Guide to browsing the CFHTLS exposures catalogs
From the various information located in the exposure FITS header, the Elixir detailed statistics and the QSO observer and coordinators qualification criteria, a string is built to fit nicely within the CFHTLS DATA web pages. The principle is to have a single line per acquired exposure with all the best possible and most complete information to properly define and characterize that exposure.

Of course not everything can fit in a single line, and the MetaData are the pieces to consult to obtain more information. This present information, uniform over all the exposures of the survey, aims at setting the first order of data quality assessment of the survey from the CFHT viewpoint (all the statistics proposed on these CFHTLS DATA pages are derived from these catalogs of exposures).

A color code is used to differentiate at first sight the 3 survey components:

Deep Survey Wide Survey Very Wide Survey

A typical entry for a given exposure looks like this:

 
 730857 | 03BQ06 Dec 25 21:06:43 03 | L01 D1         |  2:25:59.8  -4:29:46 2000 |  360 | r | 1.11 | 0.70 0.80  1209 |P 1 V Q|
 
 

The separator is the "|" character and there are a total of 10 fields, some having several sub-fields. The following sections describe each of the 10 fields individually.


 730857 | 03BQ06 Dec 25 21:06:43 03 | L01 D1         |  2:25:59.8  -4:29:46 2000 |  360 | r | 1.11 | 0.70 0.80  1209 |P 1 V Q|
 

The filename: the CFHT's odometer number

The CFHT uses a unique filename identificator for each exposure. Started in the 90s as a six digits code, it'll probably reach soon the million and will then pass to a 7 digits code. There can't be two exposures with the same odometer number and this is the best identificator for a given exposure.


 730857 | 03BQ06 Dec 25 21:06:43 03 | L01 D1         |  2:25:59.8  -4:29:46 2000 |  360 | r | 1.11 | 0.70 0.80  1209 |P 1 V Q|
 

The observing run: semester and QSO run

This field is the granularity element used at CFHT to document the CFHTLS: it changes at each observing run, marking a new instrument setup. The first three letters define the semester (03B for the second semester of the year 2003) and the last three letters define the QSO observing run (Q04 stands for the fourth run of the given semester).


 730857 | 03BQ06 Dec 25 21:06:43 03 | L01 D1         |  2:25:59.8  -4:29:46 2000 |  360 | r | 1.11 | 0.70 0.80  1209 |P 1 V Q|
 

The date of observation: Hawaii local time (not UT!) at the END of the exposure

A string of characters giving the month, the day the local time and the year (year coded only on two digits). Note that this time corresponds to the time the file is created on disk at the end of the exposure.


 730857 | 03BQ06 Dec 25 21:06:43 03 | L01 D1         |  2:25:59.8  -4:29:46 2000 |  360 | r | 1.11 | 0.70 0.80  1209 |P 1 V Q|
 

CFHTLS Run Identificator and Field Name: survey component and pointing

There are 6 run identificators for the CFHTLS (Deep=L01/L04, Wide=L02/L05, VeryWide=L03/L06) and those appear in this log in the first position of the string. The field name as entered by the CFHTLS individual coordinators in the QSO's PH2 tool is provided next.


 730857 | 03BQ06 Dec 25 21:06:43 03 | L01 D1         |  2:25:59.8  -4:29:46 2000 |  360 | r | 1.11 | 0.70 0.80  1209 |P 1 V Q|
 

Field coordinates: right ascension, declination and epoch

The fractional seconds on the declination have been removed, but those coordinates are still precise enough to define the pointing within a few arcseconds. Note that these are the telescope control system coordinates for the observed field, not the accurate coordinates of the center of field that could be derived from precise astrometry analysis of the image (the difference is only a few arcseconds).


 730857 | 03BQ06 Dec 25 21:06:43 03 | L01 D1         |  2:25:59.8  -4:29:46 2000 |  360 | r | 1.11 | 0.70 0.80  1209 |P 1 V Q|
 

Exposure time: nearest integer value in seconds

The shutter precision is 1 ms, hence it is safe to truncate the exposure time to the nearest integer as the exposures are always defined to be taken with an integral value exposure time.


 730857 | 03BQ06 Dec 25 21:06:43 03 | L01 D1         |  2:25:59.8  -4:29:46 2000 |  360 | r | 1.11 | 0.70 0.80  1209 |P 1 V Q|
 

Filter: single letter definition

The filters used for the CFHTLS are u*,g',r',i',z' but for simplicity of file handling, the filters are simply defined as u,g,r,i,z.


 730857 | 03BQ06 Dec 25 21:06:43 03 | L01 D1         |  2:25:59.8  -4:29:46 2000 |  360 | r | 1.11 | 0.70 0.80  1209 |P 1 V Q|
 

Airmass: airmass at the time of the start of the exposure

The airmass is coded with two decimals for a better precision on data mining.


 730857 | 03BQ06 Dec 25 21:06:43 03 | L01 D1         |  2:25:59.8  -4:29:46 2000 |  360 | r | 1.11 | 0.70 0.80  1209 |P 1 V Q|
 

Image quality: center and outer "ring" median image quality in arcseconds

As described in the "Technical Considerations" of the MegaPrime pages (1.3 Evaluation of the image quality across the field of view), these two image quality values allow for defining the overall quality over the field of view. A typical offset of 0.10 to 0.15 arcseconds between the center and the the outer ring is to be expected. When no data are available, the values are set to "nan".

