ID VST-SPE-OCM-23100-3050

NOVA - Kapteyn Institute OmegaCAM

Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 1

OmegaCAM

OmegaCAM Data Flow System User Requirements

Including formal list of Calibration Requirements

Issue: VERSION 1.1

Date: 29 Oct 2001
Prepared by: Valentijn, Begeman, Boxhoorn, Deul, Kuijken, Rengelink
Purpose of printout: FDR

Approved by:
Konrad Kuijken, OmegaCAM PI.

Edwin Valentijn, NOVA programme manager.

www.astro.rug.nl/∼omegacam

This document is prepared by the Odoco Document Control System.

 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 2

CHANGE RECORD

Changes between Version 1.0–FINAL DESIGN REVIEW and Version 1.1–FINAL

DESIGN REVIEW
Additions
New section on 2-3 day cycle specifications – 2.2 DFS requirements.
Specify all three levels of TSF’s, as given in VST-SPE-OCM-23100-3064,
in requirements.
Detailed estimates of observing times for photometric checks – 2.2 DFS
requirements, req.562, req.563, req.564.
Include tilt determination – req. 571 Camera focus/tilt.
CalFile– 562S Sky brightness – req.562.
Reference to req.563 in CA – req.564.
OmegaCAM DID – 1.2 Applicable documents.
Once/year dark dome test – req.533.
Processing of calibration data follows telescope schedule – 2.2 DFS re-
quirements, 6.1 Data reduction software requirements.
Updates
Clarify fast recipe for Technical Specifications conformance – req.562,
req.563.
Erroneous references to darkcurrent check for req. 547 Quick detector
responsivity check removed.
Nonexistent CalFile– 561 removed – req.533.
Stars have to be observed during the night – req.525.
req. 571 Camera focus/tilt is no longer a workhorse/doit – 1.4 Abbrevi-
ations and Acronyms, 5.10 On site quick look analysis.
Exposure times TBC during commissioning – req.561.
More accurate description of the algorithm – req.523.
Lamp procedure TBC. – 5.4 Detectors operational specific calibrations.
Removed reference to QCO – seq.– 631.
Mention acceptance of multi-extension FITS files – seq.– 631.
Target-related template parameters (only) where applicable – 4.4 Observ-
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 3

ing Templates.
Reworded sentence about DFS-pipeline – 5.10 On site quick look analysis.
Reworded sentence about modules – 6.1 Data reduction software require-
ments.
Rotator offset angle – 4.4 Observing Templates.
Use plots for analysis – req.571.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 4

TABLE of CONTENTS
1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.1 Scope of this document . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2 Applicable documents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.3 Reference documents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.4 Abbreviations and Acronyms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2 SCIENTIFIC REQUIREMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.1 Scientific requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2 DFS requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3 INSTRUMENT CONCEPT- Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.1 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.2 Focal plane geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.3 CCD details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.4 Filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.5 Control electronics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.6 Instrument Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4 OBSERVING MODES and STRATEGIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.1 Observing Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
4.2 Observing Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.3 Filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.4 Observing Templates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.5 Field correctors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
5 BASELINE CALIBRATION REQUIREMENTS . . . . . . . . . . . . . . . . . . . . . . . . . 41
5.0 Documentation system, Odoco . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
5.1 Functional Checks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
5.2 Detector Electronics specific Calibrations . . . . . . . . . . . . . . . . . . . . . . . . . . 48
5.2.1 Req.– CCD read noise - doit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
5.2.2 Req.– Hot pixels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5.2.3 Req.– CCD gain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
5.2.4 Req.– Electromagnetic Compatibility . . . . . . . . . . . . . . . . . . . . . . . . 50
5.2.5 Req.– Electrical cross talk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.3 Detectors specific calibrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
5.3.1 Req.– CCD Dark Current - doit . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 5

5.3.2 Req.– CCD Particle Event Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

5.3.3 Req.– CCD Linearity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
5.3.4 Req.– CCD Charge Transfer Efficiency . . . . . . . . . . . . . . . . . . . . . 53
5.3.5 Req.– CCD Cold Pixels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
5.3.6 Req.– CCD Hysteresis, strong signal . . . . . . . . . . . . . . . . . . . . . . . . 55
5.4 Detectors operational specific calibrations . . . . . . . . . . . . . . . . . . . . . . . . . 55
5.4.1 Req.– Bias - doit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
5.4.2 Req.– Flat-field - dome key bands + user bands - doit . . . . . . . 61
5.4.3 Req.– Flat-field - twilight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.4.4 Req.– Flat-field - night sky . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.4.5 Req.– Flat-field - Fringing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
5.4.6 Req.– Flat-field - master flat and weight map . . . . . . . . . . . . . . . 65
5.4.7 Req.– Quick detector responsivity check - doit . . . . . . . . . . . . . . . 66
5.4.8 Req.– Illumination correction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.5 Astrometric Calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.5.1 Req.– Position of Camera in focal plane . . . . . . . . . . . . . . . . . . . . 69
5.5.2 Req.– Telescope Pointing and offsetting . . . . . . . . . . . . . . . . . . . . . 70
5.5.3 Req.– Telescope and Field Rotator tracking . . . . . . . . . . . . . . . . . 70
5.5.4 Req.– PSF Anisotropy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
5.5.5 Req.– The astrometric solution for templates - doit -see 6.3.4 72
5.5.6 Req.– The astrometric solution for Guide CCD’s . . . . . . . . . . . 72
5.6 Photometric Calibration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
5.6.1 Req.– Shutter Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.6.2 Req.– Photometric Calibration - monitoring . . . . . . . . . . . . . . . 78
5.6.3 Req.– Photometric Calibration - zeropoint keybands - doit . . 80
5.6.4 Req.– Photometric Calibration - zeropoint user bands . . . . . . . 81
5.6.5 Req.– Filter band passes - user bands vs key bands . . . . . . . . . . 82
5.6.6 Req.– Dependency on rotator angle/reproducibility . . . . . . . . . . 83
5.6.7 Req.– Linearity (as a function of flux) . . . . . . . . . . . . . . . . . . . . . . 83
5.6.8 Req.– Detection limit and ETC calibrations . . . . . . . . . . . . . . . . . 83
5.6.9 Req.– Secondary Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5.7 Internal alignments, optics etc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.7.1 Req.– Camera focus/tilt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.7.2 Req.– Ghosts - ADC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 6

5.8 Effect of Telescope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

5.9 Workhorses and End to end tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
5.10 On-site quick look analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Fig 6. Data Flow of Science and Calibration data . . . . . . . . . . . . . . . . . . . . . 89
6 DATA REDUCTION SOFTWARE SPECIFICS . . . . . . . . . . . . . . . . . . . . . . . . . 90
6.1 Data reduction software requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
6.2 Estimate of data volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
6.3 Baseline Requirements for the Image Pipeline . . . . . . . . . . . . . . . . . . . . 101
A1 LIST of CALIBRATION REQUIREMENTS . . . . . . . . . . . . . . . . . . . . . . . . . 105
A2 LIST of RAW CALIBRATION DATA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
A3 LIST of DFS I/O CALIBRATION FILES . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
A4 LIST of DFS INPUT REFERENCE CATALOGUES . . . . . . . . . . . . . . . . . . 110
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 7

Lay-out of the sections

URD OmegaCAM

1.0 Introduction
2.0 Scientific Requirements
3.0 Instrument Concepts
4.0 Observing Strategies
5.0 Calibration requirements
6.0 Data reduction
7.0 Preparatory Programmes
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 8

1 INTRODUCTION

OmegaCAM is planned as the first instrument for the VLT Survey Telescope
(VST) on ESO’s Paranal site. It is expected to operate for a period of ten years,
and at least during the first 3-5 years of operations of the VST the OmegaCAM
is foreseen to be the only instrument on this telescope. OmegaCAM is a 16,384
x 16,384 pixel (16k x 16k) imaging camera which will image a field of 1 square
degree of sky.
The instrument is envisaged to execute dedicated observing programmes defined
by individual users or teams. About 2/3 of the available observing time will be
allocated by ESO’s OPC. The remaining time is labeled as guaranteed time for
the consortia involved in the construction of the telescope and the camera. Both
small dedicated programmes, and bulk wide field sky surveys, are expected.
The VST and OmegaCAM are built to provide an observing facility for the
purpose of selecting targets for follow-up observations at the VLT, but also
to conduct stand-alone observing programmes that require wide-field imaging.
The camera and its associated data reduction will facilitate accurate photom-
etry and astrometry over its entire field of view, following the requirements on
the VST and its instrumentation. Primary Performance Characteristics for the
VST wide-angle CCD camera are laid out in the Memorandum of Understand-
ing (MoU) between OmegaCAM and ESO in Section A.4.1, while guidelines for
its implementation in ESO’s Data Flow System (DFS) are given in A.4.2 of the
MoU which in turn refers to VLT-SPE-ESO-19000-1618.

1.1 Scope of this document

The present document provides the deliverable ’Data Flow System User Re-
quirements’ as proposed in VLT-SPE-ESO-19000-1618: ”it defines the user re-
quirements including operational scenarios which must be supported and there-
fore may impact definition of observing modes, Exposure Time calculator (ETC)
or pipeline procedures”.
The OmegaCAM User requirements related to the Instrument Software are spec-
ified in a separate document [VST-SPE-OCM-23100-3060].
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 9

Standardization of observing modes, calibration procedures and related data

reduction procedures play an import role in the OmegaCAM instrument con-
cept and the present document includes the specification of the standardized
observing scenarios for science observations (templates) and the basic require-
ments for the qualification, quality control and calibration of the instrument
both during switch-on, commissioning phase (CP) and routine phase (RP).
Implementations of these requirements shall be further specified in the Cali-
bration Plan document . Details of the related data reduction are given in
the Data Reduction Specification document. Instantiations of the Calibra-
tion Plan will form the Schedule of the telescope, which for the commissioning
phase shall be written by the Instrument Consortium and will form a part of the
Commissioning plan. During routine operations the schedule shall be written
by ESO.
Figure 1 sketches that part of the organization of the OmegaCAM consortium
which relates to the deliverables of data reduction and calibration issues. The
sketch is not exhaustive, but is meant to highlight the basic relation between
National tasks (mostly NL as far as data reduction is concerned, with prod-
ucts/results marked in blue) and tasks relating to deliverables to ESO (on the
right side, with deliverables marked in red).
For some Dutch National tasks collaborations are foreseen with other European
data centers for wide field imaging, such as Terapix (France), the Astronomical
Observatories of Capodimonte and Padova, and the Sternwarte München.
The deliverables to ESO are driven by the Memorandum of Understanding be-
tween the Consortium and ESO, which in turn refers to ’Data Flow System User
Requirements’ VLT-SPE-ESO-19000-1618 (DFS-req-16 in the sketch), which
in turn specifies the hierarchy of deliverables: User Requirement Document
(URD) and Calibration Plan document and various S/W deliverables.
The consortium has implemented a documentation control system with a strict
documentation format for the definition of calibration req.’s, and named this
Odoco (OmegaCAM document control). This system is explained in detail in
Section 5.0. The green encircled region in the figure indicates the various
documents and files which will be guarded by Odoco.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 10

OmegaCAM data reduction and calibration deliverables scheme

National Tasks ESO Tasks

NOVA-ISC MoU
PI

Odoco DFS-req -16

Community
Input
PDR/FDR
cal.req’s URD
Workpackages
preparatory PDR/FDR
programmes
CP

recipes modules manuals

pipeline pipeline
hardware software

expertise calibration operations

support system pipeline DMD
telescope

source lists

Figure 1
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 11

1.2 Applicable documents

The following documents, of the exact issue shown, form a part of this doc-
ument to the extent specified herein. In the event of conflict between the
documents referenced herein and the contents of this document, the contents
of this document shall be considered as a superseding requirement.
[] VST-PLA-OCM-23100-3090 OmegaCAM Calibration Plan
[] VST-PLA-OCM-23100-3051 OmegaCAM Data Reduction Specifications
[] VST-PLA-OCM-23100-3100 OmegaCAM Commissioning Plan
[] VST-PLA-OCM-23100-3010 Project Management Plan and Project Plan
and Schedule
[] VST-PLA-OCM-23100-3020 Product Assurance Plan
[] VST-SPE-OCM-23100-3551 ICD OmegaCAM - ESO-DMD
[] VST-PLA-OCM-23100-3080 Technical Operations and Maintenance Plan
S/W deliverables depending on the URD
[] OmegaCAM Data Interface Dictionary
[] OmegaCAM Template Signature Files
[] OmegaCAM Exposure Time Calculator

1.3 Reference documents

The following documents are referenced in this document.
[] VLT-SPE-ESO-19000-1618 1.0 21/04/1999 — VLT Data flow for the VLT
instruments Requirement Specification
[] MoU OmegaCAM - ESO
[] VST-PLA-OCM-23100-3030 Safety Analysis and Compliance Assessment
[] VST-TRE-OCM-23100-3040 Design Analysis, Performance Report
[] VST-TRE-OCM-23100-3041 Design Analysis, Performance Report: Mechan-
ical
[] VST-ESO-OCM-23100-3042 Design Analysis, Performance Report: Detec-
tor System
[] VST-TRE-OCM-23100-3043 Design Analysis, Performance Report: Electronic
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 12

[] VST-SPE-OCM-23100-3050 Data Flow System User Requirements

[] VST-SPE-OCM-23100-3060 Instrument Software
[] VST-SPE-OCM-23100-3064 OmegaCAM Observation Software Design De-
scription
[] VST-PLA-OCM-23100-3070 MAIT and Alignment Plan
[] VLT-MAN-ESO-19000-2050 FTU FITS Translation Utility User Manual
[] SExtractor v2.1.3 User’s guide
[] Eclipse Developer’s Guide
[] Eclipse User’s Guide
[] The LDACTools Library v1.2 User’s guide
[] Pipeline Documentation Ver. 1.3
[] Swarp v1.20 User’s guide

1.4 Abbreviations and Acronyms

Abbreviations and Acronyms used in this document
A/D Analog/digital
ACS-dbase Astronomical Calibration Source database
ADC Atmospheric Dispersion Corrector
ADU Analog to Digital Unit
AGN Active Galactic Nucleus
BRD Baseline Requirements Document
CA Calibration Analysis
CAP Calibration Analysis Procedure
CCD Charge Coupled Device
CO Calibration Observation
CP Commissioning Phase
CalP Calibration Plan
CTE Charge Transfer Efficiency
CVS Code Version System
DFS Data Flow System
ESO European Southern Observatory
ETC Exposure Time Calculator
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 13

FOV Field of View

FWHM Full Width at Half Maximum
GRB Gamma Ray Burst
GT Garanteed Time
HZSS High Redshift Supernova Search
ICS Instrument Control Software
IWS Instrument Workstation
ISO Infrared Space Observatory
IST Instrument Science Team
KBO Kuiper Belt Object
MoU Memorandum of Understanding
NOVA (Dutch) Nederlandse Onderzoekschool Voor Astronomie
OaPd Astronomical Observatory of Padua
OB Observation block
OD Observation Description
OPC Observing Programme Committee
OT Optical Transient
OT Observing Template
PPP Photometry Preparatory Programme
PSF Point Spread Function
QC0 Quality Control zero
QC1 Quality Control one
QSO Quasi-Stellar Object
RPE Relative Pointing Error
RP Routine Phase
RSRF Relative Spectral Response Function
SCP Supernova Cosmology Project
SED Spectral Energy Distribution
SSO Solar System Object
S/N Signal/Noise
S/W Software
TBC To Be Confirmed
TBD To Be Defined
TBW To Be Written
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 14

TCS Telescope Control System

TP Target Package
TSF Template Signature File
URD User Requirement Document
USM Universitäts Sternwarte München
VLT Very Large Telescope
VST VLT Survey Telescope
WFI@2.2m Wide Field Imager at the ESO 2.2m telescope
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 15

2 SCIENTIFIC REQUIREMENTS

2.1 Scientific requirements

The scientific case for the construction of a wide field imaging facility to be
dedicated to “supply complete databases for the VLT science” was presented
by Astronomical Observatory of Capodimonte in 1997. After consideration of
specific research projects, of the specific needs for VLT target selection and a
survey of existing and planned wide field imaging facilities worldwide, a number
of scientific requirements were extracted which drove the design of the telescope
and also provided the references for the design of the associated camera.
OmegaCAM and the VST are entirely dedicated to wide field optical imaging,
and the design must comply with the scientific requirements derived in the VST
science case and the MoU which are repeated below (items 1–11). Further
scientific requirements (items 12–19) have been derived from considerations of
specific scientific use cases, described later in this section.
These scientific requirements propagate as requirements on Instrument S/W
[VST-SPE-OCM-23100-3061], a set of Baseline Calibration requirements (Sec-
tion 5) and requirements on the Data reduction (Section 6).

List of science driven requirements (Sc. Req’s):

1 - A telescope diameter of 2.6m
2 - 1 deg2 field of view
3 - the highest efficiency achievable
4 - CCD filling factor as close as possible to 100%
5 - good seeing sampling of at least 2 × 2 pixels within seeing element (good
match is 15 µm, i.e. 0.2100 pixel−1 scale)
6 - 80% of energy within 2 × 2 pixels across the whole field
7 - achromatic PSF or less than 2% effect
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 16

8 - astrometry with an internal accuracy of 0.01-0.0200 rms (goal) and 0.0500 rms
(formal), an accuracy of 0.100 rms with respect to an external standard and
an absolute accuracy of 0.2-0.300 rms (goal) and 100 rms (formal value).
9 - 1% accuracy in flat fielding
10 - multi-color photometry
11 - narrow band imaging
12 - four observing modes: stare (single exposures), jitter (few pixels shifts be-
tween sub-exposures), dither (shifts larger than the largest gaps in the CCD
mosaic) and SSO - solar system objects, facilitating non-siderial tracking
13 - 10-12 filters available on-line
14 - accurate photometric calibration of all broad-band filters to a level of 0.01
mag rms (goal) and 0.05 mag rms (formal) in the zero point in instru-
mental magnitudes and 10% in the color transformation terms from the
instrumental to the Johnson-Cousin standard photometric system
15 - astrometric and photometric performance monitored and administrated
throughout the life time of the instrument; certain data items related to
these calibrations need to be delivered separately to the consortium. Specifi-
cations of these items are given in a separate Interface control document be-
tween ESO-DMD and the OmegaCAM consortium [VST-SPE-OCM-23100-
3551]
16 - observing programs must allow for new inclusions or modifications on a
day by day basis
17 - if required, raw, observational data must be made available to the propo-
nents, for analysis within 2-3 working days
18 - the “pipeline data reduction must be able to fully reduce an average day
worth of science and calibration exposures within 12 hours elapsed time, op-
erator preparation of the execution shall not exceed half an hour” following
the specification of Annex A.4.1 of the MoU.
19 - observing (scheduling), calibration and data reduction stategies should
support the following type of survey strategies :
— Basic observing modes on single fields
— Deep integrations of single fields
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 17

— Wide-area real-time variability searches.

