EIS Definitions and Conventions
30 August 2004
The EIS survey system is designed to administer and document (via
integration with a robotically maintained WEB service) multiple
surveys and strategies in their entirety, from survey definition to
release of science-grade products by nominally operating in
unsupervised mode. Achieving this requires the introduction of
well-defined procedures to enable adaptive, automatic and
sequential addressing of information. Occasionaly, these may be
need to be properly explained, which will be done in this document
In addition, the system also produces a large variety of survey
products requiring clarification on the different data types, FITS
header keywords and conventions.
The purpose of this document is to serve as a quick reference
document for users of data products produced by the EIS survey
system. This is a ``live'' document which will evolve time as
For some surveys involving external collaborators, inconsistencies
may be present between adopted conventions which, while inconvenient,
are unavoidable. Below some of these differences are described and
their origin explained.
Ideally, the standard adopted by EIS to name a field is to use the
survey name and/or region as the root, followed by a numerical
suffix generated by the particular algorithm used in the
construction of the "Observation Block" (OB) and the associated
survey strategy (as defined by a set of pointings, instrument and
passband). Typically, this numerical suffix is assigned in an
increasing sequence based on the R.A. and Dec. defining the
pointings. This should be done when a survey is first defined. In
this way, all possible identifiers of a field utilized by the EIS
system (both internally and externally, e.g. OB name, object name,
target name, field name) and their graphical representations on the
EIS WEB pages form a homogeneous set.
Historically, however, the conventions have evolved in time due
to several factors:
- At their outset, prior to the introduction of a systematic
global approach, several surveys were defined case-by-case
and in isolation.
- Over time, some pre-existing and on-going surveys had their
geometries changed; methods for addressing positions were
reviewed; surveys were extended, amalgamated or otherwise
re-defined using modified algorithms.
- In the absence of pre-determined standards, collaborative
surveys co-administered by independent groups have led to
different notations. The interface to a common scheme (when this
requirement is inevitably identified) becomes particularly
difficult leading unavoidable to inconsistencies.
Hence, in practice, the desired ideal case is not always attainable
and requires some retroactive adaptations. The following noteworthy
discrepancies between these different conventions are:
- There exist surveys for which addressing information was
assigned in isolation at the survey outset and which were also
assigned a sequential value to be incorporated into, and
referenced by, the EIS system at a later date. For these
surveys, while the multiple nomenclatures persist
simultaneously, sequentially numbered fields and/or regions
referred to in EIS indices do not correspond to numerical values
which may appear in region and/or field names assigned prior to
introduction of the EIS system. For example, in the
the field indexed by the EIS system (including the web service)
as "1" does not necessarily correspond to the field name
originally defined to be "F01" in the GOODS survey
- The current EIS terminology prefaces fields with root names
derived from the particular survey name whereas the same fields
may be refer to by a subjective name (e.g.XMM-06 =
SHARC-2). This applies to the OB names as well as the OBJECT
keyword in image FITS header. Thus, the use of this keyword to
rename the distributed images is discouraged since it may lead
- Sequential TILEID numbers appearing in the FITS header of an EIS
product are automatically assigned by the EIS system when
binding mosaic images and their contributing tiles into a set of
related products for a given release. The binding depends on the
geometry and completeness of the contributing reduced
data. Therefore TILEID numbers do not necessarily correspond to
other indices such as the field identifiers used by the Survey
teams, or the field identifiers employed by EIS algorithms.
EIS Definitions and Conventions
- FIELD: An individual pointing of an instrument defines a
field. Fields are thus instrument specific and if exposures
at the same coordinates are taken with different instruments, they
constitute different fields. Each field belongs to exactly one
- REGION: Region is the area covered contiguously by fields
belonging to the same strategy. Each region belongs to exactly one
- STRATEGY: An observing strategy is a combination of telescope,
instrument, filter, exposure time and expected limiting magnitude.
Strategies are grouped by wavelenght domain (optical/infrared) and
by limiting magnitude (deep/shallow). Each of these groups is also
referred to as Strategy.
- SURVEY: A survey is a group of strategies covering one or more regions.
- FILTERS: All filters listed on the EIS WWW pages have the format
''passband''+''ESO Filter ID''. Details about these filters can be
found on the telescope pages which are directly linked from the filter names.
