas de galaxias
Abstract: The gravitational lensing effect is a physical phenomenon that distorts the shapes of extended luminous sources located behind a matter distribution as a galaxy system, named as the lens. The introduced shape distortion can be quantified through the measured ellipticity of the sources, i.e. galaxies located behind the lens. These quantities encode information about the lens surface density distribution. One of the main sources of errors come from the fact that the observed background galaxies have their own ellipticity components besides those introduced by the lensing effect. This can be overcome by adopting a statistical methodology that combines the measured ellipticity of many sources, thus, assuming that they are randomly orientated. Then, the observed term that is related with the source's intrinsic ellipticity is averaged out and only the term related with the lensing effect is left. The precision of the lensing estimator would be related with the number of sources combined and can be optimised by adopting stacking techniques, which artificially increase the number density of sources. This procedure combines the observed signal of many lenses, thus allowing to obtain the information of the mean surface density distribution of the combined lenses. Besides the advantage of upgrading the lensing signal, it also reduces the intrinsic scatter introduced due to the differences in the halos formation histories by smoothing the density distribution and blurring the effect of substructure. This powerful technique allows to measure the total mass of the halos that host low-massive galaxy systems, for which the individual weak lensing signal is too weak to assess this estimate. The application of stacking techniques can also provide information about the mean elongation of combined halos if the main orientation of the surface mass is taken into account when combining the lenses. In this work we describe how lensing catalogues can be combined in order to measure the halo properties in which galaxy systems reside, such as the mean mass of low massive galaxy systems and the projected shapes of galaxy clusters. We revise the application of this technique using simulations as well as observational data.
15h: coffee and discussion