— Wide-area surveys with full sky coverage

The rest of this Section provides further details and justifications for this list.
Note on req. 12:
Wide area coverage and deep integration on single fields, usually require that the
gaps in the CCD mosaic are covered by the combination of several exposures
(≥ 5) with a dithering larger than the maximum gap in the CCD mosaic.
For some programmes (e.g. variability monitoring, weak lensing), it is more
important that at corresponding positions in the sky the PSF shows a minimum
variation. In this case, the best strategy is to obtain multiple exposures with
only a small jittering to remove bad pixels.
Note on Sc. Req. 13:
Because of flexible scheduling, the need to be able to monitor transient or vari-
able objects in several filters, and the measurement of atmospheric extinction,
the number of filters available on-line must be on the order of a dozen. 5 of
these are required for the continuous maintenance of the photometric scale at
4 key bands (see Section 5.6).
Note on Sc. Req. 14:
Accurate calibration to the Johnson-Cousin standard photometric system is
strictly required by many stellar population studies. Required accuracy for such
programmes is 0.01 mag. (Also many other programs, even those not using
standard filters, will require a calibration link to a standard photometric system.)
To provide the necessary infrastructure for this goal a proper monitoring of the
sky conditions, in particular of the extinction coefficient, is required, next to the
continuous monitoring of the overall responsivity variations of the instrument.
Data acquisition and baseline calibration plan shall be designed to achieve the
goal, but this will not be provided by the standard pipeline, which is required to
achieve the nominal 5% photometric accuracy. Additional dedicated off-ESO-
pipeline data-analysis will be required to qualify 1% accuracy.
Note on Sc. Req. 15, (i):
A major fraction of the observing time will be spent on wide area surveys in dif-
ferent filters. Some of the projects may require several years to be completed.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 18

To ensure uniform quality of the products, accurate monitoring of the perfor-

mance of the instrument is required. This performance shall be continuously
administrated and available for trend analysis. This administration should be
integrated in the DFS infrastucture and/or its peripheral infrastructure ready
for inspection and analysis.
Note on Sc. Req. 15, (ii):
ESO’s processing of the OmegaCAM data is described as ”the removal of instru-
mental fingerprints from the raw data”, essentially leading to astrometrically
and photometrically calibrated images. Clearly, more reduction steps have to be
taken in order to obtain extracted source parameters and the error bars on these
parameter values. Also the stacking of images with user provided qualification
criteria (modes deep and mosaic) or the achievement of high, 1%, photometric
accuracy requires re-processing of the data with user provided criteria. In order
to provide the end-user insight in the quality of the data the administrative
data relevant for the OmegaCAM photometric and astrometric calibration need
to be distributed to the consortia National Data Reduction Centers. The Inter-
face document ICD [VST-SPE-OCM-23100-3551] between ESO-DMD and the
OmegaCAM consortium specifies the essential data items for this distribution.
Note on Sc Reqs. 16, 17:
The main purpose of OmegaCAM + VST is to perform the target selection for
VLT follow-up. For variable targets (e.g. microlensing events, GRBs optical
transients, supernovae etc...) this requires flexibility in the observing programs
and near real time data analysis. Even as a stand-alone facility, a medium size
telescope such as VST can become competitive with larger ones if it allows for
a flexible operation and prompt access to the data.
SCIENTIFIC USE CASES
We found in our national communities a widespread scientific interest for wide-
field imaging, covering virtually all fields of astrophysics from the Solar System
to cosmology and ranging from the search for special unique objects to statisti-
cal studies of classes of objects, thus motivating our efforts in the construction
of OmegaCAM.
A few representative scientific projects were sketched in the proposal to build
OmegaCAM, submitted to ESO in October 1998. They were intended to high-
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 19

light specific topics of interest for our communities and which were not pre-
sented in the VST science case documents. Though in some cases the research
strategy of these projects will need to be refined following the most recent ad-
vances in the field, they can still serve as examples of the scientific problems to
be addressed with OmegaCAM. Here, we summarize various scientific projects
which are anticipated for the VST/OmegaCAM and we identify a set of science
driven requirements on the instrument, its operations and its data reduction
system. These projects can be taken as use cases for the OmegaCAM project.

A. Scientific case: Microlensing

The unprecedented combination of OmegaCAM and VST is ideal for studies
of microlensing in the inner Galaxy, allowing the simultaneous detection and
monitoring of the > 14 on-going microlensing events expected per square de-
gree. A small survey, based on two-hour sampling of 10 fields over one month
would yield the microlensing optical depth to brown dwarfs in the disk and
bulge, a regime unexplored by current surveys. A larger effort could be made to
create a bulge microlensing survey that will be unparalleled worldwide. Larger
detection efficiencies for faint and rare microlensing anomalies such as those due
to extra-solar planets are guaranteed by the superior site and field size; if they
are numerous, a major bulge survey could yield tens of planets located a few AU
from their parent lenses every year. The technical advantages of OmegaCAM on
the VST together with the rapid sampling intrinsic to microlensing monitoring
will also produce, simultaneously, a catalog of faint or rare eclipsing objects
that can be used to constrain the binary mass function into the regime of short-
period Jupiter-sized companions. Metallicity and limb-darkening measurements
for faint bulge stars and lens kinematics can be obtained through rapid (5–10
minute) photometric monitoring and VLT spectroscopy of on-going, caustic-
crossing events. Such a capability would require on-site, near real-time analysis
of the data in order to find the microlensing event candidates. Such a mi-
crolensing survey can be interleaved with other projects at any phase of the
moon.
Special requirement on DFS and/or its peripheral infrastructure– As-
tronomical target objects for VLT follow up observations must be identified
from OmegaCAM image data within 1-2 days after data taking. The user
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 20

will have to work on calibrated images and to execute identification/source

extractions
Clearly, such special operations will need additional arrangements, in which
balances have to be found between services provided by ESO and ad hoc data
reductions done by the user at or near Paranal.

B. Scientific case: Optical Transients to Gamma-Ray Bursts

Gamma-ray bursts are short outbursts of highly energetic photons. In 1997 it
was finally shown that GRBs do show afterglows and are situated at cosmo-
logical distances. The HETE-2 experiment (launched on October 9, 2000) will
be able to localize about 50 GRBs per year with an accuracy better than 10
arc-minute (i.e., radial distance from the true burst position) in the medium-
energy X-ray band using the WXM. These localizations shall be calculated on
board the spacecraft within 10-100 seconds of burst onset, depending on burst
duration and temporal structure (GRBs have durations ranging from 10 ms to
1000 seconds). The GRBs coordinates will be transmitted to ground station in
near real time (< 10 seconds), immediately passed to GCN and from there to
the community of ground-based optical, IR, and radio observers in searches for
GRB counterparts and their after glows.
By exploiting the large field of view of OmegaCAM it will be possible to monitor
the first few minutes/hours of the GRBs optical afterglow, even before this has
been identified.
An interesting prediction of current theories is that GRBs may be beamed, in
which case many more optical transients (OTs) (GRBs without γ-rays) than
GRBs with γ-rays will exist due to a lower bulk Lorentz-factor of the emit-
ting material. Estimates of the fraction of γ-ray rich over γ-ray less GRBs
range over more than 4 orders of magnitude, with predicted event rates from
∼ 4 10−3 OT/square degree /yr (isotropic case) to 30 OT/square degree/yr
(pencil beam GRB). Any statistics on this fraction will provide extremely im-
portant information on the beaming factor and, therefore, intrinsic properties
of GRBs. A fast, wide-area optical transient survey independent of gamma-ray
detections will address the issue.
Special requirement on DFS and its peripheral infrastructure – As-
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 21

tronomical target object for VLT follow up observations must be identified

from OmegaCAM image data within 1-2 days after data taking
Also these special operations require additional arrangements, in which balances
have to be found between services provided by ESO and ad hoc data reductions
done by the user at or near Paranal.
Special requirement on Operations – Observing schedule must allow for
modifications on a day by day basis

C. Scientific case: Search for Supernovae at high redshift

Type Ia SNe are probably the best distance indicators to test cosmological
models. Even though two major programs (SCP & HZSS) aimed at the study
of SNe at intermediate redshift (z = 0.4 − 1.0) are already running today, it
will require a few more years to accumulate sufficient statistics and to settle
definitively the debate on the value of the Hubble constant and of the density
parameter. Good statistics are also needed to test the present assumption that
SN Ia properties show no evolution with redshift.
Given the expected limiting magnitude for OmegaCAM, and assuming that a
single field is observed once per month for one year we expect to find 50-100
SN Ia. Most of these will be in the redshift range z = 0.3−0.6, but a significant
fraction (∼ 5%) will be at higher redshift.
Once the targets have been identified the current approach is to spectroscopi-
cally confirm the candidates and measure their redshifts. For the faintest, and
potentially more interesting candidates, this will require an 8-m class telescope
and long exposures. Whereas this remains the best option, an alternative strat-
egy can be devised that relies on multicolor photometry for the confirmation
and classification of the SN candidates (very similar to the photometric redshift
approach used for galaxies).
In addition to contribute to the direct measurement of cosmological param-
eters, this approach could allow for the first time the study of the evolution
of the SN rate with redshift, which contains unique information on the star-
formation history, the initial-mass function of stars and the progenitor scenario.
In particular, estimates of the rates of core collapse SNe at high redshift can
significantly improve our understanding of the intrinsic nature and age of the
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 22

populations involved in the SN explosions and eventually can be used to probe

the star-formation and heavy-element enrichment history of the Universe.
A high-redshift SN search will obviously exploit the wide field-of-view of Omega-
CAM and also, because of the need to disentangle the SNe from their parent
galaxies, will especially benefit from the good seeing at Paranal.
Special requirement on DFS and its peripheral infrastucture– For VLT
follow up candidates must be identified within 3-4 days

Clearly, also these special operations will need additional arrangements, in which
balances have to be found between services provided by ESO and ad hoc data
reductions done by the user.
D. Scientific case: A wide-area proper motion survey
In order to locate and study intrinsically faint stars, a deep wide-field survey
(∼ 100 square degrees) in two colors will be conducted on a regular basis so as
to select those objects with high proper motions. Down to faint absolute mag-
nitude limits, this will allow all objects with halo kinematics within a distance
of 100pc in the directions surveyed to be located.
For example, if a significant population of halo white dwarfs exists, one such
object should be found by this survey per square degree, even if their cooling
ages are 16 Gyr. A proper motion survey is the only way to detect intrinsically
faint stars from the ground: at faint apparent magnitudes, galaxies dominate
the counts to such an extent that faint-star counts are virtually impossible.
Only by pre-selecting objects with proper motions can galaxies be eliminated,
and the faint end of the stellar luminosity function reliably studied.
Special requirement on Instrument – 0.0100 accuracy in relative astrom-
etry

E. Scientific case: A deep 20 sq degree wide-intermediate and narrow-

band survey
This program consists of a deep survey of a total of 20 square degrees towards
several lines-of-sight at various galactic latitudes and longitudes in a set of
broad-, medium-, and narrow-band filters, as well as a deeper I-band survey in
a 5-square degree subfield. The many filters allow a significantly better object
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 23

classification and permit the determination of accurate spectral energy distri-

butions of essentially all objects down to a limiting magnitude of 26 in the
B band. Key research programs will be related to galactic structure, gravita-
tional lensing, low surface-brightness galaxies, and the cosmological evolution
of galaxies and of quasars. Some of these projects rely solely upon VST obser-
vations, while others are of a preparatory nature for detailed VLT observations.
Sub-projects are:
• E1. The contribution of hot subdwarfs and white dwarfs to the Galaxy:
A complete identification of these stars in a subvolume within a volume of radius
of ∼ 30 kpc will solve the controversy about the scale length of their distribution
and thus, will constrain the structure of the halo, and that of the thin and
the thick disks. Furthermore, the luminosity function of these objects will be
determined for the first time (and with precision) thus allowing constraints on
the star-formation history of the Milky Way. Data from OmegaCAM will allow
the identification of hot subdwarfs in the dwarf-galaxy satellites of the Milky
Way yielding constraints on their structure and star-formation history as well.
Special requirement – 0.01 mag absolute photometric calibration
• E2. Galaxies in the nearby Universe (z < 0.1):
A complete census of galaxies down to very low absolute luminosities and very
low surface brightnesses will allow the determination of the luminosity density,
the star-formation rate, and the luminosity function in the local Universe with a
previously unrealized precision. In particular, the contribution of the faint end of
the luminosity function will be clarified. The goal is to obtain reliable statistics
from an appropriately large volume, including a wide range of density regimes,
and with the same selection criteria as applied to high-redshift galaxies. This
knowledge of the local Universe can then be used in comparison with results
of analyses of objects at higher redshifts in order to constrain the evolution of
galaxies and its large-scale structure. The philosophy of this survey is distinct
from that of the Sloan Digital Sky Survey that is more limited to intrinsically
brighter objects.
• E3. Formation and evolution of high-redshift galaxies:
Following the pioneering drop-out technique introduced by Steidel and Hamil-
ton (1992), we will archive a sample of several tens of thousands of galaxies
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 24

at redshifts above 2.5, ≥ 50× more than presently known. These data, and
with follow-up VLT observations, will allow comparative studies of the evolu-
tion of the luminosity density, star-formation activity, oxygen abundances, and
luminosity function in several redshift intervals. Analyses of such data will also
constrain the models for the formation and evolution of galaxies as well as those
for the formation of structure. Statistically complete sets of these objects will
be selected for detailed spectroscopic follow-up with the VLT in order to study
the evolution of the scaling relations of spiral and elliptical galaxies in the field
and in clusters.
• E4. Quasars and AGNs:
The reconstruction of spectral energy distributions as well as surveys of vari-
ability will allow the identification of a large number of AGNs and QSOs over a
large redshift range. This data will constrain the formation epoch of QSOs, and
will serve as a set of faint, optically selected AGNs that will be needed to solve
the longstanding debate on the origin of the X-ray background. Moreover, with
these data, the role of the surrounding density on the AGN phenomenon can be
reliably studied. Finally, the survey will identify a large number of QSOs that
are good targets to study the large-scale structure by the aid of Lyα absorption
systems. All of these projects will need VLT follow-up spectroscopy.
• E5. Weak lensing on different mass scales:
The survey will provide the data to study the distribution of luminous and
dark matter on various scales (10 h−1 Mpc down to 10 h−1 kpc) via lens-
ing techniques. The mass selected (i.e., lensing selected) sample of clusters
and groups of galaxies can directly be compared with N-body simulations of
structure formation, avoiding the usual assumptions about mass-to-light ratios,
luminosity-temperature relation of clusters, etc. We will measure the mass-to-
light ratio of clusters, its scatter, and its dependence on the cluster mass, and
we can quantify any bias caused by the traditional optical clusters selection.
Statistically, we will also be able to measure small density fluctuations by the
large-scale structure (‘cosmic shear’) and thus constrain the amplitude and
shape of the power spectrum directly.
F. Scientific case: Discovering new Solar system objects
OmegaCAM is expected to make a large inroad in solar system astronomy. A
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 25

belief commonly held is that the entire solar system has been inventoried – new
comets and asteroids will continue to be found and catalogued, but these are
just other members of the principal solar system populations. Recent years have
seen this view sharply challenged. The outer solar system (in particular from
Neptune outward) is now known to contain the Kuiper Belt, a vast population of
cold primitive bodies, of which the Pluto-Charon binary is the most well known
member. Two new irregular satellites of Uranus were discovered as recently as
1997, and there is theoretical evidence that the inner solar system still harbours
undiscovered populations in the form of terrestrial and Venus trojans. It is clear
that our inventory of the solar system is far from complete, and this situation
needs to be rectified if we ever hope to understand solar system formation as
a general, galactic-wide phenomenon.
• F1. Kuiper belt objects:
A solar system survey that the OmegaCAM is particularly well equipped to
carry out is the search for the largest (brightest) Kuiper Belt objects. The size
distribution for the Kuiper Belt is only well established between 21 ≤ mR ≤ 26,
since this magnitude region is well suited to conventional telescopes and CCDs.
However, at brighter magnitudes, the surface density drops rapidly so that very
large sky coverage is needed for positive detections. In spite of the difficulties,
it is essential to extend the size distribution to large sizes (bright magnitudes)
because this is the most informative area regarding theories of planetary growth.
The numbers and sizes of the largest bodies in the Kuiper Belt will provide
strong constraints on the ”runaway growth” stage of accretion in the outer solar
system, and should shed light on the still unsolved problem of the accretion of
the outer planets.
Since the objects of interest are bright (mR ∼ 20), such a survey could be
carried out during bright time. A sky coverage of ∼ 1000 deg2 at mR ∼ 20 can
be achieved in 1 year with ∼ 5 hours of observation each month. Extrapolating
from the known size distribution, the survey is expected to yield ∼ 20 Kuiper
Belt objects brighter with mR ∼ 20 in 1 year. With this rate, one could equal
Clyde Tombaugh’s survey (the largest outer solar system survey to date) in
merely 1.5 years. Such a survey would answer the intriguing question whether
other ”Plutos” (bodies in the 1000 km-radius range) exist. Even without the
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 26

spectacular discovery of more ”Plutos”, the expected yield of bright Kuiper

Belt objects will provide an as-yet unavailable sample of bright targets ideal for
detailed physical studies.
The typical angular motion of Kuiper belt objects is small (∼ 3”/hr) and the
observations can be done with standard siderial tracking and guiding (using
observing Mode– Stare N=). The data are preferably accessed within 3-4
days after the observations and could be processed with the standard image
pipeline. (The slow motion of the KBOs means that they can be recovered even
a few days after discovery, so that real-time data processing is desirable, but
not imperative. It is sufficient, for example, to flatten one night’s data within
∼ 1 − 2 days, and to spend another ∼ 1 − 2 days searching for candidates.
After 4 days, a typical Kuiper belt object will have moved only ∼ 300” and
should be easily recoverable with conventional CCD detectors.)
• F2. Discovering fast moving objects such as asteroids near Earth and
Venus at L4, L5:
At the other end of the solar system, i.e., the inner solar system, opportunities
for exciting new discoveries also exist. It would be valuable to carry out searches
for terrestrial and Venusian trojans, i.e., asteroids located at the L4 and L5
Lagrangian points of the Earth and Venus, respectively. In contrast to the
small angular motion of Kuiper Belt objects the angular rates of these inner
solar system bodies is ∼ 140”/hr for terrestrial Trojans and ∼ 240”/hr for
Venusian Trojans. For some (sometimes uncertain) populations the orbital
parameters can be predicted, implying that the searching for such faint objects
can be facilitated by a non-siderial tracking of the telescope on the anticipated
fast motion of the objects. As the typical exposure time of OmegaCAM is not
expected to exceed 500 sec, the remaining uncertainty in the angular speed of
the targets will nominally lead to larger image smearing than the effect of the
tracking imperfections of the telescope without guiding. Thus there appears
no compelling need for auto-guiding in such a non-siderial tracking mode.
Special requirement – Non-Siderial tracking observing mode - Mode–
SSO N=2-5
The option of a non-siderial tracking observing mode could support all kinds
of other solar system observations.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 27

