Planck Science Case
PLANCK Simulations Workflow (PSW) Graph (WS-PGRADE)
This application relates with Astronomy, Astrophysics and Cosmology. This is an evolution of the Y1 Planck workflow and gateway that implements also meta-workflows. The application implements the simulations of the Planck LFI mission. The computing and storage needs related to the simulations of the Planck mission implies the use of Grid DCIs; in particular we used the capability of the SSP to access gLite resources (Planck Virtual Organization). Moreover the application must be executed a large number of times varying the input parameters.
To explore different parameters a set of Planck simulations workflows are executed in parallel on different resources, thereafter the outputs are combined. The output data is produced in VOTable format and then combined into a single VOTable. The VOTable format is used in order to obtain IVOA compliant data. The meta-workflow implements native WS-PGRADE workflows and AstroTAVERNA ones as shown in Figure 7.
The Cosmic Microwave Background (CMB) preserves a picture of the Universe as it was about 380 000 years after the Big Bang, and can reveal the initial conditions for the evolution of the Universe. Planck's main objective was to measure the fluctuations of the CMB with an accuracy set by the fundamental astrophysical limits. The spacecraft charted the most accurate maps yet of the CMB.
The workflows are used to develop a web application of the Planck simulation software based on SSP. This web application allows Astronomers to execute a large number of simulations needed to study the cosmological parameters; it helps in the challenging task of identifying and correcting instrumental and observational systematics.
Planck Simulation meta-workflow
The Planck workflow consists of a very simple pipeline which is constituted of different software modules. The basic steps of the pipeline are described below:
- the CMB power spectrum is created with cmbfast starting from cosmological parameters;
- the CMB maps are built starting from the CMB power spectrum with synfast code being part of the HEALPix package;
- the CMB is combined with foregrounds with their own frequency dependent intensities and the final sky is convolved with the beam pattern for each of the detectors considered in the simulation;
- the map is contaminated by introducing instrumental noise which is computed and added to the "observed" sky signal, therefore the TOD (Time Ordered Data) is built.
The knowledge level increases over the time, hence new details are introduced and the whole computational chain is iterated many times, even during the operative phase of the mission.
In order to speed up calculations, we can assume a perfect overlapping between samples in two consecutive scan circles of the spacecraft when it remains in the same pointing position. In this way the sky signal is always the same for all the 60 scan circles corresponding to the same pointing position, so we can simulate it only once. We refer to this "fast" simulation procedure as "short" run; "long" runs instead correspond to complete simulation procedures where each scan circle is kept distinguished from the other ones.
The workflow produces as output a VOTable. The AstroTAVERNA component concatenates the various VOTables to produce a unique VOTable.
A simulation starts from a set of values associated to cosmological parameters. The simulation builds an ideal sky, contaminates it and extracts new maps; a new set of parameters is obtained starting from them. As shown above, different software components contribute to build the whole pipeline run; this typical modular structure fosters the reuse of single simulation modules to build new applications and workflows.
We produced more simulations by changing the value of the Dark Matter content, then we averaged over the produced maps. The results are compared with Planck observed data.
- Planck Collaboration, Ade, P. A. R., Aghanim, N., Arnaud, M., et al. Planck early results. I. the planck mission. A&A, 536:A1, 2011.
- G. Taffoni et al. Enabling grid technologies for planck space mission. Future Gener. Comput. Syst., 23(2):189–200, February 2007
- Planck Collaboration, Planck intermediate results. XVI. Profile likelihoods for cosmological parameters, 2014A&A...566A..54P
A simulation starts from a set of values associated to cosmological parameters. The simulation builds an ideal sky, contaminates it and extracts new maps; a new set of parameters is obtained starting from these maps.
As shown above, different software components contribute to build the whole pipeline run; this typical modular structure fosters the reuse of single simulation modules to build new applications and workflows.