- OBSERVATION BLOCK (OB): as defined in the ESO Data Flow System
- REDUCTION BLOCK (RB): one RB is a group of raw images obtained in the same night
which share the same position, passband and form a
sequence in time. The content of one RB is usually
identical to an executed OB. However, the system
also allows RBs to be formed by concatenation of
similar sequentially executed OBs. In addition
individual exposures can be discarded based on
configurable constraints on their properties
(e.g. background counts, DIMM seeing values,
integration time, etc.)
- REDUCED IMAGE: A "reduced image" is the "nightly" product of a
reduction process of a so-called "Reduction Block"
- STACKED IMAGE: A "stacked image" is the product of an advanced
process in which "reduced images" sharing the
same instrument, passband and position on the
sky are grouped into so-called "Stack Blocks"
(SB). In the assembly of SBs, individual
"reduced images" can be discarded according to
configurable constraints on their properties
(e.g. measured seeing, amplitude of PSF
distortion, visual grade, OB completeness). The
images of the thus assembled SB are re-scaled
and accumulated in a weighted co-addition
procedure to generate a single final output
- MOSAIC IMAGE: A "mosaic image" (or simply "mosaic") is an
image formed from several contiguous (reduced
or stacked) "images" produced by the EIS
survey system, which have been subsequently
re-scaled to share a common zero point and
co-adjoined as "tiles" to form a single,
- MOSAIC TILE: A "mosaic tile" (or simply "tile") is an image
(reduced or stacked) produced by the EIS survey
system which has been subsequently re-scaled
such that it shares a zero point common to other
tiles and a large-scale "mosaic image" (or
simply "mosaic") of which it forms a component.
- VERSION: A release "version" represents a wholesale, global
change: either introducing a new set of products,
affecting or superseding the quality or properties of
more than 5 per cent of previously release products.
- REVISION: A release "revision" represents localized changes
which may affect individual products with a released
data set, while leaving others unaffected. Individual
products found to be anomalous may be removed or
replaced. New products may be added. The total
revisions must not alter more than 5 per cent of a
pre-existing set of products.
- ZP: FITS header keyword ZP contains the value of the zero point
(magnitudes) computed from night solutions. In a
reduced image the ZP in the header is given by ZP =
ZP' + KX. where ZP' is the zeropoint at zero airmass
(determined from the linear fit), K is the extinction
coefficient and X is the observed airmass. Therefore,
the Vega magnitude is given by mag (Vega) =
-2.5*log(flux) + ZP. In the case of a stack the ZP is
the zeropoint at zero airmass where the flux is the
number of counts directly measurable on the images
(note that the reduced images are normalized to 1
sec). It can also be set to a default value in which
case the associated error is set to a negative value.
- ZPERR: FITS header keyword ZPERR contains the value of the
uncertainty (magnitudes) associated with the
determination of ZP. This is estimated to be the
scatter relative to the best linear fit to the flux
measurements of standard stars as a function of the
airmass. It can also take the values of (-1) for
nights with poor fits; (-2) for nights without
standards; (-3) for nights in which the zeropoint
deviates significantly from a ``typical'' (normally
- ZPMOSAIC: FITS header keyword ZPMOSAIC contains the value of
the zero point (magnitudes) applied to the
processed tiles in order to refer instrumental
magnitudes to the common photometric system of the
mosaic, MOSAICID. The value for ZPMOSAIC is
determined by a re-scaling process in which
individual reduced images are harmonized to become
tiles sharing a common zero point for the entire
- ZPMOSERR: FITS header keyword ZPMOSERR contains the value of
the uncertainty (magnitudes) associated with the
determination of ZPMOSAIC.
- ZPMOSSYS: FITS header keyword ZPMOSSYS contains the
photometric system in which ZPMOSAIC and
ZPMOSERR are expressed.
- MOSAICID: FITS header keyword MOSAICID contains the
the unique identity of the mosaic to
which an individual tile belongs to.
- MOSTILES: FITS header keyword MOSTILES contains the
total number of individual tiles contributing to
the larger mosaic, MOSAICID.
- TILEID : FITS header keyword TILEID contains the unique
identity of the individual tile
within the larger mosaic, MOSAICID. TILEID does not
necessarily map directly to any other FITS keyword
value (e.g. OBJECT).
This document was generated using the
LaTeX2HTML translator Version 2002 (1.62)
Copyright © 1993, 1994, 1995, 1996,
Computer Based Learning Unit, University of Leeds.
Copyright © 1997, 1998, 1999,
Mathematics Department, Macquarie University, Sydney.
The command line arguments were:
latex2html EIS2004-08-27T15:26:18527.tex -split 1 -no_navigation
The translation was initiated by eis data account on 2004-08-30
eis data account