Transcript PDC

PLATO Data Center:
Purpose and Structure
Laurent Gizon (PDPM)
Hamed Moradi (PDC Project Office)
The PDC is in charge of the
validation, calibration, and
analysis of the PLATO
observations.
It delivers the final PLATO
science Data Products.
SGS Structure
•
•
•
•
Mission Operations Center (MOC, flight-critical)
Science Operations Center (SOC, mission-critical)
PLATO Data Center (PDC, science-critical)
Science Preparatory Activities (scientific
specification of software)
PLATO Ground
Segment
Operations
Ground
Segment
Mission
Operations
Centre
Science
Ground
Segment
Science
Operations
Centre
Ground Station
Network
SOC
PLATO Data
Center
PDC
PLATO
Science
Preparation
Management
PSPM
• The definition phase objectives of the SGS are to
establish the technical (PDC) and scientific (PSPM)
requirements baseline for the SGS and to develop the
operations concept, architecture and interfaces.
• The definition phase activities of the PDC and PSPM are
organized according to the following guidelines:
– PSPM provides the scientific specifications of the software
– PDC translates the scientific specifications into technical
specifications
– PDC implements the technical specifications
– PSPM checks that the PDC software is consistent with the initial
scientific specifications. This validation by PSPM occurs within
the PDC – a normal part of the development QA process.
Data Levels
• Telemetry: Baseline: 109 Gb/day uncompressed (8.7 Mb/s
compressed, during 3.5 hr each day).
• Level 0: Depacketized light curves, centroid curves, and
selected imagettes (~1600), for each telescope (32+2)
• Level 1: Analysis of imagettes to validate and optimize
performance of on-board treatment. Implementation of onground instrumental corrections, such as CCD corrections
and jitter corrections. Then computation of average light
curves and centroid curves for each star (science-ready).
• Level 2: PLATO science Data Products (next table). Final
DP is a list of confirmed planetary systems, fully
characterized by the transit curves, the stellar seismic
parameters, and the follow-up observations. High scientific
added value.
PLATO Data Products
Calibrated light curves and centroid
curves
DP1
L1
Planetary transits and their parameters
DP2
L2
Asteroseismic mode parameters
DP3
L2
Stellar rotation and activity
DP4
L2
Stellar masses and ages
DP5
L2
Confirmed planetary systems and their
characteristics
DP6
L2
Ancillary Observations
• Essential information for the success of the mission: input catalog,
follow-up observations, etc.
• Support for on-board processing, on-ground calibration, and
scientific data analysis
• Stellar properties: effective temperature, absolute luminosity, radius
[Gaia], chemical abundances, v sin i, activity, properties specific to
multiple stars.
• Follow-up observations to confirm planets (at several wavelengths
when possible)
• Other relevant complementary observations: hires spectra,
astrometry, imaging, spectro-polarimetry, etc..
• The ancillary data are in support of the processing activities and are
accessed by the PDC via the main data base.
Ancillary observations:
star catalogs
stellar parameters
follow-up observations
DP0
spectroscopy
radial velocities
interferometry
astrometry (Gaia)
DP1
DP3
DP5
DP4
DP2
DP6
Exoplanet
Analysis System
PDC Architecture
• Main Data Base centralizes DPs and
ancillary data (PDC-DB @ MPS)
• PDC develops and implements code at
ESA SOC to produce DP1
• Data Processing Centers (PDPCs)
manufacture DP2-DP6
• Distribution of DPs and long term archive
under SOC responsibility
PLC
Science Team+PDPM
science activities
PLATO Instrument
Manager
MOC
SOC
data access
coordination
PDC
L. Gizon, MPS
main database &
system architecture
R. Burston, MPS
data treatment
Algorithms
R. Samadi, LESIA
data treatment
implementation
I. Pardowitz, MPS
Exoplanet
Analysis System
N. Walton, IoA
SOC includes a processing center for the validation and calibration of the data
ESA overall coordination (oversight) of science data releases, data access
Stellar
ancillary database
and distribution
Analysis System
R. Burston, MPS
T. Appourchaux, IAS
PDC designs and implements software to be run at the SOC
PLATO Instrument
Manager (P. Bodin)
PDC WBS
300 000
PDC Overall
Management
MPSSR
L. Gizon (PDPM)
310 000
System Architecture
and Management
320 000
