UC-IEG-V

From EGI Knowledge Base

Jump to: navigation, search

Use Case title: Detection and Extraction of compact objects in CMB maps

Short description: The Planck satellite and its experiments has been designed to measure the existence of anisotropies in the Cosmic Background Radiation. The structure of this radiation is a extremely relevant information for the astrophysics community interested in models describing the evolution on the universe at its early stages.

Over the past couple of years the group of astrophysics of the Institute of Physics of Cantabria has developed several methods based on wavelets and image filtering in order to detect compact objects in CMB maps as part of the working group “Compact Source Extraction” of the Planck mission.

This working group has generated realistic simulations of the sky for nine frequencies of the future Planck mission. The analysis that is being performed in int.eu.grid consists of three steps.

In the first step the code reads a map from the sphere and, as a function of certain input parameters (see Figure 5). As analysis tool we have installed in the Software runtime environment the Healpix libraries (http://healpix.jpl.nasa.gov/).

In the second step the map is divided in flat regions or patches:

Each region is filtered in order to increase the signal/noise ratio. See Figure

In the last step the compact objects on each patch are detected and catalogued. The process has to be repeated for nine maps, one for each Planck frequency. Every map is analyzed with several filtering techniques in order to find out which filter separates better the compact sources from the rest of radiation sources. The number of parameters that have to be analyzed is very large, including analysis of the patch sizes, overlap regions, etc…

In the next step of integration in int.eu.grid we plan to design a workflow tool as a plug-in for Migrating Desktop which performs all the steps of the calculation without the user having to take care manually of the information flow.

Actors involved:

Personal tools
hidden pages