The urgency of fast data reduction when dealing with Trojans of the terrestrial
planets: in the extreme case, a Venusian Trojan will have moved 1.6 deg in 24
hours and will be difficult to recover with common CCD cameras. It should be
easily recoverable, however, with wide-field cameras like OmegaCAM. For this
reason, a survey for very fast moving objects such as near-Earth asteroids and
terrestial Trojans will necessarily include follow-up time roughly 24 hrs after the
discovery observations.
Special requirement – Provide users access to the raw data within 24
hours within Chile

2.2 DFS requirements

GENERAL
The Science requirements (Section 2.1) 15, 16, 17, 18 (see also Notes 15i, 15ii)
propagate directly to requirements on the DFS pipeline infrastructure and the
DFS peripheral infrastructure (housing operators manually running tools) and
are not repeated here.
Prominent derivatives from these requirements include:
• Provide a quality control on all observational data– this task is matched by
ESO’s QC0 and QC1
• Provide a quality control on all results of the processing of OmegaCAM
data, such as Calibration files and image data— this is provided by a trend
analysis. Trend analysis is used as a tool to monitor the behaviour of
the instrument. The results of trend analysis are not used to interpolate
(calibration) results to be fed back to the pipeline.
• Parallel processing of image data
• Distribution of administration and calibration data to National data cen-
ters/Users facilitating further source extraction re-processing and error bar
evaluation.
The Interface Control Document [VST-PLA-OCM-23100-3551] specifies the
essential data items for this distribution.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 28

QUICK ACCESS - 2-4 DAY CYCLE

The use cases:
- Supernova/ Optical transients
- Microlensing
- Kuiper Belt objects and fast moving asteroı̈ds
require some fast follow-up observations within typically a few days (1-2 lensing)
to 3-4 days (supernova). For brevity we call this the 2-4 day cycle.
In the near future it is unlikely that the data transfer rates required to elec-
tronically transfer the raw image data to Europe will be available to the VST
observers, but this might become possible in a couple of years, since the tech-
nology already exists.
For the time being, it appears that a 2-4 day cycle involves some data reduction
to take place at, or near, Paranal. For most observing programmes it is enough
to provide the user with the raw data and some calibration data. The user
would then have to run some quick data reduction steps to identify the events
which require fast follow-up observations.
As an indication how frequently and with what kind of data volumes the 2-4
day cycle would be loaded, the Consortium has made a small inventory on the
estimated usage for the GT programmes. This might be used as an indication
for the total usage, but beware individual programmes which might ask for
exceptional data rates and we advise the OPC to include this as a consideration
in the allocation procedures for any (also non 2-4 day cycle) programme.
Supernova IA searches: GT about 10 nights per year (10 times 40
frames).
For the search with OmegaCAM, typical exposure time could be 45-60 min
(splitted in 3-5 dithered exposures). This makes 30-40 science frames per
average night. For the search, image processing through archive calibration
files (master flats and bias) is sufficient.
Micro-lensing: GT about 15 nights/year
We would spend (GT) two 1-week campaigns a year on monitoring the bulge,
but there could also be a more extensive effort by combining with open time
or other partners. This programme will mostly be useful for generating tighter
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 29

limits on the optical depth in very short-duration events, which do not require
follow-up per se. However, there are microlensing events all the time in each
field, and it is likely some of these will be useful to study in real time with
VLT (spectroscopically). In that case it is necessary to search for these events,
which requires to analyse ALL data of those 2 weeks in near-real time. The
programme could combine a microlensing search for brown dwarfs with a transit
search for hot jupiters in the same fields.
Solar system Objects: Open Time up to 2-3 nights/month with 60
frames/night.
As already indicated by the several SSO oriented use cases described in Section
2.1 there are a number of programmes which need a fast turn around time.
Although the interest for such programmes is great we have not assigned much
GT time to this and it is very difficult to estimate the usage or anticipate on
OPC allocations for this mode. One possible OT proposal relayed to us asks
for typically 5 nights/month for this mode for which roughly half of the data
require the 2-4 day cycle.
Two SSO programs that need fast follow-up:

(1) Search for Near-Earth asteroids: this has 2 levels of turn-around time. The
objects really nearby and not yet discovered need immediate follow-up obser-
vations, basically during the same night or within 1-2 nights after discovery.
Somewhat more distant objects objects in the asteroid belt require a turn -
around time of around two weeks.
The searches must be done in both cases on the whole data set. The exposures
will have 5-10 min integration time. The shorter the turn-around cycle time is
the better.
(2) Search for Kuiper Belt Objects and MBOSSes: a few epochs per object
need to be observed. Turn-around times up to the order of one month could
be allowed, but it is preferably done within one week; this also depends on
the survey method (one or two exposures per field) and how one combines the
survey with other scientific programmes. Typical exposure times for this search
will be 10-15 min with few images per field per night.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 30

SUM-UP 2-4 day cycle

The present rough estimate indicates that the consortium would want to make
use of a 2-4 day cycle for a maximum of 25 nights/year with on average
about 60 frames/night of its GT. The supernovae search (10 nights) has a high
priority on our programme. Regarding open time, only for the SSOs one might
expect of the order of 24 nights/year (also with on average 60 frames/night).
Some of these programmes can be optimized by combining these programmes.
As the lower priority programmes will not all run year after year we estimate
that in practice the 2-4 cycle would be used for a fraction ranging from 10%
to 20% of the total time.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 31

3 INSTRUMENT CONCEPT- Summary

This Section summarizes those characteristics of the camera which potentially

can have an impact on the User requirements on calibration and data reduction.
Particularly relevant are the focal plane geometry, the cosmetics of the CCD’s
and the chosen set of both narrow and wide band filters.

3.1 Description
OmegaCAM is a wide field optical imager, featuring a 4 × 8 mosaic of 4k × 2k
pixels CCD’s for a total imaging area of 16k × 16k pixels. It will cover the
VST field of view of 1◦ × 1◦ and at the same time it will adequately sample
the best seeing foreseen at Paranal. The standard observing mode is with a
two-lens field corrector. In addition, a single lens plus atmospheric dispersion
corrector can be used.
The instrument features a two-blade photometric shutter. An exposure is
started by moving the blade that obscures the CCD’s to a rest position, and
finished by moving the other blade from its rest position to the position that
obscures the CCD’s.
The focal plane geometry of the CCD’s is given in section 3.2 and the Omega-
CAM filters are described in section 3.4.

3.2 Focal plane geometry

The layout for the CCD array is sketched in Fig 3.2. It consists of a, roughly
square, science array of 16384×16384 pixels, and four auxiliary CCD’s for
guiding and image analysis.
With the preferred LL,RR arrangement for orientations of the CCDs’ readout
ports the gaps between sensitive areas are 6.0 and 2.0 mm. Vertically there
will be 7 2.0mm gaps, leading to a total gap of 14mm vertically.
16384 pixels of 15µm comprise 246mm. Adding in the gaps, the total light-
sensitive part of the science array is thus 259 x 260mm in size. The full unvi-
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 32

gnetted field of the VST is 1.47 degrees, or 372mm in diameter. The planned
array with gaps fits in the field.

Fig 3.2 Sketch of layout of CCD arrays. Units of X and Y axes are in mm.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 33

3.3 CCD details

An extensive description of the CCDs is given in VST-TRE-OCM-23100-3042

3.4 Filters
The OmegaCAM filters are to be decided on jointly by the VST, OmegaCAM
consortium and ESO. ESO has set up a VST/OmegaCAM Instrument Science
Team (IST) for advise on such issues.
Discussions between VST and OmegaCAM have resulted in the following list of
filters:
• Broad band (6 filters)
Given the importance of the Sloan survey in the North, the rather large
width of g’ for stellar work, and the desireability to maintain small colour
terms with respect to the standard Johnson system for many applications,
the top priority broad band filter set is:
– 2 Johnson B, V
– 4 Sloan u’ g’ r’ i’

• Narrow band – current preferences in bold

– O[III]/100Å (FWHM) at redshifts of 2000 and 6000 km/s.
– Hα/80Å(FWHM) at redshifts of 2000 and 6000 km/s

The broad-band filters will be monolithic. The narrow-band filters are seg-
mented, made up of two halves. A scheme where each half is made with a
different passband is under investigation. Half the array would then be illumi-
nated through one passband, half through the other. After a second exposure
with the camera rotated through 180 degrees, the whole field will have been
covered through both passbands, which can serve as each other’s off-band
image.
In addition a composite filter, comprising a u’, B, V, and i’ quadrant for rapidly
obtaining photometry measurements is envisioned. For details see section 5.6
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 34

Note that the f/5.5 beam is transfered to f/6 in front of the filters due to the
corrector optics.
Note that the fast beam makes it impossible to define truly rectangular filter
profiles. A relative wavelength spread of at least 0.4% is unavoidable. At
5000Å, for instance, this represents a 20Å broadening of any filter profile.
Thus, an O[III] filter with a FWHM of 50Å= 3000km/s, has a flat part of
40Å= 2400km/s. For Hα filter with FWHM 60Å= 2600km/s, only 46Å=
2000km/s is flat.
For this reason it is not useful to define rectangular filter transmissivity profiles
for the narrow-band filters. This may reduce the number of coatings, and hence
the cost, of the filters.

3.5 Control electronics

The OmegaCAM subsystems to be controlled by the control electronics include:
• Filters Unit
• Shutter Drive Unit
• Filter identification system
• Instrument status and diagnostic system
• Interlock system
The design is given in Hardware Control Electronics, VST-TRE-OCM-23100-
3043

3.6 Instrument Software

The Instrument Software shall be self-supporting and will internally verify whether
commands have been successfully executed. Malfunctions or special status of
the instrument will be displayed on the instrument monitor.
QC0 and the DFS-QC1 will not further check or double check these items.
The User requirements on the OmegaCAM instrument software are given in a
separate document VST-SPE-OCM-23100-3060.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 35

4 OBSERVING MODES and STRATEGIES

OmegaCAM employs a fixed set of observing modes. These observing modes

use a particular dithering method (Section 4.1 URD) and either use a set of
key-filters or User-filters (Section 4.3 URD). The observing modes shall be
implemented in observing templates (Section 4.4 URD).
Supervising observing strategies subscribe additional instructions for the schedul-
ing and pipeline data reductions (optional for ESO-pipeline, Section 4.2 URD).
The calibration procedures, often involving a trend analysis for qualifica-
tion/quality control purposes, are designed to support the set of observing
modes and employ as much as possible identical observing modes.
A prominent concept is the designation of keyfilters, which are used to monitor
the behaviour of the instrument and atmospheric transparency: the observa-
tional data obtained with the keyfilters will be subject to extensive adminis-
tration and trend analysis. Contrary to the keyfilters, the data obtained with
userfilters, in so-called user mode, will not be subject to trend analysis. User-
mode data are cross-calibrated nightly to the key-filters. Observations taken
with the Atmospheric Dispersion Corrector are considered as user mode.
The information about strategies and modes shall be strictly carried through
the DFS by, amongst others, FITS-header items.

Glossary of OmegaCAM terms

Key filter Used for photometric calibration, instrument and atmo-
sphere monitoring
Key bands Sloan u’ and i’ and Johnson B and V
Mode– Observing technique
Mode– Dither Shifts larger than the gaps in the CCD mosaic
Mode– Jitter Few pixels shifts between sub-exposures
Mode– Stare Independent exposure(s) with identical pointing
Mode– SSO Solar System Object / Non-Siderial tracking
Strategy– Observing strategy
Strategy– Deep Does deep integrations
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 36

Strategy– Freq Frequently visits (monitors) the same field

Strategy– Mosaic Maps areas of the sky larger than 1◦
Strategy– Standard Consists of a single observation

4.1 Observing Modes

In order to overcome:
i) the effect of the blank columns and rows in between individual CCD chips,
ii) bad (hot or cold) pixels (columns) of the CCD chips,
iii) the effect of cosmic ray events,
dithering (i.e. obtaining very slightly offset images of the same patch of sky) will
play a prominent role in the data acquisition and its associated data reduction.
Dithering can involve from N = 1 to 11 different positions on the sky.
There are two spatial domains to which the dithering can be optimized:
I Dither offsets matching the size of the largest gaps of the CCDs in the
focal plane (= above item i).
II Jitter offsets matching the smallest gaps in CCDs, e.g. dead columns and
bad pixels (= above items ii and iii).
While dither mode will provide maximum, complete, sky coverage with a
minimum effect of the gaps in between the CCD’s, its context map will be very
heterogeneous (complex). Conversely, jitter mode will result in a very homo-
geneous context map, but the gaps in between CCDs will remain unobserved
and sky coverage is incomplete. Depending on the scientific objectives of the
observations the User can choose between dither mode with maximum sky
coverage (completeness) and jitter mode with maximum homogeneity of the
acquired data, which is expected to propagate as lower effective noise in the
reduced data, as is desired for e.g deep fields.
Both dither and jitter observing modes will be supported. There will also be a
stare mode, for which there are no offsets, and a mosaic mode, which is used
to generate images larger than a single OmegaCAM field.

Mode– Dither N= will be operated with N pointings on the sky, with offsets
matching the maximum gap between arrays, i.e. ∼ 380 pixels (∼ 80 arcsec). In
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 37

standard observing mode N =5, maximum supported (pipeline!) value of N is

11. Mode– Dither N= covers all the gaps in the focal plane, thus the resultant
image has no gaps within the field of view, but the context map describing the
origin of the various pixels in the resultant map will be very complex. Dither
mode is optimized for maximum sky coverage. Mode– Dither N= is used for
general surveys, which want to achieve complete sky coverage.

Mode– Jitter N= will be normally operated with N=3-5 pointings, with quite
small offsets (∼ 5 pixels). Maximum N = 11. Mode– Jitter N= is optimized
for observations which require maximum homogeneity of the context map, and
for which the acquisition of information in the wide CCD gaps is not critical.
Mode– Jitter N= will have a minimum of discontinuous variations of the PSF
over the field of view; also all the data from a single sky pixel originate from a
single CCD chip.

Mode– Stare N= will have one fixed pointing position, but the same position
of the sky might be re-observed subsequently (N>1). Mode– Stare N= can
be used to take all kinds of snapshots or calibration frames, but will also be
the main workhorse for monitoring optical transients or fast moving objects.
Facilities for direct on-line comparison of single Mode– Stare N=2 frames
both for variable target monitoring and for instrument qualification monitoring
will be an important aspect of this mode.

Mode– Mosaic N= will facilitate surveys that aim to image areas of the sky
larger than 1◦ diameter, or which for any other reason want to combine data
from two adjacent partly overlapping fields. The prime characteristics of this
mode are:
– the coordination of the definition of a set of pointings on the sky - in the
OB planning
– the combination of the data from different fields into one data product
In principle, this mode should be handled by the de-dithering (=co-addition)
process, (this can be interpreted as a requirement on the co-addition code,
namely produce an output image which is the rectangle of the overlapping
area of several input images). If the co-addition (particularly the set of quality
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 38

control parameters) and the source extraction methods are handled correctly,
this mode will form a natural extension of the dither mode and does not require
additional developments. Maximum image size supported for this mode is TBD;
for mosaiced source lists the maximum covered sky area is unconstrained and
data volumes could well run into the Tbyte regime. It is not planned for the
standard DFS pipeline to produce mosaiced images.
Mode– SSO N= will support the data taking of Solar system objects with
a non-siderial tracking of the telescope. No-autoguiding is required for this
mode as in general the uncertainties of the predicted angular velocities of yet
to be discovered objects are larger than the expected jitter of the telescope
when the autoguiding is disabled. This mode resembles very much the Mode–
Stare N=, however the wanted quick (<24 hour) data reduction for many fast
moving targets implies that the DFS pipeline will not support the processing of
these data fully. However, for observations which require a slower response (4
times 24 hour or longer) or which only require de-biasing and flatfielding, the
standard pipeline can be configured to handle these data.

4.2 Observing Strategies

An observing strategy employs one of the basic observing modes and defines
a number of additional instructions for specifically the scheduling of the obser-
vations. The observing strategy shall be recorded in the FITS headers of the
observations. Optionally, this header information can be used in data reduction
pipelines, particularly those operated by the Consortium when addressing the
combination (e.g. stacking) of images. It is not expected that the ESO pipeline
will recognize strategies, as the ESO pipeline will not combine various runs.
We discriminate strategy:
Strategy– standard which consists of a single observation (observation block),
Strategy– deep which does deep integrations, possibly taken at selected at-
mospheric conditions over several nights; the standard image pipeline will not
combine images taken over several nights.
Strategy– freq which frequently visits (monitors) the same field on timescales
ranging from minutes to months and has overriding priority on the telescope
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 39

schedule
Strategy– Mosaı̈c maps areas of the sky larger than 1◦ , which is essentially
an item for the scheduling, as the pipeline has to produce uniform quality
data anyway. The combination of various field centers into one image is not
considered a standard pipeline task.

4.3 Filtering
OmegaCAM employs the following type of filters (see Section 3.4 for details):
- Broad-band filters (Johnson B,V and Sloan set)
- Intermediate band (e.g. Stromgren and deep sky filters 200A)
- narrow band ( e.g. O[III], Hα )
A set of keyfilters will be defined which will be used for the photometric cal-
ibration. These keybands will be used for monitoring the instrument and the
atmosphere at timescales from hours up to years. The keyfilters will be used
for standard, beginning of the night checks, the monitoring of the atmospheric
extinction and for a variety of trend analyses relating to the photometric cal-
ibration of the instrument. The photometric system of the instrument shall
be continuously maintained at these keybands. One composite keyfilter will
be available, which will contain in each quadrant a different passband, namely
the four different keybands. The composite filter will be extensively used for
quick checks, extinction measurements and other monitoring of the photometric
system.
The observing modes related to the photometric calibration (req.’s 5.6) will
employ the concept of keyfilters. The other filters (Userbands) will be cross
calibrated versus the keyfilters (see req.5.6.5). Monitoring of the atmospheric
extinction will be done with the keyfilters (req.562).
The 4 keyfilters are Sloan u’ and i’ and Johnson B and V.