Data Processing
Algorithms
LESIA
R. Samadi
MPSSR
R. Burston
310 001
PDC-DB
1 Central Data Base
330 000
Data Processing
Development
MPSSR
I. Pardowitz
340 000
340 001
Input Catalogue
ASDC-ASI
360 001
PDPC-C
370 001
Stellar Analysis
System
PDPC-I
T. Appourchaux
380 000
Data Analysis
Support Tools
MPSSR
PDPC-L
N. Walton
370 000
IAS
350 001
M. Deleuil
360 000
Exoplanet Analysis
System
IoA-Cam
PDPC-A
P. Giommi
350 000
Ancillary Data
Management
LAM
5 Data Processing Centers
L. Gizon
380 001
PDPC-M
WP32 Data Processing Algorithms
(Talk by Samadi)
WP35 Ancillary Data Management
(Talk by Deleuil)
WP36 Exoplanet Analysis System
(Talk by Walton)
WP37 Stellar Analysis System
(Talk by Appourchaux)
WP31 System architecture
and main database (Burston)
• System architecture, archives, data base,
system management
• Data flow design and management, export
system, network
• Simulation of data stream
WP33 Data Processing
Development (Pardowitz)
• Write and implement core-processing
software that will run at the SOC
• Requires a good understanding of system
interfaces with SOC and operational
procedures
• For phase A, study jitter correction to
prove feasibility
WP34 Input Catalog (Giommi)
• Implementation of the PLATO input
catalogue, under Italian responsibility.
• This activity is related to the target and
field characterization activity in the PSPM
segment of the PMC.
• WP34 delivers the validated PIC to the
PDC-DB
WP38 Data Analysis
Support Tools (Gizon)
• PDC documentation management
• Tools to support the analysis of individual light
curves and to provide feedback to L2 processing
pipelines (exoplanet and stellar).
• PDPC-M is the place were consortium scientists
inspect light curves, assess DP validity and
update ranking of planet candidates
• Search tools and VO activities
• Internal PDC web site
• In particular, PDC web site makes FU info
accessible to FU observers
Time table
• Regular meetings with ESA to specify interfaces PDC-SOC
• End Feb 2011  5th PDC Meeting in K-Lindau. Identify final
problems. Invite PSPM Leaders.
• WP Leaders deliver reports to LG by March 2011
• PDC document delivered to PCL in May 2011
• June 2011: Decision on PLATO selection
• End June 2011, Phase B1 meeting.
• December 2011: End phase B1
•
•
•
•
November 2018: Launch of PLATO
3+2+1 years in space
Several releases of DPs during and after space mission
PDC must remain operational up to ~3 yrs after the end of
the space mission in order to confirm last planets.
Cost
Thank you
Data validation
–
Validate onboard software:
•
•
•
–
Validate onboard setup:
•
–
Check onboard processing using ground copy of
onboard software and the imagettes of ~1600 stars
Validate distortion matrix model, 2D sky background
model, PSF model fits
Validate computation of masks and windows
Fine tuning of onboard software algorithm. For
example choose number of parameters needed to
describe PSF. Especially during configuration mode.
Monitor health of each telescope and assess
quality of the data
Data corrections
–
–
–
–
–
–
–
Correction for jitter. Performed independently for each
telescope; requires PSF knowledge, stellar catalog, and
distortion matrix.
Integration time correction, sampling time correction
Statistical analysis over the 40 telescopes to identify
cosmic ray hits, hot pixels, and possibly deficient
telescopes
Average light curves and centroid curves over all
telescopes (weighted average).
Compute error based on scatter
The ~1600 stars for which imagettes are available receive
a more sophisticated treatment. PSF fits to improve
photometry (contamination from neighboring sources
taken into acount). Imagettes are downloaded for all
stars for which a serious planetary candidate has been
identified.
Long term detrending probably moved to PDC
Data volumes
• Telemetry rate: 109 Gb/day uncompressed
• Over a 6 yr mission: 30 TB uncompressed
• The volume of archived L0, L1 and HK data is
expected to be 10-50 times this amount (reformatting
and calibration history), i.e. 300-1500 TB
• The volume of the science data products is likely to
be negligible in comparison (although the complexity
of the data may be high).
• Ancillary data base: basic stellar observations and
parameters, spectra, Gaia specific obs, etc. How big?
The overall data volume should not exceed a few PB,
which is not problematic.