4.4 Observing Templates

The Observing Templates for OmegaCAM are defined in Instrument Software
Functional Specifications (VST-SPE-OCM-23100-3064). This definition is based
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 40

on the requirements as set out in the present document. The observation modes
described in the previous section imply three basic OTs for science observations.
- TSF– OCM img obs dither
Observe with N pointings (default 5) in the sky, with offsets > maximum
gap between detectors (> 750 pixels) between exposures.
- TSF– OCM img obs jitter
Observe with N pointings (2-5) in the sky, with small offsets (≈ 5 pixels)
between exposures.
- TSF– OCM img obs stare
Observe N (default 1) exposures with one fixed pointing position.

4.5 Field correctors

The VST and OmegaCAM employ two different field correctors:
Two lenses field corrector: In the U–I bands (0.320–1.014 µm) the worst
case (edge of field) diffraction encircled 80% energy diameter is 1.33 pixel at
Zenith. The transmission is 93%.
One lens plus Atmospheric dispersion corrector - ADC: In the B–I bands
(0.365–1.014 µm) the ADC produces a diffraction encircled 80% energy diam-
eter of 2.0 pixels (worst case at edge of field) at a Zenith distance of 50◦ . The
transmission is 88%. At 70◦ Zenith angle this is 2.88 pixel (2.63 average over
the field). The ADC cannot be used in the U band.

For observations near Zenith the two-lens corrector will be preferred because of
its higher throughput. At larger zenith distance (value depends on wavelength),
the ADC will yield better imaging performance.

The standard (=key) observing mode is with the two-lens corrector.

The ADC observing mode is considered a User mode.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 41

5 BASELINE CALIBRATION REQUIREMENTS

5.0 Documentation system, Odoco

The trajectory of the data through the DFS shall be guided by the Omega-
CAM documentation system, Odoco. The set-up of the specifications of the
calibrations and its data reduction is done in purely requirement driven fashion.
The calibration documentation system (Odoco) is a collection of files which
contains a description of both the requirements on OmegaCAM calibrations, as
well as detailed descriptions of the envisaged Calibration Observations (CO’s)
and their analysis and data-reduction. The Odoco is meant to avoid unnec-
essary duplication of work for the documentation of the various stages of the
project; from the definition of basic requirements, to implementations and final
operation and user manuals. It should also aid the documentation of develop-
ment work. The Odoco was originally developed for ISO and provided a full
uplink system (IOCD). The present version, adapted for OmegaCAM, contains
essentially only the part which deals with text, pseudo code, recipe’s and auto-
matic document creation. The Odoco system is essentially a set of TeX macros,
with LaTeX emulation, together with some supporting C routines. The whole
present document is generated by Odoco, but particularly this section, listing
the baseline calibration requirements, uses some more advanced Odoco options.
The very strict document control of the original IOCD will not be maintained,
but CVS (Concurrent Versions System) will be used for local versioning control.
Official versions of documents shall be filed separately.
The contents of the Odoco will be continuously evolving and fonts are chosen
for optimal reading on a computer monitor screen.
In Odoco, calibrations are specified under requirement subsections (req.’s)
which are labeled with 3-digit numbers. Each subsection contains a number of
items: e.g. the objective of the requirement, a description of a specific Calibra-
tion Observation or a cross-reference to the use of a CO that has been defined
under another requirement. Also, the end-results have been specified and the
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 42

text contains various descriptions for both Template Signature Files (TSF’s)
which define the creation of the observation blocks (OB) and for the off-line
data analysis. Overall priorities have been defined (essential, very important,
desirable) and are specified under the item priority.
The chosen items for the descriptions of the requirements (req.’s) match well
to the items needed for the recipes for DFS data reductions. In section 6.1
complete listings of both the req. items and the recipe items are given. The
req.’s as listed in the Odoco will eventually evolve into the deliverable recipe’s.
The Odoco is designed to provide a comprehensive and accessible documenta-
tion system of the various activities that relate to the OmegaCAM calibrations.
It serves a variety of purposes, and facilitates the extraction of text from the
Odoco data base into complete documents. The Odoco can provide the follow-
ing documents:
1. A listing of the Baseline OmegaCAM Calibration Requirements. Odoco
contains an up-to-date listing of all the baseline requirements for the Omega-
CAM calibrations, i.e. for each requirement (req.) the text under the items:
Objective, When performed/frequency, Required accuracy, Priority.
See section 5 of the URD.
2. Full documentation of the OmegaCAM calibration plan. A detailed
description of all the OmegaCAM calibration requirements and their imple-
mentations. See Section 5 of the Calibration Plan. Summary sections (two
digit sections) have been introduced for a variety of calibration activities:
e.g. detector specific calibrations, photometric calibrations etc. A general
overview of the OmegaCAM calibrations can be obtained by printing the
summary sections of the Odoco. In order to further ease the readability of
this document, both each requirement and each calibration analysis proce-
dure text item begins with a ’one-liner’ stating the overall idea.
3. A description of the Template Signature File necessary to produce
observation blocks, TSF’s. When a requirement can not be fulfilled by
means of data analysis of observations made for another requirement, Odoco
contains a detailed description of the instrument configuration and proce-
dures under the items Sources, observations and TSF, (TSF, Template
Signature File). Note, the term selfstanding has an important meaning:
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 43

when a requirement is selfstanding, it will normally propagate as a single

dedicated calibration observation, with a single dedicated data reduction
task. Conversely, non-selfstanding requirements will have more complex de-
pendencies and often involve a data reduction of data taken for another
selfstanding requirement.
Under the item TSF the hierarchical structure of observation specifications
is detailed (when applicable) on different lines:
– first line: observing Strategy–
– second line: observing Mode– N=
– third line: generic/base TSF–
– fourth line: specific/dependent == TSF–
4. Description of Calibration Analysis (CA). For each requirement, a spec-
ification of the data analysis related to the requirement is given under
the item CA. Standard functionalities can be quoted in the optional item
Needed functionalities for CA. A detailed description of the implementa-
tion, which could include guidelines for the data analysis or pseudo code is
given under the items CAP (Calibration Analysis Procedure). Inputs and
Outputs defines the various calibration tables. Thus a document listing all
the text of the items CA, CAP and Inputs and Outputs gives a complete
overview of the Calibration data reduction analysis.
5. A reference document for timelining OmegaCAM Calibration obser-
vations. The items When performed/frequency and Estimated time
needed can be used to design a detailed Schedule of calibration observa-
tions both in the commissioning phase (CP) and during the Routine Phase
(RP).
6. A listing of the various requirements for an Astronomical Calibration
Source data base, (ACS-dbase).
7. The recipes belonging to the execution of requirements.

As Odoco can provide various documents with a different filtering of the source
of information, each printout contains a table, listing the selection criteria.
Also, the status of the printout is marked (formal issue, or private workcopy).
Each printout contains this section.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 44

On the following pages a print-out is included which is believed to be relevant

for the present document.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 45

OmegaCAM - USER REQUIREMENTS

Baseline Calibration Requirements

Issue: VERSION 1.1

Date: 29 Oct 2001
Prepared by: Valentijn, Begeman, Boxhoorn, Deul, Kuijken, Rengelink
Purpose of printout: FDR

Selected items from the Odoco file system

Summary sections •
Items:
Objective •
Fulfilling or fulfilled by
When performed/frequency •
Sources, obs.,. . .
Inputs
Outputs
Required accuracy, constraints •
Estimated time needed
Priority •
Template Signature File.
Recipe
Calibration Analysis spec’s
Needed Functionalities
CA implementation (pseudo code)
Status of Req.
FLAG

This document is prepared by the Odoco Document Control System.

 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 46

LAY-OUT of BASIC CALIBRATION REQUIREMENTS

DETECTOR RELATED

5.1 Functional Checks

5.2 Electronics specific
5.2.1 CCD read noise - doit CalFile– 521
5.2.2 CCD hot pixels CalFile– 522
5.2.3 CCD gain CalFile– 523
5.2.4 Electromagnetic compatibility
5.2.5 Electrical cross talk
5.3 Detectors specific
5.3.1 CCD Dark current - doit CalFile– 531
5.3.2 CCD Particle event rate CalFile– 532
5.3.3 CCD linearity CalFile– 533
5.3.4 CCD Charge transfer efficiency CalFile– 534
5.3.5 CCD Cold pixels CalFile– 535
5.3.6 CCD hysteresis CalFile– 536
5.4 Detectors operational
5.4.1 Bias - doit CalFile– 541
5.4.2 Flat field - Dome CalFile– 542(L)
5.4.3 Flat field - Twilight CalFile– 543
5.4.4 Flat field - Night sky CalFile– 544
5.4.5 Flat field - fringing CalFile– 545
5.4.6 Flat field - master flat CalFile– 546(W)
5.4.7 Quick detector check CalFile– 547(r)
5.4.8 Illumination correction CalFile– 548
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 47

LAY-OUT OF BASIC CALIBRATION REQUIREMENTS

cont’d

5.5 Astrometric
5.5.1 Focal plane position of camera CalFile– 551
5.5.2 Telescope Pointing
5.5.3 Telescope and rotator tracking
5.5.4 PSF anisotropy CalFile– 554
5.5.5 The astrometric solution - doit SeqFile– 634(CR)
5.5.6 Astrometry - Guide CCD’s CalFile– 556
5.6 Photometric
5.6.1 Shutter timing
5.6.2 Photometric monitoring CalFile– 562(uBVi)
5.6.3 Zeropoint key bands doit CalFile– 563Z,I
5.6.4 Zeropoint user bands CalFile– 564
5.6.5 Filter band pass - user− >key CalFile– 565
5.6.6 Rotation angle - ADC, rotator
5.6.7 Linearity
5.6.8 Detection limit
5.6.9 Secondary standards CalFile– 569
5.7 Alignment
5.7.1 Camera focus/tilt
5.7.2 Ghosts - ADC
5.8 Telescope
5.9 End to end tests
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 48

5.1 Functional Checks

The commissioning plan [VST-PLA-OCM-23100-3100] lists a large set of ac-
ceptance tests. Most of these test are engineering tests and need not to be
repeated here. Those engineering tests which can be executed with require-
ments from the URD/CP contain the proper reference to the URD in the
commissioning plan.
The technical operations and maintenance plan [VST-PLA-OCM-23100-3080]
describes activities during Routine Phase (RP). Both the URD, CP and the
DRS specifies the requirements and the activities for fulfilling these require-
ments during RP. The label RP always points to activities that shall be followed
in the technical operations and maintenance plan.

5.2 Detector Electronics specific Calibrations

Section 5.2 contains the requirements for the characterization of the detec-
tor system on electronics level, while Section 5.3 lists more detector specific
calibrations. The separation between these Sections is somewhat artificial. In
Section 5.4 more daily characterizations are listed, which involve the flatfielding
and de-biasing. Requirements which are foreseen to become ‘workhorses’ are
labeled with doit, e.g. a quick daily evaluation of the read noise serves as a
daily health check.
The CCD’s are operated at one port; electrical cross talk is not expected
to significantly affect the observations. However, as a check, the absence of
significant cross-talk is verified (req.525).
The hot pixel (req.522) and cold pixel (req.535) characterization are com-
bined in the weight maps (req.546).
As the standard read-out time of the arrays is already very fast, ∼ 40 sec , no
extra fast read-out mode is supported in the characterization.
CCD rebinning mode and windowing mode are not supported in the calibra-
tion and data reduction procedures.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 49

5.2.1 Req. — CCD read noise - doit

Objective:
Measure the CCD read noise (in ADU’s) as a standard health check.

Pairs of zero-second bias exposures are used. The rms scatter √ of the dif-
ferences between two exposures is computed and divided by 2. Monitor
variations with trend analysis. This is the first order daily health check.
When performed/frequency:
daytime- Commissioning, during all operations: daily health check.
Required accuracy, constraints:
Readout noise less than 5e− .
Variation in readout noise w.r.t. previous readout noise less than 0.5e− .
These are lab values. The corresponding limits in ADU can be calculated
using the e− /ADU conversion factor from req.523.
Priority:
essential
5.2.2 Req. — Hot pixels
Objective:
Determine CCD bad/hot pixels.

5σ outliers in the master bias frame are bad-hot pixels. These pixels should be
recorded and ignored (assigned a weight of 0) in dedithering and dejittering.
For this purpose the bad/hot pixel map is used to assign a weight of zero to
the affected pixels in the weight map (req.546). The search for hot pixels
would also identify traps.
When performed/frequency:
daytime- Commissioning, in RP twice per week.
Required accuracy, constraints:
Number of hot pixels to be determined by experience/lab values. The total
number of bad pixels (hot pixels + cold pixels) is less than 80000. Difference
in number of hot pixels w.r.t. the previous version, less than 100.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 50

Priority:
essential
5.2.3 Req. — CCD gain
Objective:
Determine CCD gain and variation with time
Determine the conversion factor between the signal in ADU’s supplied by the
readout electronics and the detected number of photons (in units e− /ADU)
and monitor variations in time.
The gain factors are needed to convert ADU’s in raw bias-corrected frames
to the number of electrons, i.e. detected photons.
Take two series of 20 dome flatfield exposures with wide range of exposure
times. Derive the rms of the differences of two exposures taken with sim-
ilar exposure (integration time). Exposure differences of pairs should not
exceed 4%. The regression of the square of these values with the mean
level yields the conversion factor in e− /ADU (assuming noise dominated by
photon shot noise). Compare with previous measurements, as a qualification
(trend analyis).
When performed/frequency:
daytime- Commissioning, in RP once week.
Required accuracy, constraints:
Accuracy: In units of e− /ADU, from lab values or found empirically. Trend
analysis better than 1%. On-site quality check.
Quality check: Difference with previous version less than 10%.
Priority:
essential
5.2.4 Req. — Electromagnetic Compatibility
Objective:
Verify whether any external source (e.g. dome drives, control systems) is not
interfering in the CCD overall detector system, leading to additional, mostly
non-white noise.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 51

Technical specifications require less than 20% effect on read-out noise, for ex-
ternal interference and less than 10% effect on read-noise for internal Omega-
CAM interference.
If electronic interference occurs then this will put constraints on the operation
of the instrument. For example, if interference occurs during movement of
the telescope, one cannot read the CCDs and move the telescope at the
same time,.
Interference is detected by measuring the read noise (req.521) under opera-
tional conditions. This means doing bias measurements while the telescope
and/or dome are moving.
When performed/frequency:
Day time; Commissioning; once a year; every time a major system change
has been made; To be determined by experience
Required accuracy, constraints:
Difference between read noise under operational conditions and the standard
read noise measurement should be smaller than 20% for external and 10%
for internal causes of interference.
Priority:
essential
5.2.5 Req. — Electrical cross talk
Objective:
Although crosstalk is not detectable in the WFI, and only one part per CCD
is used, the sharing of one FIERA by 16 CCD’s opens up the possibility of
cross talk.
Observe a bright (mag 5-8) star at 16 different chips (1 FIERRA serves 16
chips).
When performed/frequency:
Nighttime Commissioning.
Required accuracy, constraints:
10−5
Priority:
desirable
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 52

5.3 Detectors specific calibrations

5.3.1 Req. — CCD Dark Current - doit

Objective:
Measure CCD dark current (in ADU/pixel/sec) for qualification purposes of
the detector chain (qualification and trend analysis). The particle event rate
will be determined on the fly.
Repeating the test with the dome lights on will provide information on pos-
sible light leaks.
Three one hour exposures are taken with the shutter closed. After rejection
of the cosmic ray events, the signal above the bias level is the dark signal.
For the reduction of the science observations the subtraction of the sky
brightness will include the dark current, and a separation of both contribu-
tions is normally not required.
Do a trend analysis.
When performed/frequency:
Daytime (if dome and camera are proven to be light tight enough) - Com-
missioning; once per week. Alternatively, one dark frame per day could be
taken, followed by a trend analysis once/month.
Required accuracy, constraints:
Dark count rate should be less than 3 e− /pixel/hour excluding bad pixels.
Accuracy of determining particle event rate 1 ADU/cm2 /hour.
Particle event rates should be identical for each chip.
Priority:
very important
5.3.2 Req. — CCD Particle Event Rate
Objective:
Determine CCD particle event rate by evaluating dark current measurements.
Verify the absence of a local radiation source affecting the detector. The
data will be inspected for significant differences of the rates on different
chips, and will be screened for local effects.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 53

When performed/frequency:
Commissioning and when dark current is measured.
Required accuracy, constraints:
better than 1 ADU/cm2 /hour
Priority:
desirable
5.3.3 Req. — CCD Linearity
Objective:
Characterize the linearity of the system over the full dynamic range of the
A/D converter.
Both the overall absolute linearity of the system and the pixel-to-pixel vari-
ation in linearity are of interest.
The overall linearity of the system can be obtained by measuring the counts
as a function of exposure time for a series of dome flats. The data to use
for this can be the same as for the measurement of the Gain (req.523, q.v.)
The pixel-to-pixel variation in the linearity is obtained by dividing a flatfield
with a mean exposure level of more than 30000 ADU by a flatfield with an
exposure level of less than 1000 ADU. Pixels that deviate more than 5σ from
the mean, in this divided image, have an anomalously high nonlinearity. This
map of nonlinear pixels may be used in conjunction with the hot and cold
pixel maps to produce a map of bad pixels.
In addition, during a cloudy night, once per year, the linearity will be checked
by taking various exposures with a variety of exposure times on the dome
screen.
When performed/frequency:
daytime- Commissioning, in RP once per month, dark dome test once/year
Required accuracy, constraints:
better than 1% on the photometric scale
Priority:
essential
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 54

5.3.4 Req. — CCD Charge Transfer Efficiency

Objective:
Characterize horizontal and vertical transfer efficiency (CTE) per single trans-
fer (in units of the fraction of the charge actually transferred).
(taken from WFI@2.2m:) Ten flatfields are taken with 50 vertical and 50
horizontal overscan pixels and a mean exposure level of about 20000 ADU’s.
The mean is computed and corrected for the bias. Average signal levels are
determined in the two overscan regions as well in the light sensitive pixels just
preceeding the respective overscan pixels. Any signal found in the overscan
pixels was due to non-unity CTE lost from the neighbouring light-sensitive
pixels. The fractional charge still remaining in the light-sensitive pixels is the
CTE.
When performed/frequency:
daytime- Commissioning, in RP once half year
Required accuracy, constraints:
CTE > 0.999995 per parallel or serial shift.
Priority:
desirable
5.3.5 Req. — CCD Cold Pixels
Objective:
Identify cold pixels.

From a set of 5 low-level flatfield exposures a mean image is computed. This

mean image is smoothed. The smoothed mean is used to flatfield the mean
image. In this flatfielded image, pixels that are smaller than the mean minus
5σ, are taken to be cold pixels. Make sure to differentiate between hot and
cold pixels.
Cold pixel maps are used, together with the hot pixel maps (req.522), to
identify pixels that should be ignored (assigned a weight of 0) in further
processing.
When performed/frequency:
daytime- Commissioning, in RP once per 3 months.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 55

Required accuracy, constraints:

Quality Check: Number of hot pixels to be determined by experience/lab
values. The total number of bad pixels (hot pixels + cold pixels) is less than
80000. Difference in number of cold pixels w.r.t. the previous version, less
than 100.
Priority:
very important
5.3.6 Req. — CCD Hysteresis, strong signal
Objective:
Quantify the effect of CCD signal reminiscence.
Reminiscence of a strong signal (a saturated star) in subsequent observa-
tions (“ghosts”) is a potentially debilitating problem for data reduction and
interpretation. The absence of this effect should be verified by observations
of very bright objects and subsequent dark exposures.
If “sources” are detected in the dark frames at the pixel positions of bright
sources in the first field, signal reminiscence is a problem, which can charac-
terized by the decay time of a strong signal.
The CCD readout mechanism means to cure for reminiscence.
When performed/frequency:
Commissioning.
Priority:
desirable

5.4 Detectors operational specific calibrations

In this Section req.’s are listed which are essential related to the daily operations
of the acquisition of the science data.
For all detector and photometry related calibrations each CCD is characterized
independently of the others. (Only in the case of astrometric solutions, data
of various chips is combined).
A set of calibration lamps together with a dome screen is used to monitor
the health of the instrument and to measure the fine structure of the flatfield.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 56

The calibration lamp system contains two sets of 4 commercial 12-24V halo-
gen lamps each. Each set is operated independently. Each set is stabilized
in current supply (one unit for whole set). Lamps are switched on/off with a
gradual increase/decrease of the current over a timespan of 3-5 minutes. Im-
plementation TBC. When operated this way, the nominal lamp instabilities are
expected to be of the order of 0.05%/hour for a timespan of 200 hours of lamp
operations (private comm. Philips Labs). For a nominal 1-2 hour/day of oper-
ations the lamp instability is expected to be of the order 1.5-3 % per month.
The lamp instabilities are expected to be strictly linear after 100 hours of oper-
ations. When operating two sets at different rates, say one set 1 hour/day the
other set at say 1 hour/two weeks a full characterization of the lamp stability
can be achieved at accuracies better than a few percent on a monthly basis.
Also continuity after failiure of one lamp can be obtained. The accuracy is
better than required as also other factors, like dust on the lamps and back-
ground light will affect the effective illumination of the screen. Altogether, the
system is expected to provide control over the illumination of the screen with
an accuracy better than 5-10%, which will be used for a daily health check
on the overal throughput/health of the detectors (req. 547 Quick detector
responsivity and health check). This activity provides a deliberate redundancy
with flatfield measurements on the dome screen (req. 542 Flat field – dome
key bands–doit), in order to provide the necessary cross-check in the off-line
calibration analysis procedures. Next to the health checks taken during the
night, a standard health check using a photometric standard field (also provid-
ing the absolute photometry is specified in req. 562 Photometric Calibration
– Monitoring). The system of lamps, flatfields measurements, quick checks
using the lamps and health checks on the sky on photometric standard fields is
designed to support the photometric calibration of a Survey System, for many
years to come. The system provides redundancy faciltating cross-checks and
has a typical maintenance/update frequency once/month.
The calibration of science data can be divided in three steps.
1 Removing the effects of bias and differential gain.
2 Relating the overall gain, and hence counts S(x, y) to a photometric scale
3 Relating the x, y coordinates to an astrometric reference system.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 57

The raw science and standard images record S(x, y) counts in pixel x, y, that
are related to the incident photon flux I(x, y, λ) by:
R
S(x, y) = b(x, y) + G g(x, y, λ)I(x, y, λ)dλ,
with G the ADU conversion factor, g(x, y, λ) the quantum efficiency or gain
as function of position and wavelength, and b(x, y) the bias offset.
The photometric and astrometric calibration (steps 2 and 3) are the subject of
Sections 5.5 and 5.6, respectively. Here we list the calibration data necessary to
remove the effects of bias and gain variation over the image. These calibration
data include:
• Bias to subtract residual pattern in the bias offset.
• Flatfields to correct for non-uniform gain.
• Fringe maps to remove the fringe-patterns
• Weight maps to determine the relative contribution of each pixel when
image data are combined,
A first-order approximation of the bias level in an image is provided by the
median of the overscan region. A more accurate determination of the bias
offset takes into account the following two effects: i) the bias level grows to
its asymptotic level in the first few hundred lines, and ii) the bias level depends
on the total signal in a given line. Therefore, an initial bias correction—the
overscan correction, is applied by averaging the overscan pixels for each line,
and subtracting this value from that line. Also, experience with the WFI has
shown the presence of residual patterns in the bias offset over the image area.
Under the assumption that these patterns are also present in OmegaCAM data
its characterization by means of a master bias frame is specified in req.541.
The gain, g(x, y, λ), incorporates the wavelength-dependent pixel-to-pixel vari-
ation in transmissivity of the different lightpaths through the telescope optics,
filters and detectors. The gain can be approximated with
g(x, y, λ) = gDQE (λ)gf f (x, y),
with gf f (x, y) the pixel-to-pixel variation in the gain, and gDQE (λ) the over-
all detector quantum efficiency (zeropoint), at gf f = 1, which is subject to
photometric calibration (Section 5.5).
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 58

The characterization of the pixel-to-pixel variation of the gain, the flatfield, is

obtained by observing a spatially uniform source of illumination. A normalized
version of such an image provides a measure of the relative variation of the gain
over the image area, gf f (x, y). Note that this flatfield measures a combination
in the pixel-to-pixel gain variation and the variation in transmissivity of the
different light paths through the telescope optics and filters.
The ideal flatfield observation is:
i) uniformly illuminated
ii) bright, to minimize errors due to photon noise,
iii) of constant color, preferably a color that is the same as the night sky.
Several methods to determine a flatfield will be operational. However, each
method suffers from different drawbacks. The various characterizations of the
flat field, the dome, twilight, and night sky flats, are specified under the
req.542-545, while the eventual master flatfield to be applied to science and
standard field observations, is constructed from a suitable combination of the
dome, twilight and night sky flat fields (req.546).
Dome flats (req.542) are obtained by observing in telescope screen park po-
sition a fixed domescreen relatively uniformly illuminated by the calibration
lamp with a stabilizing power supply. The illumination is not sufficiently uni-
form to measure large scale variations, and the color is very different from the
night sky.
Twilight flats (req.543) are based on a bright, uniform source of illumination
(twilight sky), that is unfortunately not of constant brightness and color. Unfor-
tunately the ’twilight gradient’ precludes measurement of the largest scale gain
variation. Also bright objects may be visible even at twilight, which provides
an additional complication.
Night sky flats (req.544), obtained by combining a large number of science ob-
servations, most closely mimic the illumination properties of the science frames
themselves. These are the only flatfields usable for measuring the largest scale
gain variations. Unfortunately, the assumption that the illumination is uniform
(except for astronomical sources of course), has proven to be invalid on WFI
(e.g Manfroid et al, 2000). It remains to be seen how large this problem of
”sky concentration” will be for OmegaCAM. Computing the night sky flat is
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 59

also a computationally expensive process, both because of the large numbers of

frames involved, and because the need for a proper masking of bright objects
in the field. Fortunately, this measurement may also be usable for the fringe
correction.
The master flatfield (req.546), to be applied to science and standard field
observations, is constructed from a suitable combination of the dome, twilight
and night sky flat.
The approximation that the pixel-to-pixel variation in the gain is independent
of wavelength is in fact incorrect. Interference effects, mainly in the filters
and thinned silicon layers of the CCDs, introduce wavelength dependent gain
variations that vary on small angular scales. Since most sources are continuum
sources, and only the convolution ∫ g(λ)I(λ)dλ is measured, this effect can be
ignored when measuring source fluxes. However, due to variable strength of
several sky lines, mostly apparent at the long wavelengths, the background will
exhibit so-called fringing patterns, which can change during the night. This
requires an additional calibration step for bands redward of R: the construction
of suitable fringed background images (req.545).
The weight map is an important auxiliary file, which is used in several image
processing steps. The weight map is intimately linked to the flatfields and
therefore its construction is also addressed in this section.
Whenever individual pixels are combined, either in constructing source lists
necessary for photometric and astrometric calibration, or in the coaddition of
different frames, the OmegaCAM-reduction pipeline uses variance weighting
(weight= 1/σ 2 ). The inverse variances are recorded in weight maps.
The debiased images record S(x, y) counts in pixel x, y, that are related to the
photon flux I(x, y) by:
S(x, y) = GgDQE gf f (x, y)I(x, y),
Since photon shot-noise is much larger than the read-out noise, the rms-noise
in the raw data is given by:
σS (x, y) = G(gDQE gf f (x, y)I(x, y))1/2 = (GS(x, y))1/2 .
Once data has been flatfielded (S 0 = S/gf f ), the counts are given by:
S 0 (x, y) = GgDQE I(x, y),
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 60

and the rms-noise by:

σS 0 (x, y) = G(gDQE gf f (x, y)I(x, y))1/2 )/gf f = (GS 0 (x, y)/gf f )1/2
The photon flux I(x, y) is a sum of a uniform background Iback plus sources
Isrc (x, y). Since, the surface brightness of the sky is (much) larger than the
surface brightness of most sources, the rms-noise is given by:
σS 0 (x, y) = G[gDQE Iback /gf f (x, y)]1/2 = (GSback
0 /gf f )1/2
Hence, the rms-noise in an image is the product of a factor ((GSback0 )1/2 ) that
−1/2
is constant over one image, but will vary between images, and a factor (gf f ),
the inverse of the square-root of the flatfield, which varies over the image.
To aid in the construction of weight maps for each individual science image
master weights are constructed (req.546). These master weights are equal to
the master flatfield, except that pixels that are hot (req.522) or cold (req.535),
as well as pixels that have a gain outside a user defined range are assigned a
weight of zero. Individual weight images for each science image can then be
produced by determining the background level Sback , and dividing the master
weight by GSback . These individual weights can be further improved by detect-
ing which pixels are affected by cosmic rays or satellite tracks, and assigning
those a weight of zero too.
5.4.1 Req. — Bias - doit
Objective:
Determine master bias frame.
The signal in raw scientific frames contains a component that is due to a
bias current. This component shows up as an offset to the signal. The
bias-offset has the following characteristics: i) the bias level grows to its
asymptotic level in the first few hundred lines, and ii) the bias level depends
on the total signal in a given line. Therefore, an initial bias correction—the
overscan correction, is applied by averaging the overscan pixels for each
line, and subtracting this value from that line.
In addition, the bias offset exhibits a residual pattern, which is measured by
the master bias frame. To construct the master bias a series of 10 zero-
second bias exposures is overscan-corrected, and then averaged, rejecting 5σ
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 61

outliers (σ = dispersion of the 10 bias exposures of individual pixels), due

to particle hits during read-out. The resulting master bias frames will be
used for the correction of all frames. For each master bias frame the mean
value for each CCD chip will be determined and evaluated in a trend analysis.

As the readout noise dominates the rms scatter in the bias frames, while
the shotnoise of the sky background dominates the rms scatter on the sky
images, which is nominally much larger than the readout noise, it is sufficient
to characterize
√ the bias value at individual pixels with an accuracy of (readout
noise/ 10).
A comparison with a previous master bias frame will be done as an evaluation
of the overall health of the instrument and the quality of the data. This will
thus measure short-term variations. Long term variations can be assessed
using trend analysis.
A comparison of the mean level with laboratory values will be used as an
overall quality check.
When performed/frequency:
daytime- Commissioning, in RP initially daily. Later the frequency is to be
determined by experience.
Required accuracy, constraints:
The required
√ accuracy per pixel in the master bias frame is “nominal read-
outnoise/ 10”.

For the quality check: Deviation of the mean level of master bias (bias level)
from lab values < 10%.
Priority:
essential
5.4.2 Req. — Flat-field - dome key bands + user bands - doit
Objective:
Determine master dome flat frame for both Keybands and Userbands.

During the lifetime of OmegaCAM the dome flatfields shall be measured for
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 62

the 4 keybands and the key-composite filter at least once/week. Thus at

least within 3 days of the taking of science data a dome flatfield in the key
passband will be available.
The sequence of dome flatfields in the keybands, acquired over periods of
months to years will be used to perform a trend analysis on the long term sta-
bility of the instrument and lamp which illuminates the dome flatfield. With
the exception of the effects of the unstability of the lamp, this trend analysis
is redundant with that obtained from both req. 563 photometric zeropoint
and that of req. 562 Photometric calibration - Monitoring. Thus, by com-
bining the results of these req.’s an accurate description of the behaviour of
the lamp is feasible. The prediction of the behaviour of the lamp is a result,
which will be used as an input for req. 546 Quick detector responsivity
check.
The dome flatfields will be used on an individual CCD chip level. The relative
variations of the quantum efficiciency between individual CCD chips will be
measured by req. 563 Photometric Calibration - zeropoint key bands -doit.
The redundancy between various measurements of req. 563 Photometric
Calibration - zeropoint key bands -doit and req. 542 Dome flats at key-
bands will be used to evaluate the relative chip-to-chip gain variations, and
in due time, when advanced insight in this item is achieved, this knowledge
might be used to further optimize observing scenarios.
For the userbands the dome flatfield will only be taken when during the period
of week that particular passband has been used for science observations. No
trend analysis will be done on the data taken in the User pass bands.
When performed/frequency:
Daytime, daily. For the keybands the dome flatfields will be measured at
least once/week. For the Userbands the dome flatfields will be measured at
least within 3 days that the Userband has been used for science observations.
When filters have been changed in the cassette the presence of dust and/or
scratches might require new flat field exposures.
Required accuracy, constraints:
Accurately measuring pixel-to-pixel gain variations as small as 1%, Adding 5
exposures of 20,000 counts satisfies this requirement.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 63

Priority:
essential
5.4.3 Req. — Flat-field - twilight
Objective:
Determine master twilightflat frame, using observations of the twilight sky.
Twilightflat observations will be attempted for each passband that is ob-
served during that night. If insufficient twilight time is available then the
twilightflat observations are taken preferably in the previous or subsequent
night. In addition twilightflats of the 4 keybands will be taken at least once
a week.
In order to minimize the spatial gradient in the sky brightness, the observa-
tions need to be made on the solar circle, i.e. the great circle through the
zenith and the sun’s position, at a zenith distance of about 20◦ in the solar
antidirection. Preferably, the field of view does not include stars brighter
than TBD magnitude.
When performed/frequency:
Evening and morning twilight.
An attempt will be made to observe twilightflats for all bands observed during
the night. For all observed bands, twilightflats will be observed within a
maximum of 2 nights. In addition, twilightflats for the keybands will be
obtained at least once a week, irrespective of whether keybands were used
for science observations during that week.
Required accuracy, constraints:
Mean levels should be approximately 20000 ADU.
Priority:
essential
5.4.4 Req. — Flat-field - night sky
Objective:
Create Night Sky flat frame.

The flatfield that most closely reproduces the actual gain variation of the sci-
ence and standard observation, can be obtained by averaging a large number
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 64

of science and standard observations, taking care of properly masking the

contaminating object. While such a night-sky flat (aka supersky) could, in
principle, be superior in quality to the twilight flat, the procedure to obtain
this flat can be computationally very expensive. On the other hand night-sky
flats may also be suitable for fringe removal.
It remains to be seen to what extent the problem of sky concentration,
i.e. nonuniform illumination due to stray light/reflection affects the quality
of night-sky flat.
Because these night-sky flats can only be obtained from actual observations,
we cannot guarantee their availability for bands for which only standards were
obtained. It is, therefore, not clear how routine building of the skyframes
should be incorporated into our photometric calibration scheme.
A minimum of 5 images in a night in a given band is required to optimally
fulfill this requirement.
When performed/frequency:
daytime, daily
Required accuracy, constraints:
This procedure would benefit from a prior detection and masking of bright
objects. If no masks of bright objects are available, then a minimum of 15
frames should be included.
Priority:
very desirable
5.4.5 Req. — Flat-field - Fringing
Objective:
Determine the fringe pattern of the background.
Fringing due to variable strength of several skylines, mostly apparent at the
long wavelengths, requires a different approach to background subtraction.
Normally, after flatfielding, the background can be expected to be flat over
the entire image, and a median of the image, excluding 5σ outliers, would
in principle be sufficient to subtract the background.
In images that suffer from fringing we have to deal with a background that
is variable on small (<< 10 ) scales within the image, and can not be dis-
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 65

tinguished from sources. The image itself can, therefore, not be used to
determine the background. However, given the fact that most observations
are taken in jitter or dither mode, the information of several images can be
combined to determine a background. This average should include enough
observations to properly exclude contamination from sources, and, because
the standard jitter/dither patterns only include 5 pointings, one background
computation per jitter/dither is probably not sufficiently accurate. On the
other hand, because the fringing pattern varies with time and telescope po-
sition, a straight mean (the supersky) over an entire nights worth of data is
also not usable.
A suitable strategy to construct a fringed background image, usable for sub-
traction, thereby removing the fringe pattern, remains to be determined. If
the fringe pattern is stable over the night, a decomposition of the night-sky
flat in an additive and multiplicative term is feasable. The assumption that
the high-frequency spatial component in the night-sky flat are fringes, while
the lowest frequency components represent gain variations has been used
with reasonable success.
When performed/frequency:
Commissioning and when long wave science frames are taken.
Priority:
very important
5.4.6 Req. — Flat-field - master flat and weight map
Objective:
Determine the master flatfield, to be used to correct for the pixel-to-pixel
gain variation from the raw image data. Also use this flatfield to create a
master weight map, to be used when co-adding image data.
Four different measures of the variation in the gain are available: the dome
flat (req.542), the twilight flat (req.543), the night-sky flat (req.544) and
the illumination correction (req.548). A suitable choice of the final master
flatfield, based on a combination of one or more of these flatfields, and,
optionally, the illumination correction will
A method whereby the dome flat is used to measure the pixel-to-pixel (small-
scale) variation, and either the twilight or night-sky flat is used to measure
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 66

the large scale variation, would provide a first-order approximation of the

master flatfield. This master flatfield could then be used to flatfield the
science and standard images.
Experience at FORS has shown that a suitable combination of twilight and
night-sky flats provided the best determination of the gain variation. Experi-
ence with WFI indicates that some care has to be taken to address the issue
of sky concentration. An optimal algorithm that takes this into account,
(using the illumination correction), will be based on experiments with WFI
and OmegaCAM data.
The master flatfield is proportional to the inverse variance in the flatfielded
data and can therefore be used to build a master weight image. Weights
of zero are assigned to hot (req.522) and cold pixels (req.535), as well as
pixels that have a relative gain outside a user defined range.
The master weight maps are the basis on which individual weight maps are
created (seq.– 632).These individual weight images also assign a weight of
zero to cosmic-ray events and satellite tracks.
When performed/frequency:
New weight and flag images should be constructed whenever a new flatfield
is constructed.
Priority:
essential
5.4.7 Req. — Quick detector responsivity check - doit
Objective:
Quickly check the overall health in terms of responsivity by observing the
dome screen with the composite filter.
Together with req. 521 read-noise this item forms the most important
day-to-day health check. The expected lamp intensity is characterized in
req. 542 Dome flat. This measurement will lead to a go/non-conformance
flag and day report. The results will have to be inspected on the site, as this
is a daytime health check of the instrument.
Trend analysis on the raw data will be redundant with that of req. 542 dome
flat.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 67

The equivalent of this req. on the sky is provided by req. 562 Photometric
Calibration - Monitoring
When performed/frequency:
Commissioning, daytime, every day of operations both during CP and RP.
Required accuracy, constraints:
1%
Priority:
very important
5.4.8 Req. — Illumination correction
Objective:
Characterize the illumination correction.
The zeropoint is determined individually for each CCD in req.563. The
gain variation over individual chips is characterized by the twighlight and
sky flatfields (req.533 and req.534) under the assumption of an ideal flat
illumination over the field of view. In practice this ideal flat illumination
can be affected by stray light (sky concentration) and the flatfield has to be
corrected for this.
An initial verification that this effect is indeed present will be obtained when
constructin g catalogues of secondary standards (req.569). In case it is
found to have an amplitude over a single chip larger than 1% the effect has
to be characterized by measurements of a standard field. The master flatfield
(req.546) will apply this information when needed.
When performed/frequency:
Verification of effect during commissioning. Measurement, during RP, once/month
Required accuracy, constraints:
better than 1% for the amplitude over a single CCD.
Priority:
essential
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 68

5.5 Astrometric Calibration

The aim of the astrometric calibration is to accurately determine the projection
matrix for converting pixel positions to sky coordinates. This will be done
automatically in the data reduction pipeline.
For Mode– SSO N= the astrometric solution can be derived provided some
observational limits, restricting the curvature of the stellar track, are met. When
stellar tracks become more curved than the limits, all astrometric quality checks
will have to be relaxed. All source positions will be derived using the first-
order moments of the pixel distribution of the detected objects. The resulting
positions, in such cases, will be the weighted averages of the stellar tracks and
that of the (than not) moving targets. The exposure time limitation is defined
by the curvature of the stellar tracks in the field, which should be less than
1% over the full track. Estimates for limitations on tracking for OmegaCAM
are defined by requiring the curvature of the stellar tracks to be less than 0.1
FWHM of the PSF (< 0”6). Assuming tracking in Ra, a track length of less
than 0.1 degree at any declination is allowed. For a track length of 0.2 degree
only declinations between 0 and +/-20 deg or +/-70 to +/-90 deg are within
the above limits. For the other modes there are no such restrictions.
The astrometric calibration can be derived using two fundamentally different
methods.
The first, traditional, method is to derive the full projection matrix separately for
each pointing, ignoring any detailed prior knowledge (apart from some rough
initial estimates such as pointing, orientation and platescale). This method
has great freedom and allows for instrument independent determination of the
projection matrix. It can thus be used on a variety of instruments, but requires
more input data elements to achieve an accuracy equal to that of the next
method.
The second method allows for a separate determination—and use—of instru-
ment specific characteristics of the geometry that do not change (much) among
different pointings and the characteristics that do change with pointing. Fixed
characteristics are, e.g., position of the chips relative to each other, the po-
sition of the rotator axis, the optical deformation at the position of the focal
plane and perhaps pointing accuracy. Variable characteristics include items like
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 69

the flexure of the telecope and its instrument. The latter items need to be
determined for each pointing separately. This method can be robust and has
less degrees of freedom facilitating accurate astrometry with a small amount
of data points. It is, however, tightly fixed to the instrumental geometry. A
geometric model must be obtained and associated parameter values must be
determined before standard astrometric reduction can be done.
Both methods will make use of an astrometric reference catalog. Current
catalogs have positional accuracies that are not extremely high on the scale of
the field of view of OmegaCAM. To achieve higher positional accuracy within
a given pointing set and to allow accurate co-addition without degrading the
PSF the astrometric solution will make use of the overlap among the pointing
set.
The methods can be applied both to the main camera CCD chips and to the
guide CCDs, as the latter can be viewed as a auxiliary CCDs of the main camera.
The requirements outlined here determine the calibration data necessary to
perform an astrometric solution using prior knowledge of instrument specific
characteristics.
The algorithm for the derivation of the astrometric solution is detailed in Section
7.3, while its use is detailed in seq.– 634.
5.5.1 Req. — Position of Camera in focal plane
Objective:
Determine the position of the chips with respect to the rotator axis of the
telescope. This is part of the static astrometric calibration of the camera. It
involves the determination of the chip position, scale, and orientation with
respect to a perfect pixel plane. This has to be done with the ADC in and
out.
This procedure produces the astrometric pre-solution. In fact, the expected
pointing and other a-priori positional offsets are expected to be small; hence
the standard astrometric solution can already be obtained without a pre-
solution. However, a first inspection and verification of the pre-solution is a
task to initialize the system.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 70

When performed/frequency:
Each mechanical change of the camera. Each user supplied filter, once a
year.
Required accuracy, constraints:
Internal precision: 0.3 pixel. External precision limited by reference catalog
Priority:
Desirable
5.5.2 Req. — Telescope Pointing and offsetting
Objective:
Verify the pointing and the offsetting of the telescope for both optical con-
figurations (ADC in and out).
The pointing model is provided independent of the OmegaCAM S/W, but a
verification of both the pointing and the offsetting accuracy is required.
Perform on-site spot checks of the pointing model. The data from the Guider
CCD can be used for this.
Also in the data reduction pipeline, as a standard check in the astrometric
solution, the pointing error is determined.
When performed/frequency:
Commissioning, after each change of the pointing model, and to be deter-
mined by experience.
Required accuracy, constraints:
1 arc second
Priority:
Very important
5.5.3 Req. — Telescope and Field Rotator tracking
Objective:
Verify that the rotator performs properly and simultaneously that the tele-
scope is tracking correctly.
Up to Zenith distances of 60 degrees and wind speed of 18 m/s with a
dynamical component of 30%, the free tracking of the telescope shall be
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 71

better than 0.2 arcsec rms. With closed-loop autoguiding, the rms deviation
shall not exceed 0.05 arcsec
Two methods will fulfill this requirement.
First, check at various telescope positions the global performance of the
rotator plate which is driven by the pointing model. When the rotator plate
is not performing optimal, the objects are elongated in a circular pattern
(concentric rings) with the rotator plate axis at the center. This inspection
is closely related to the determination of the point spread function.
Second, check for the tracking of the telescope to find functional dependency
with telescope position. This is purely a verification. As an internal check,
do this for each OmegaCAM observing mode. Similar to the rotator plate
inspection the point spread function is used as the measuring tool. When
the telescope is not tracking correctly the shapes of stellar objects are sys-
tematically elongated. The amount of elongation may not exceed a certain
value, corresponding to the basic tracking requirements given above.
These two functional checks are merged into one analysis as they essentially
use the same technique for verification and because they are coupled. Non-
conform tracking can cause non-conform rotation.
The offset information from the Guider CCD’s is another element in the
functional check. When offsets for guide stars are becoming too large, rotator
plate errors or telescope tracking errors are apparent as different patterns of
offsets during an exposure time.
When performed/frequency:
Commissioning, at each change of the pointing model, in RP to be deter-
mined by experience. In CP check once with the ADC in and out.
Required accuracy, constraints:
First method: 1 arcsecond in the dependencies. Second method 0.1 pixel
dimension or 0.1 FWHM
Priority:
very important
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 72

5.5.4 Req. — PSF Anisotropy

Objective:
Determine the PSF anisotropy.
Detailed characterization of the Point Spread Function at various positions
in the focal plane shall be provided. Monitor optical defects and possible
variations in time. Do this for both optical configurations of the telescope,
ADC in and out.
Guide CCD recordings are used for the analysis.
When performed/frequency:
Commissioning
Each optical change to the telescope,
Remount of the detector assembly. Once per three months.
Required accuracy, constraints:
better than 1%
Priority:
desirable
5.5.5 Req. — The astrometric solution for templates - doit -see 6.3.4
5.5.6 Req. — The astrometric solution for Guide CCD’s
Objective:
Perform astrometric solutions for the Guide CCD’s and hand over the solution
to the Instrument S/W for locating Guide stars.
Note, the Guide CCDs can be read out separately and ’stand alone’.
When performed/frequency:
Commissioning.
Required accuracy, constraints:
1 arcsec rms for the accuracy with respect to the external standard;
External precision is driven by the position reference catalog. This is in the
case of the USNO-A2 catalog of the order 0.3” with possible systematic
excursions to 1”.
Priority:
Essential
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 73

5.6 Photometric Calibration

The basic requirement for the photometric calibration of the broad-band filters
is to achieve an accuracy of better than 5% on the photometric scale in ‘instru-
mental magnitudes’ as assigned to the units of the resultant output image of
the “image pipeline”. The accuracy of the colour transformation terms of in-
strumental to standard systems should be better than 10% on the photometric
scale.
In order to maintain this accuracy on a routine basis over years of operation, a
set of requirements are specified in this section.
The descriptions of these requirements involve the following OmegaCAM spe-
cific ingredients:
• key passbands (X = u’,B, V and i’ in Johnson (B,V) and Sloan (u’, i’)
system)
• two lens correctors (near Zenith, the baseline, key configuration) and an
atmospheric dispersion corrector -ADC for operations in User mode at
larger Zenith angles.
• a composite key filter (X = u’, B,V and i’ in each quadrant),
• a standard polar field, observable throughout the year,
• 8 equatorial fields, containing both primary and secondary standard stars
(Landolt fields - see section 7.1)
• a dome lamp and a fixed dome screen equipped with a stabilized current
supply,
• 32 CCD’s are operated simultaneously, with the exception of the composite
filter which ‘feeds’ 8 CCDs simultaneously in one passband.
• data rates should stay within limits that allow processing and storing of the
data with the currently anticipated technology.
• A standard atmospheric extinction curve is adopted and all atmospheric
extinction in various passbands is taken as a scaling of this curve.
• The photometric monitoring employs observing strategy – freq which has
overriding priority on the scheduling and which employs observing mode –
stare and its associated trend analysis.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 74

The prime concept of the OmegaCAM photometric calibration is to continu-

ously maintain the photometric scale in the keybands, even when the science
programme does not require the usage of these passbands during a particular
night or period. This continuity is used by the data reduction (calibration and
its trend analysis) and is meant to ease the maintenance of the photometric
system on a routine basis.
The usage of a standard extinction curve results into a high rigidity of the
pipeline processing, and provides a tool for error estimates, quality checks,
recognizing non-conform data and provides a framework for successful pipeline
processing of incomplete data.
We model the characterization of the photometric system in terms of a series
of gains, where for each aspect of the calibration we distinguish a gain g0 at a
pre-determined fixed moment and the variation of that gain as function of time
g(t), the latter being mostly analyzed by a trend analysis.
All photometry is determined on an individual CCD chip basis (N = number
of chip, 1...32), apart from the atmospheric extinction which is common for all
chips.
Most gains depend on passband (X), but not the variation of the atmospheric
extinction, which is assumed to scale with the standard extinction curve.
In Figure 5.6 an overview is sketched of the various requirements which form
the photometric calibration.
The fixed and variable gains of the various calibrations are defined as follows:
Atmospheric extinction: use scaling of standard extinction curve represented
by gsel.e (X) :
middle of the night: ge (0) × gsel.e (X) req.563
during the night : ge (t), t in hours req.562

Zeropoint effective DQE - req.563

middle of the night: gDQE (0, N, X)
on different nights: gDQE (t, N, X), t in nights

Flat field - req.542 and others

for period of 7 days: gf f (week, N, X)
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 75

different weeks: gf f (t, N, X) t in weeks ; every week a new ff is assigned,

replacing gf f (week, N, X)
For a given star observed in passband X at a given position in the field of view,
at a given moment (t) the relation between the output (Iobs of the detectors)
and the zero-airmass intensity is given by the general expression:

Iobs,X = ge (0)ge (t)gsel.e (X)×gdqe (0, N, X)gdqe (t, N, X)×gf f (week, N, X)×Iref,X

Colour term:
The primary standard stars have been measured with presumably the same filter
passbands, but with CCD detectors which have a different relative spectral
responsivity. This implies that the effective gDQE (0, N, X) when observing
these primary stars depends on the colour of these stars: gDQE (0, N, X, X −x).
The photometric calibration involves the solution of the general equation above
along a different path, with different unknowns, and different knowns.
The initialization of photometric calibration is to carefully process backwards
and forwards through the basic equation. Particularly the settling of the sec-
ondary standards, which cover a larger field of view than the primary standards
is tricky. On one hand the preparatory programme will provide this informa-
tion, on the other hand OmegaCAM calibrations can self-calibrate the secondary
standards, which has the advantage that it avoids the extra bootstrapping via
another telescope and detector system.
Normalizations:
gf f (week, N ) unity at central pixel of each chip N
ge (0) and ge (t) multiplication factor of standard extinction curve represented
by gsel.e (X)
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 76

Monitoring the Photometric Calibration

1 Week
Requirement Applied filters Field 1/3Night 1 Night Run 1 Month 1 Year Years
extinction
5.6.2 Monitoring u’ V
B i’
Polar ❈ clouds ❈ ❈ ❈ ❈ ❈

5.6.3 Zeropoint u’ B V i’ u’ V
Eq. 1 (❈) ❈ ❈ ❈
B i’
Eq. 2 ❈ ❈

per chip!
zeropoint
Eq. 3 ❈ ❈
Eq. 8 ❈ ❈

5.6.4 Zeropoint υυυ

i j k υ ?
Eq. 1
Eq. 2
❁
Eq. 3

Eq. 8 ❁

5.6.5 User key u’ B V i’

u’ V
B i’
Eq. n

υυυ
i j k υ ?
❁

5.4.2 Flat−field u’ B V i’ u’ V
B i’
Dome ❈ ❈ ❈ ❈

υυυ
i j k υ ?
❁
lamp

5.4.7 Quick check u’ V Dome ❈ check ❈

B i’

Fig 5.6. Overview of tracking the photometric calibration at various time

scales. From left to right, the figure indicates the requirement number, the
requirement name, the used filters and standard fields. The stars indicate
at which frequency the measurement is done. For further details, see text.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 77

5.6.1 Req. — Shutter Timing

Objective:
Verify the actual timing of the shutter.
Exposure timing has to be accurate to ±0.2% at 1 sec exposure time at any
position of the focal plane (OmegaCAM Technical Specifications). The ex-
posure timing signal is provided by PULPO. For the following considerations
we assume that the PULPO timing signal is accurate to better than 0.1 msec
(i.e. it is not the dominating source of inaccuracy).
The shutter mechanism consists of a pair of chasing carbon fibre blades.
Their movement is controlled by the Shutter Control Unit (Shutter CU)
such that it results in an identical effective exposure time all over the frame.
These two movements may or may not overlap in time depending on the
exposure time and the blade traveling time.
The opening blade starts moving immediately (µsec’s) after the falling edge
of the TTL signal (provided by PULPO). This is the beginning of the exposure
procedure. The closing blade starts moving immediately (µsec’s) after the
rising edge signal was detected and ends (about 1 sec later) when the closing
blade completely covers the aperture, which marks the end of the exposure
procedure. Therefore, the duration of an exposure procedure is always:
exposure time + blade travel time (ca. 1 sec)
Two types of delays affect the effective exposure time: The delays of the
start of the blade movements after the opening/closing TTL signal edge
(i.e. absolute exposure time) and position dependent delays during blade
movement (i.e. exposure homogeneity).
The open/close delays are up to 0.05 msecs due to signal polling of the
Shutter Control Unit software. These values are well within the requirements
(shutter open time error: ±0.2% at 1 sec corresponding to ±2 msecs).
Deviations from this occur only in case of a severe shutter failure which is
detected by the Shutter CU and PULPO independently followed by operator
actions.
Position dependent delays (requirement: 0.2% at 1 sec exposure time) will
be monitored in regular intervals of 3 months.
Dome flatfields of 10 sec and 0.1 sec exposure time will be taken for both
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 78

shutter blade movement directions. Illumination level shall be such that the
CCD’s are at about 60% to 80% full well for the 10 sec exposure. Exposure
times will have to be evaluated during Commissioning.
When performed/frequency:
Commissioning, once per 3 month, further to determined by experience.
Daytime
Required accuracy, constraints:
Timing error less than 0.2%.
Priority:
desirable
5.6.2 Req. — Photometric Calibration - monitoring
Objective:
Monitor any short term variability related to the transparency of the at-
mosphere (atmospheric extinction) or due to instrumental instabilities (e.g.
effective DQE) with a minimum sampling of at least 3 times/night. This
provides a daily overall health check of the instrument and detectors. A
further trend analysis has to provide information on long term stability.
The variation (r.m.s.) of the flux detected by the autoguider shall also be
used as an indicator (put in the FITS header) of the sky conditions. This is
to be done for each science observation.
This monitoring is done on a standard polar field, which will be repeatedly
observed at the beginning, middle and end of the night with the compos-
ite key filter (u’, B, V and i’ band), irrespective of the passbands used for
the science observations. The observations are done in the standard config-
uration, with the two-lens corrector. A direct comparison of the measured
intensity of the stars with reference values is used to qualify the overal condi-
tions of instrument and atmosphere, the actual zeropoint (both unit airmass
extinction and instrument DQE) being determined by req.563. The com-
posite filter will provide simultaneous measurements of the sky brightness in
these bands, thus providing an accurate spectrum with a spectral resolution
R of roughly 5.
The comparison of the observed signal with the expected signal from stan-
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 79

dard stars in each of the four quadrants will lead to the determination of
the product of the atmospheric extinction (= ge (0) × ge (t) × gst.e (X), (with
gst.e (X) the gain of the standard extinction curve at passband X ) and the
overall effective DQE (gDQE ) of the detector system including the optics.
As req.563 solves both gDQE and gst.e (X) × ge (0), a comparison with these
measures gives ge (t), the variation of the overall gain during the night. The
thus derived values of ge (t) at t = beginning, middle and end of the night
(could be more if the observer so wishes) provides the required monitoring.
Excursions from the standard extinction curve, due to extraordinary meteo-
rological conditions, can be traced by computing the standard deviation of
observed minus standard curve values in the various bands.
The sky spectrum shall be derived on line from the data, as a quality check on
the health of the instrument and the clearness of the atmosphere, as clouds
or cirrus will be immediately notable in the spectral shape. A reference table
containing the expected sky brightness (and thus colour) as function of lunar
phase will be used in the evaluation of the data.
The repetitive measurements on the same field, with the same filter will also
be used in trend analysis to monitor the overall long term stability of the
instrumentation and atmosphere. The redundancy of these measurements
with req. 563 zeropoint and req. 542 Flat field -dome will be used as a
cross-check on the validity of the photometric system.
In addition, the repetitive nature of these observations, make them ideally
suited for the health check as specified in requirement 3.7.3.4 of the Technical
Specification. In order to facilitate rapid analysis at Paranal an additional fast
analysis recipe is proposed, based on a predefined catalog of standard objects
in each CCD. This recipe assumes that the pointing accuracy, achieved when
guiding, is better than 6 arcsec. The recipe identifies regions of interest in
each CCD for each object in the input catalog. These regions are then
extracted after which a simple moment analysis of the brightest object in
the region will determine total flux in ADU, position and FWHM for each
object of interest, as well as the background flux in ADU.
Fulfilling or fulfilled by:
Selfstanding; a corresponding requirement on detector level is
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 80

req. 546 Quick check

When performed/frequency:
beginning, middle and end of night; any moment atmospheric conditions are
suspect. QC1, CA
Required accuracy, constraints:
all photometry better than 1-2% on the photometric scale; the zeropoint of
the night should be determined first req.563.
Priority:
night monitoring: essential, long term trendanalysis: important
5.6.3 Req. — Photometric Calibration - zeropoint keybands - doit
Objective:
Determine the zeropoint of the overall detector chain, separately for each
CCD chip, in all four keybands (no composite filter) and the true atmospheric
extinction at midnight by measuring standard stars in the 4 key passbands
in one of the eight equatorial fields and the polar field. Do this every night
whatever the science programme on the telescope may be. Optionally, add
one observation of another equatorial standard field with higher airmass to
obtain a redundant, classical, measurement of the atmospheric extinction.
The keybands plus the two-lens corrector form the standard for this require-
ment, the use of the ADC is considered as User mode (see req.564).
The zeropoint corresponding to the DQE of each of the 32 CCD chips will be
determined on an individual chip basis gDQE (0, N, X). Thus the composite
filter can not be used for this. However, additional data will be acquired
with the composite filter for redundacy. In case the relative gain variations
of individual CCD chips appear small and well characterized by the overall
flatfield (which is not really expected) then the req. might be fulfilled with
only the composite filter, substantially releaving the data rate and workload.
The combination of the data of req. 562 Monitoring taken at the middle of
the night with the present zeropoint data will be used to solve separately for
the effect of the extinction and DQE at the middle of the night. A standard
extinction curve will be used as a reference, for error analysis, and to support
the derivation of the solutions in the pipeline processing.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 81

As primary standards the Landolt equatorial fields will be used, possibly

extended with the WFI@2.2m preparatory programme results (Section 7 of
CP). For each of the 8 equatorial fields a solution for secondary standard
stars will be made for a larger, one degree, field of view. The acquisition of
catalogues of secondary standards is discussed in req.569. These secondary
standards data will be used for the nightly determination of the zeropoint.
During commissioning the reproducibility of the zeropoint determination
should be verified for the different observing modes Mode– jitter N=5 and
Mode– dither N=5. This also serves as an end-to-end-test.
Note that the fast extraction procedure defined in req.562 can also be
adapted to these standard star observations of the equatorial fields.
When performed/frequency:
Once in the middle of each night. The linking of external (primary) to
internal (secondary) standards will be done once during commissioning, and
the first year of operation for each standard field; after that to be determined
by experience.
It is to be determined by experience (stability of the system) whether the
nightly zeropoint measurements can be relaxed as follows: in case the keyfil-
ters are not used for science observations in a particular night, it is sufficient
to only take an exposure with the composite filter.
During commissioning the Mode– dither N= and Mode– jitter N= have
to be verified.
Required accuracy, constraints:
1% on the photometric scale
Priority:
essential
5.6.4 Req. — Photometric Calibration - zeropoint user bands
Objective:
Determine the zeropoint of the overall detector chain and the atmosphere
by measuring standard stars in the User passbands.
The zeropoints of the photometric calibration in the User bands will only be
determined for the nights that the User bands are actually used for scientific
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 82

observations. The measurements will be done on one of the eight equatorial

fields.
The atmospheric extinction will be determined in the keybands through
req. 562 monitoring on the polar field. These data are appended here
(like req.563) with a composite key exposure at the equatorial field at mid-
night, to settle ge (0, uBV i). The extinction curve will be used to transform
measured atmospheric extinction at the keybands to the User bands.
The transformation (colour term) of the user passband to the key passbands
is determined for a limited set of filters and the ADC with the keybands in
req.565.
By combining the extinction results in keybands, the passband transformation
coefficients and the direct zeropoint measurements in the User bands the
zeropoint corresponding to the DQE of each of the 32 CCD chips will be
determined on an individual chip basis.
Trend analysis on these data is not required. The instrumental magnitudes
of standard stars in each of the userbands will not be solved.
When performed/frequency:
Once in the middle of each night.
Required accuracy, constraints:
2% on the photometric scale for broad bands and 5% for narrow band filters.
Priority:
essential
5.6.5 Req. — Filter band passes - user bands vs key bands
Objective:
Characterize the transformation coefficients, including the colour term for
the OmegaCAM user passbands to the OmegaCAM key passband.
The standard keybands are calibrated in req.563 with the two lens corrector;
the characterization of the ADC at the keybands and its transformation to
the standard configuration is part of the present requirement.
When performed/frequency:
Once commissioning
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 83

Required accuracy, constraints:

10% on the photometric scale (formal spec) and 2% (goal) for broad band
filters
Priority:
very important
5.6.6 Req. — Dependency on rotator angle/reproducibility
Objective:
Verify the dependency of the photometric calibration on the angle of the
field rotator.
Measure dome flatfield at 12 field rotator angles.
Measure the polar field with 12 field rotator angles.
This also verifies the reproducibility and provides an end-to-end test.
When performed/frequency:
Commissioning
Required accuracy, constraints:
1% on the photometric scale
Priority:
desirable
5.6.7 Req. — Linearity (as a function of flux)
Objective:
Verify the linearity (ratio of input over output) of the overall detector amplification-
data reduction chain for the three different observing modes as an end-to-end
test.
Compare the resultant magnitudes derived by the image pipeline by taking
short and long exposures of the same standard field.
When performed/frequency:
Commissioning
Required accuracy, constraints:
Better than 1% on the photometric scale
Priority:
desirable
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 84

5.6.8 Req. — Detection limit and ETC calibrations

Objective:
Verify the effective detection limit/ overall througput of VST + camera and
subsequently the parameter values used for the Exposure Time Calculator.
When performed/frequency:
Commissioning
Required accuracy, constraints:
10 % in detection limit
Priority:
essential
5.6.9 Req. — Secondary Standards
Objective:
Build catalogs of secondary photometric standards, by observing Landaolt
fields, centered on each individual CCD.
A fundamental concept in the calibration of OmegaCAM data is separate
photometric calibration for each CCD. To obtain this calibration with a single
observation this requires photometric standards covering the entire FOV of
the instrument. There are currently no catalogs of photometric standard stars
satisfying this requirement. Hence, obtaining such catalogs of secondary
standards for the equatorial fields and the polar field will be part of the
calibration observations to be performed during operations.
Obtaining the necessary observations of secondary standards is a time-consum-
ing operation (see below). Moreover, because these observations should
cover all 8 equatorial fields (ref seq.– 563),approximately two (bright) nights
each month will have to be reserved for these observations, at least in the
first year of operations.
The set of observations of secondary standards will also be used to determine
the illumination correction (req.548). Note, that this constitutes a bootstrap
problem, because determining accurate zeropoints of the secondary standards
requires that the illumination correction is already known.
Because of this bootstrap problem, a straightforward determination of the
zeropoint requires that for each CCD the zeropoint and illumination cor-
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 85

rection are determined simultaneously using primary standards. Therefore,

obtaining measurements of secondary standards requires a sequence of obser-
vations that positions the primary standards in each CCD, i.e. 32 pointings
per field, per filter. Techniques for reducing this observational burden are
under investigation.
The order of priority of determining standards is first the key bands followed
by the composite filter.
When performed/frequency:
Commissioning, 2 nights of bright time, each month in the first year.
Required accuracy, constraints:
0.02 mag in individual secondary standards stars.
Priority:
essential

5.7 Internal alignments, optics etc

5.7.1 Req. — Camera focus/tilt

Objective:
Determine and verify the camera focus.
The tilt of the detector plane with respect to the focal plane and its depen-
dency on the orientation of the telescope shall be determined both from:

CalFile– 554 PSF anisotropy

and from the matrix of PSF;s provided by the present requirement.
Verify once for each filter that they have the same optical thickness (15mm
physical thickness). Do this by measuring the ”filter focus offset”.
When performed/frequency:
CP, also filter thickness only once during comissioning
Priority:
essential
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 86

5.7.2 Req. — Ghosts - ADC

Objective:
Verify the absence/presence of ghosts.
For each available filter inspect the presence of ghosts by making several
exposures near a very bright star at various angular distances from the field
center. Do this for both correctors.
The inspection will be done in first instance visually on the RTD.
Off-line the images will be fed to the standard pipeline for closer inspections.
When performed/frequency:
CP
Required accuracy, constraints:
Priority:
desirable

5.8 Effect of Telescope

The various effects of the telescope on the quality of the images produced by
the camera have been addressed by the following req.’s:

req. 524 Electromagnetic compatibility

req. 551 Position of camera in focal plane
req. 552 Telescope pointing
req. 553 Telescope and rotator tracking
req. 554 PSF anisotropy
req. 566 Dependency on rotation angle - field rotator, ADC
req. 572 Ghosts

This list appears rather complete and no additional requirements are specified.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 87

5.9 Workhorses and End to end tests

WORKHORSES
The following ‘work horses’ or ‘doit’ requirements are specified:

req. 521 CCD read noise - doit

req. 531 CCD Dark Current - doit
req. 541 Bias - doit
req. 547 Quick detector responsivity check - doit
req. 555 The astrometric solution for templates - doit
req. 563 Photometric Calibration - zeropoint keybands - doit

The following end-to-end tests (i.e. observational data which employ many
different aspects of the system and which can be used to trace reproducibility)
have been specified:

req. 566 Dependency on rotation angle - ADC, rotator/ reproducibility

req. 567 Linearity (as a function of flux)
req. 562 Photometric calibration - Monitoring/ Health check

5.10 On-site quick look analysis

Requirements for the Real Time Display, essentially requirements on how to
perform visual health checks on the acquired data, are given in the Instrument
S/W User requirement document.
Here req.’s are listed which require analysis on the site.
The first list gives the requirements for which on-site analysis is essential (listed
in order of priority).

req. 562 *Photometric calibration- Monitoring

req. 547 *Quick detector responsivity check - doit
req. 521 *CCD read noise - doit
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 88

req. 531 *CCD Dark Current - doit

The second list gives the requirements for which on-site analysis is desir-
able/most practical (listed in order of priority). On-site, these activities will
output go/no-go flags.

req. 571 Camera focus/tilt

req. 552 Telescope pointing
req. 553 Telescope and rotator tracking
req. 566 Dependency on rotation angle - field rotator, ADC
req. 524 Electromagnetic compatibility
req. 551 *Position of camera in focal plane
req. 554 *PSF anisotropy
req. 572 Ghosts

In the lists above the req.’s which produce a CalFile– are marked with a *.
These req.’s have also to be processed off-line at ESO HQ (e.g. DFS pipeline,
DFS operations, calibration pipeline, QC1, Quality control, trendanalysis).
DFS-pipeline modules, extracted from the off-line ESO HQ version, (including
thosee used for calibration pipeline and QC1) could fulfill these task on-site
with relative limited extra effort. Such modules will run with a stripped Calfile-
date base As a desirable side effect, this creates the possibility to also quickly
verify any other req.’s with extracted DFS-pipeline modules in the case of
un-expected events.
The filling of the calibration database (CalFile– )should however be exclusively
handled by the off-line pipeline at HQ.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 89

Fig 6. Data Flow of Science and Calibration data

relational db Sybase RECIPE RESULTS
ESO-archive DFS operation Calibration archive

calibration daily CalFile- 521, 547, 562, 563, 564

ESO-Archive
Sybase calibration weekly CalFile- 523, 531, 541, 546

raw-cal data calibration monthly CalFile- 533, 535

8 Tb/yr
calibration yearly CalFile- 535, 551, 561, 565

Seq631- Initialisation
ESO-Archive
Sybase Seq632- bias + flat f
raw science data Seq633- weights
20-30 Tb/yr
Seq634- astrometry
Seq635- photometry 1 Tb/yr
SeqFile 631 632 633 634 635 636
Seq636- Co-addition only descriptors
calibrated /co-added image

MD
15 Tb/yr ESO - D
MD
ESO - D A CENT
ERS
/ DAT
USERS
Source extraction Source lists
TERS
/ DATA CEN 3-5 Tb/yr
USERS
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 90

6 DATA REDUCTION SOFTWARE SPECIFICS

6.1 Data reduction software requirements

This section is printed both in the User requirement document (URD) and
the data reduction specification document (DRS). It is meant to summarize
the general framework, requirements and limitations of the OmegaCAM data
reduction.
The MoU refers to the OmegaCAM pipeline:
• .... a pipeline, the operation of which can be driven entirely by the infor-
mation available in the raw file headers
• the pipeline shall be able to combine without operator intervention all ex-
posures obtained of one field through one filter and during one night
• on the computer platform provided by the consortium, the pipeline shall be
able to fully reduce an average day worth of science and calibration exposures
within 12 hours elapsed time. Operator preparation of the execution shall
not exceed half an hour
• The data products of the pipeline shall be images with the above astrometric
and photometric calibrations.

Following VLT-SPE-ESO-19000-1618, the following deliveries shall be prepared

by the Consortium:
• Descriptions and implementations of calibration req.’s needed for the sup-
port of the instrument modes described in Section 4 .
The baseline requirements for these calibration procedures are specified in
Section 5 of the URD while details of the implementations are given in
Section 5 of the Calibration Plan. The consortium will specify the req.’s
and the implementations, and will deliver modules and recipes. Integra-
tion will be done under ESO responsibility. Detailed specifications of the
implementations are given in the Data Reduction Specification document
(DRS).
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 91

• Support for a Quality Control System The consortium will provide the
necessary routines/procedures, as components of the DFS pipeline infras-
tructure, but not for QC0 which is generic for all VLT compliant instruments
and which will be provided by ESO. The integration of the quality control
routines will be done under ESO responsibility. The execution of quality
control will be done on at least two ‘platforms’ within the context of the
DFS. Currently, the DFS infrastructure does not support the storage of QC
parameters, neither its trend analysis. It is highly desirable to improve upon
this situation.

The Quality Control will involve the following:

• QC0 on-site, on-line consistency check
The primary function of the QC0 quality check is to confirm that a given
exposure is consistent with the definition of the Observing Block (Observing
Description, Target Package, Constraint Set), of which the exposure is a
part. To this end relevant keywords from the FITS header are compared to
relevant parameters of the OD, TP, and CS. In addition, the size of the FITS
file is checked, to verify that the data transfer was completed succesfully.
Passage of QC0 is assumed to be a necessary and sufficient precondition for
input into the calibration or image pipeline. Implementation will require a
full specification of the possible ODs, TPs, and CSs.
• QC1 off-line Quality Control provides a full assessment of the quality of the
raw data and the derived data products.
QC1 includes both the measurement of Qualitity Control Parameters (in-
cluding image statistics, seeing, and errors on astrometry and photometry)
and the subsequent assessment of these parameters to arrive at an overall
statement of the quality of the data. It is useful to distinguish between QC1
measurements at different processing stages:
- pre-calib off-line quality check of incoming raw calibration data.

- pre-image off-line quality check of incoming raw science data.

The data coming from the telescope should be checked for consistency
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 92

with expected values. These checks include: i) check that global illumi-
nation (median) is within a range expected for that observation (seq.–
631 ), and ii) check that image quality (PSF) is consistent with seeing
(to detect tracking problems, for instance, seq.– 633)
- post-calib off-line quality check of CalFile– .
The CalFile– produced by the calibration pipeline should be checked for
validity as specified in the req.s before being put into the Calibration
data base, thus securing that only valid CalFile– sare used by the image
pipeline. Trend analysis will be the main tool for monitoring the validity
of CalFile– s.Trend analysis is run outside the pipeline. It is used as a
tool to monitor the behaviour of the instrument. The results of trend
analyis are not used to interpolate entries in the db to be fed back to
the pipeline.
- post-image off-line quality check of reduced science data.
The final characterization of the quality of the reduced science data
includes the following items.
- Validity of the processing operations.
- Quality of the processed data (sensitivity and resolution)
- Photometric check (consistency between different CCDs)
By comparing the quality of the processed data, with the quality of the
input data, the validity of the processing steps can be assessed. By
comparing the quality of the processed data with the expectations as
determined by the ETC and seeing-statistics, overall performance of the
instrument and ambient observing conditions can be determined.
Example: a powerful check on the astrometry, seeing, image quality,
tracking rotator, tracking alt-az, co-addition is given in seq.– 636 which
for Mode– dither N=and Mode– jitter N=compares the PSF over the
FOV between the input and output images.

A different projection of the various planned quality control tasks is:

- Operational checks verify on a low level whether, people, machines
and computer recipes are operating according to expectations (e.g. not
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 93

down). These items relate to the execution of implementations and

will be specified in ESO’s operations plans, but not in the URD or the
Calibration Plan.
- Functional and maintenance checks on mechanical and electronic
components on the engineering level, which generally have the character
of a go/no-go result. These checks are given in the maintenance manual.
- Instrument S/W self-supporting checks. The instrument S/W per-
forms various self-checks. Malfunctions are displayed on the Instrument
workstation.
- Routine (day-week-month-year) calibration procedures, as listed in de-
tail as req.’s in Section 5. These will be handled by the calibration
recipes. E.g. for daily checks the URD specifies req. 533 Quick detec-
tor responsivity check.

• Automatic Pipeline Processing

The OmegaCAM pipeline is the collection of modules and recipes integrated

in ESO’s DFS pipeline infrastructure environment.
The pipeline produces data corrected for the instrument signature.
It is useful to distinguish between the calibration pipeline and the image
pipeline.
I The calibration pipeline produces calfiles (CalFile– ), as outlined in
Section 5. The calibration pipeline operation schedule is driven by the
specifications given in each requirement about the frequency at which
the characterizations have to be executed at the telescope. Assuming
a smooth dataflow from telescope to ESO HQ the processing of these
calibration data will follow the telscope schedule. CalFile– s are valid
for a certain period as specified by the requirements and they are times-
tamped by the validation and implementation procedures. It is at the
moment not clear whether the current DFS infrastructure can run the
calibration pipeline in an automatic mode, or whether it should be run
”by hand”
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 94

II The image pipeline applies the calfiles to raw science data, to produce
astrometrically and photometrically calibrated science data. The image
pipeline is designed to operate as much as possible as a single process
(in spite of its many sub-processes, as a black box, with a minimum
of storage of intermediate results in order to constrain the OmegaCAM
data storage requirements within acceptable volumes, and to further fa-
cilitate parallel processing on external platforms with a minimum of in-
terdependence). The sequential steps of the image pipeline are specified
in Section 6.3 and are named seq.– and involve:
- Image arithmetic (de-bias, flat field)
- Cosmic ray and bad pixels correction
- Image statistics
- Astrometric solutions
- Image combination (co-addition)
- Flux calibration
The integration of the pipeline will be done under ESO responsibility.
Part of the Data Reduction Software is a set of instrument sample data and cal-
ibration data to be processed as representative cases. Results of the processing
of such data in a test environment will also be delivered.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 95

Coding standards:
The software developments shall be made in ANSI C, according to programming
conventions, data structures, and low-level functionalities provided by ESO. A
list will be established further (low-level) ESO-provided functionalities.
The complete description of a recipe includes:
1. Definition
2. General description
2.1 Context of utilisation
2.2 Observing procedure
2.3 Calibration data
2.4 Resulting data format
2.5 Expected pre-processing, post-processing
2.6 Pipeline calibration

3. Mathematical description
3.1 Mathematical justification
3.2 Error propagation/evaluation
3.3 Articles publication

4. Pseudo-code
5. Validation
Of course, this only applies for ”complex” recipes, there is no need to produce
a so detailed algorithmic description for trivial tasks.
As examples of high-level algorithm description, the calibration and operation plans for already
existing VLT instruments are a good reference. The instrument scientist tried in each case to
describe the process to apply to both calibration and observation data, to produce acceptable
and useful pipeline products.
The recipe description above matches well to the more formal Odoco format
employed here for the subsequent documentation of both the basic require-
ments, the implementation and data analysis of req.’s. The present set of all
Odoco items reads:
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 96

Title
Objective oneliner + description
Fulfilling or fulfilled by when selfstanding there is a TSF
When performed/frequency
Sources, observations, instrument configurations
Inputs dependencies CalFile– ,pre-processing
Outputs CalFile– ,post-processing
Required accuracy, constraints
Estimated time needed both for observing and data reduction
Priority desirable, very important, essential
TSF - name
Recipe - name
CA description of calibration analysis (also mathematical, error propagation)
Needed Functionalities
CAP implementation of calibration analysis (or pseudo code)
Status of Req. management tool
FLAG management tool

A good example of algorithmic documentation is given in the document: ”Infrared jitter imaging
data reduction algorithms”
http://www.eso.org/projects/dfs/papers/jitter99/
The coding standards and recommendations followed by ESO and external
consortia is a set of rules issued by the C programming community. It can be
found on the Web from:
Recommended C Style and Coding Standards
http://www.apocalypse.org/pub/u/paul/docs/cstyle/cstyle.htm
The pipeline software shall be exportable to POSIX-compliant systems. In the
case of OmegaCAM, because the target platform is based on Linux (Beowulf
clusters), this requirement could be narrowed down to: the software shall be
running on Linux workstations. However, making POSIX-compliant software in
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 97

general and not attaching the software to any system’s specifics is a good idea
to preserve portability in the future.
On-line documentation for this software shall be in HTML format.
Additional documents:
[2] Recommended C Style and Coding Standards see URL above.
[3] ”ISAAC calibration plan” by PA Duc and JG Cuby VLT-PLA-ESO-14100-
1384
[4] ”Infrared jitter imaging data reduction algorithms” see URL above.

The standard software needed to execute the pipeline consists of the following
packages:
- FITS data reduction–eclipse, LDACtools, cfitsio.
- Source extraction–Sextractor.
- Astrometry–LDACtools.
- Co-addition–SWARP.
The consortium will support and maintain the LDACtools. eclipse is a package
maintained by ESO. All other packages are in common use. These packages
provide for all computationally expensive operations to be performed on the
data. All these packages are written in C.
To integrate this set of sophisticated data reduction and analysis tools, the
consortium currently uses the Python scripting language. This language pro-
vides a high level easy-to-maintain glue. All recipes and procedures are written
using this language as a layer on top of the aforementioned packages, provid-
ing transparent, common interfaces to these diverse packages. Prototypes of
the recipes will be delivered to ESO using Python. It is recognized that final
delivery may require recoding these in C.
The consortium prepares for a scripting of the pipeline modules offering all
required functionalities with a generic database interface. To this interface
functional programming (files) or object oriented (objects) databases could be
attached. These modules can be integrated by ESO in its DFS environment,
without using these interfaces.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 98

Internally, the Consortium uses CVS for its versioning control of code, recipe’s,
dictionaries and documentation and it considers to use Objectivity for storing
and distributing all calibration and source-list data.
An Interface document specifying all data items to be forwarded by ESO to the
Consortium is planned for (FDR).
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 99

6.2 Estimate of data volumes

The planned observing strategies, particularly the widely used dithering and
jittering techniques have a large impact on the eventual datavolume of both
raw and processed data. As first estimates of the amount of science data and
calibration data indicate data volumes and data rates at the edge of the storing
and processing capabilities, these items might force some high level constraints
on the data handling.
Example image data volumes:
OmegaCAM contains 256 Megapixels which are read-out with 16 bit integers.
Thus one read-out produces 0.5 Gbyte of raw data. A standard dither will do 5
read-outs at various positions on the target, with integration times of 4-6 min
each, leading to a total of 20-30 min of integration on the target. In these 20-30
minutes 2.5 Gbyte of raw data is acquired. In a 10 hour night this corresponds
to 50 - 75 Gbyte of raw data. For 300 nights a year this is 15 - 22.5 Tb of raw
data.
The flatfielding and de-biasing of the raw data require (32bit) floating point
output images and when applied bluntly would double (in fact triple) the total
data volume to 45 - 67 Tb a year. However, de-dithering will produce one
floating point output image out of 5 input images and thus leads to a data-
reduction by a factor of five. In this case, the overall reduction of the data
volume will be a factor (5/2)=2.5 and the total data volume of de-dithered,
flatfielded, debiased images will be 6-9 Tb.
Clearly, it is desirable to process the de-biasing, flatfielding and de-dithering in
one go without the archiving of intermediate steps. Alternatively, one might
consider to do this process on the fly, in which case the total data volume stays
as the original raw data of 15 -22.5Tb a year.

SCIENCE IMAGE DATA VOLUME 5 DITHERS in 20 - 30 min

raw data flat fielded co-added raw+ co-add
NIGHT 50 - 75 Gb 100 -150 Gb 20 -30 Gb 70 - 105 Gb
300 NIGHTS 15 - 23 Tb 30 - 45 Tb 6 - 9 Tb 21 - 31 Tb
+ CALIBRATION RAW IMAGE DATA - see A2 - current estimate 8Tb
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 100

ASTRONOMICAL SOURCE LIST DATA VOLUME

NOMINAL USING CO-ADDED FRAMES
25 sources per square arcmin → 105 sources/degree2
(CFHT reckons 2 105 for 4m telesc. 42 /2.52 = 2.5 )

20 - 30 parameters per source → 2−3 106 real*4/degree2 = 8 - 12 Mbyte /field

10 hour night (20 -30 CO-ADDED fields) → 160 - 360 Mbyte

330 night/year → 53 - 120 Gbyte

(when not using co-added frames factor 5 more → 0.5 Tb)

ABSOLUTE MAXIMUM
- Crowded fields
- 0.5 arcsec seeing
- reliable source decomposition on 2.5 arcsec scale
576 sources per square arcmin
23 times more than the nominal example → 184 - 276 Mbyte /field
10h night (20 -30 CO-ADDED fields) → 3.7 - 8.3 Gbyte/night
(when not using co-added frames factor 5 more)
So, a run of 10 nights on crowded field can result in 10 x 8.3 x 5 = 0.4 Tb of
source data on non co-added frames
few observing runs end up in Tb regime
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 101

6.3 Baseline Requirements for the Image Pipeline

As noted previously, it is useful to distinguish between the calibration pipeline
and the image pipeline. Here we provide more detail on the operation of the
image pipeline
The image pipeline applies calfiles to raw science data, to produce astromet-
rically and photometrically calibrated science data. The DFS pipeline applies
different recipes to raw science data, based on the observing template used.
There are currently three observing templates for science data envisioned for
OmegaCAM. These templates (see also Section 4 CP), and their corresponding
recipes are:
• TSF– OCM img obs stare, Recipe– Science Stare(N=1), or Recipe–
Science Jitter(N>1)
• TSF– OCM img obs jitter, Recipe– Science Jitter
• TSF– OCM img obs dither, Recipe– Science Dither
The recipes (Science Stare, Science Jitter, Science Dither) combine a series of
processing steps that are specified by their (intermediate) data product in seq.–
631-seq.– 636.These can be summarized by:
• Initialization and image statistics Create a new entry in the database of
derived products for each raw image. Include image statistics in the entry
(seq.– 631).
• de-biasing and flatfielding This step removes the gain-variation within
images (seq.– 632).
• Constructing weights, incorporating the detection of cosmic rays, and
satellite tracks (seq.– 633).
• Astrometry, measuring the astrometric distortions and offsets (seq.– 634).
• Apply photometry (seq.– 635)
• Dedithering, combining the images to construct a clean, maximally uni-
form, image of the FOV of OmegaCAM and perform a quality check (seq.–
636) .
The detailed specification of each seq is analogous to the specification of the
req’s in Section 5. It should however be understood that within the context of
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 102

the DFS pipeline, the recipes for the individual seq’s are to be combined into
one recipe for each Observing template (so-called superscripts). An overview
of the data flow within a recipe is given in Fig. 6.3.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 103

Image pipeline dataflow

Ingest,
images statistics, Raw
Archive validation science data
seq631

Cal 541
Bias
de-Bias and flatfield
Reduced
Cal 546 science data
seq632
Flatfield

Cal 523
ADU conv.
Make individual
weights Reduced
Cal 546W weights
seq633
Weight

Cal in1 Do astrometric

Astrometric calibration Astr. calib
reference science data
seq634

Cal563/4 Apply photometric

zeropoint+ calibration Astr.+Phot .
extinction calib
seq635 science data

Dedithered
science data
Coadd science data

Dedithered
seq636
weights

Dedithered
context
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 104

The recipe Science Stare differs from the recipes Science Jitter and Science Dither
in that the co-addition step (seq.– 636)is not included. Furthermore, Sci-
ence Dither produces one co-added image combining data from all chips, while
Science Jitter produces one co-added image per chip.
Operations seq.– 632-seq.– 635 are performed on individual chips. The format
of the raw data will be one FITS file per observation including the data for 32
chips in 32 extensions. seq.– 631 separates these raw data in FITS files for
individual chips. Only in seq.– 636 the data from individual chips are combined.
Note that seq.– 631-seq.– 636 can also be used within various recipes for the
calibration pipeline.
Clearly, the image pipeline will determine parameter values which are specific
for individual raw data frames or their 32 CCD frames, and these could be
viewed as transient calibration data. Most of this transient calibration data
is uninteresting or could be easily re-constructed, while other data have per-
sistent scientific value (such as seeing, PSF, astrometric solution, zeropoint,
sky-brightness, etc.), while other data have potential, but unspecified, value
for trend analysis or trouble-shooting. Output files of the image pipeline (flat-
fielded data, astronomically calibrated data, individual weight maps, etc.) are
called SeqFile– ’s, e.g. SeqFile– 633 individual weight map. The image data
recorded in a SeqFile– may be discarded when the data is no longer needed.
Care has been taken to propagate intermediate information from the headers of
the SeqFile– ’s to header of the pipeline product, whenever appropriate. The
ESO-DMD - OmegaCAM Interface control document discusses in more detail
the requirements for maintaining this intermediate informaation.
Both the image pipeline and the calibration pipeline shall obey the accuracies
and errors as listed in Section 5. Propagation of errors shall be computed,
where applicable.
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 105

A1 LIST of CALIBRATION REQUIREMENTS

5.2.1 Req.– CCD read noise - doit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

5.2.2 Req.– Hot pixels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5.2.3 Req.– CCD gain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
5.2.4 Req.– Electromagnetic Compatibility . . . . . . . . . . . . . . . . . . . . . . . . 50
5.2.5 Req.– Electrical cross talk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.3.1 Req.– CCD Dark Current - doit . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
5.3.2 Req.– CCD Particle Event Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
5.3.3 Req.– CCD Linearity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
5.3.4 Req.– CCD Charge Transfer Efficiency . . . . . . . . . . . . . . . . . . . . . 53
5.3.5 Req.– CCD Cold Pixels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
5.3.6 Req.– CCD Hysteresis, strong signal . . . . . . . . . . . . . . . . . . . . . . . . 55
5.4.1 Req.– Bias - doit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
5.4.2 Req.– Flat-field - dome key bands + user bands - doit . . . . . . . 61
5.4.3 Req.– Flat-field - twilight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.4.4 Req.– Flat-field - night sky . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.4.5 Req.– Flat-field - Fringing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
5.4.6 Req.– Flat-field - master flat and weight map . . . . . . . . . . . . . . . 65
5.4.7 Req.– Quick detector responsivity check - doit . . . . . . . . . . . . . . . 66
5.4.8 Req.– Illumination correction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.5.1 Req.– Position of Camera in focal plane . . . . . . . . . . . . . . . . . . . . 69
5.5.2 Req.– Telescope Pointing and offsetting . . . . . . . . . . . . . . . . . . . . . 70
5.5.3 Req.– Telescope and Field Rotator tracking . . . . . . . . . . . . . . . . . 70
5.5.4 Req.– PSF Anisotropy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
5.5.5 Req.– The astrometric solution for templates - doit -see 6.3.4 72
5.5.6 Req.– The astrometric solution for Guide CCD’s . . . . . . . . . . . 72
5.6.1 Req.– Shutter Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.6.2 Req.– Photometric Calibration - monitoring . . . . . . . . . . . . . . . 78
5.6.3 Req.– Photometric Calibration - zeropoint keybands - doit . . 80
5.6.4 Req.– Photometric Calibration - zeropoint user bands . . . . . . . 81
5.6.5 Req.– Filter band passes - user bands vs key bands . . . . . . . . . . 82
5.6.6 Req.– Dependency on rotator angle/reproducibility . . . . . . . . . . 83
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 106

5.6.7 Req.– Linearity (as a function of flux) . . . . . . . . . . . . . . . . . . . . . . 83

5.6.8 Req.– Detection limit and ETC calibrations . . . . . . . . . . . . . . . . . 83
5.6.9 Req.– Secondary Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5.7.1 Req.– Camera focus/tilt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.7.2 Req.– Ghosts - ADC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 107

A2 LIST of RAW CALIBRATION DATA

volume per req.run and per year

RawData– 523 CCD gain 10Gb 0.52Tb
RawData– 531 Dark current 0.75Gb 40Gb
RawData– 533 CCD Linearity 2Gb 24Gb
RawData– 541 Raw Bias frame 5Gb 0.51Tb
RawData– 542 Dome flat frame 2Gb 0.66Tb
RawData– 543 Twilightflat frame 10Gb 3.3Tb
RawData– 547 Quick check 0.5Gb 115Gb
RawData– 562 Monitor 0.5Gb 0.5Tb
RawData– 563 Zeropoint- Key 2.5Gb 0.8Tb
RawData– 564 Zeropoint - User 2.5Gb 0.8Tb
RawData– 565 User − >key 25Gb 25Gb
RawData– 569 Secondary Standards 32 Gb 4Tb
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 108

A3 LIST of DFS I/O CALIBRATION FILES

volume per file and per year

CalFile– 521 Readout noise 2 kb, 1 Mb
CalFile– 522 bad/hot pixel map 0.25Gb 25Gb
CalFile– 523 conversion factor e− /ADU 2kb .1Mb
CalFile– 531 dark count rate for each CCD 2kb .1Mb
CalFile– 532 Particle event rate 2kb .1Mb
CalFile– 533 CCD Linearity 2kb .1Mb
CalFile– 534 charge transfer efficiency factors .2Mb .4Mb
CalFile– 535 cold pixel map .25Gb 1Gb
CalFile– 536 CCD Hysteresis 2kb 2kb
CalFile– 541 Master Bias frame 1Gb 104Gb
CalFile– 542 Master Domeflat frame 1Gb 0.330Tb
CalFile– 542L Dome Lamp 2kb 1Mb
CalFile– 543 Master Twilightflat frame 4Gb 1.3Tb
CalFile– 544 Nightsky flat frame 1Gb 1Tb
CalFile– 545 ff-fringe 1Gb 0.3Tb
CalFile– 546 Master flatfield 1Gb 0.3Tb
CalFile– 546W Weightmap 1Gb 0.3Tb
CalFile– 547 Quick check 1Gb 0.33Tb
CalFile– 547r Quick check - day report 3kb 1Mb
CalFile– 548 Illumination correction 1Gb 10Gb
CalFile– 551 Astrometric camera/chip solution 2kb .1Mb
CalFile– 554 PSF anisotropy .1Mb 1 Mb
CalFile– 556 Astrometric solution - Guide CCD 2 kb 1 Mb
CalFile– 562S Sky brightness 1 kb 1 kb
CalFile– 562 Extinction-night report 2 kb 1 Mb
CalFile– 562u Photom + Sky 10Mb 10Gb
CalFile– 562B Photom + Sky 10Mb 10Gb
CalFile– 562V Photom + Sky 10Mb 10Gb
CalFile– 562i Photom + Sky 10Mb 10Gb
CalFile– 563Z Zeropoint - key bands 4kb 1.2Mb
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 109

CalFile– 564 zeropoint - User bands 4kb 1.2Mb

CalFile– 565 User− > key 4kb 1Mb
CalFile– 569E Primary Standard stars 1 Mb 1 Mb
CalFile– 569 Standard Catalog 1 Mb 1 Gb
 ID VST-SPE-OCM-23100-3050
NOVA - Kapteyn Institute OmegaCAM
Issue VERSION 1.1
Date 29 Oct 2001
USM – OaPd URD
Page 110

A4 LIST of DFS INPUT REFERENCE CATALOGUES

CalFile– in1 US-NAVAL Observatory A2.0 6 Gbyte

CalFile– 569E Primary Standard stars - external Landolt fields 100 Mb
CalFile– 569E Primary Standard stars - external WFI@2.2 100 Mb
CalFile– 564E Standard extinction curve 0.1Mb