The "MegaNorth" period of May 2003 (half of the mosaic operational only) was left with no value for Elixir statistic (set both values to "nan") to clearly mark these data as marginal in comparison of the standard instrumental setup.


 730857 | 03BQ06 Dec 25 21:06:43 03 | L01 D1         |  2:25:59.8  -4:29:46 2000 |  360 | r | 1.11 | 0.70 0.80  1209 |P 1 V Q|
 

Sky background: in ADU on the raw data

The Elixir statistics analysis provides the sky background (corrected for the detector bias level) on all 36 CCDs of the mosaic. The value presented here is the median value of the stack of these 36 measurements.

When no data is available, the value is set to "0".

The "MegaNorth" period of May 2003 (half of the mosaic operational only) was left with no value for Elixir statistic (set value to "0") to clearly mark these data as marginal in comparison of the standard instrumental setup.


 730857 | 03BQ06 Dec 25 21:06:43 03 | L01 D1         |  2:25:59.8  -4:29:46 2000 |  360 | r | 1.11 | 0.70 0.80  1209 |P 1 V Q|
 

The four exposure external information: sky transparency, observing rating, exposure validation, and observing group validation.

1) Sky transparency:

SkyProbe provides a robust way to define the absolute value of the sky transparency at the time of the observations (1 point every minute in the V-band only). But sometimes there can be a very slight haze that goes undetected. The service observer looks both at the SkyProbe plot (which is found to be reliable in 97% of the cases), the satellite map, and the guide star flux variations to define in absolute if the conditions are or are not photometric. The photometric flag is set manually in real-time during the night by the observer for each exposure based on these inputs.

The value of the flag can be either "P" for Photometric or "A" for Absorbed. The actual level of absorption can be derived in the V-band from the SkyProbe plot archive, and the MetaData also provide a table of the SkyProbe zero points throughout the whole night (the tabular version of the plot), allowing for a direct determination of the absorption for a given exposure (only for the V-band though).

However the "Observer Rating" and the "Coordinator Rating" flags described below can tell a lot on the level of absorption in the case the "Photometric" flag has been set to the absorbed state.

When no data is available, the flag is set to "X".

2) Service Observer rating:

This flag ranges from 1 to 4 and is set in real-time during the night by the observer at the telescope.

- 1:Taken within, or better than, the conditions defined by the Principal Investigator (the Survey Coordinator for the CFHTLS) in terms of seeing, sky background, and sky transparency.

- 2:Taken almost within the conditions defined by the Principal Investigator (the Survey Coordinator for the CFHTLS): one of the three parameters (seeing, sky background, sky transparency) is slightly out of boundaries.

- 3:Taken outside the conditions defined by the Principal Investigator (the Survey Coordinator for the CFHTLS): one of the parameters (seeing, sky background, sky transparency) is way out of boundaries, or two, or all of the parameters are out of bound.

- 4:Horrible observing conditions which most likely prevent any science to be done.

- 5:A lost cause, nothing to be done with that exposure.

Note that these three defining parameters are provided in that exposure information line as described above (Image quality, Sky background, Sky Transparency), hence if you see the Service Observer rating flag set to 2, 3 or 4, you can easily conclude the reason why the exposure is considered faulty.

When no data is available, the flag is set to "X".

3) QSO Coordinator validation:

This flag can be set to "V" for "Validated" or "O" for "Observed". The day following the night when that exposure was obtained, the QSO coordinator checks the service observer rating and other comments to define if that exposure fits within the specifications defined by coordinator. The fact that an exposure is validated does not however necessarely means that its integration time will be accounted for in the time balance allocated to the observing program: this depends on the observing group validation described below.

- V: the exposure qualifies for the science goals defined by the Principal Investigator. An exposure can be validated even if the Service Observer rating is equal to 2 since a degradation on of the parameters won't most likely affect the science dramatically.

- O: the exposure does not qualify for the science goals defined by the Principal Investigator (the Survey Coordinator for the CFHTLS). The exposure will be processed and distributed anyway but the time used to acquire it is not deducted from the time allocated to the observing program.

When no data is available, the flag is set to "X".

4) QSO Coordinator Observing Group validation:

This flag can be set to "Q" for "Qualified" or "D" for "Dismissed". Within the CFHTLS programs, exposures are always taken by groups: a set of dithered exposures for the Deep and the Wide, or a long string of various pointings for the Very Wide for example. A group lasts 2 hours maximum (1 hour is typical) and is usually composed of 5 to 15 exposures. If some exposures within the observing group are out of specifications then the entire set of exposures, included the validated ones, will be dismissed in terms of time accounting for that observing program.

However, the validated exposures can still be very useful for science programs other than the main one defined for the principal program. For example, the deep survey (which focuses on stacking as many exposures as possible) will benefit from extra validated exposures in dismissed groups taken for the supernovae program (which focuses on time sampling).

However, the Observing Group validation flag shown here holds for the given pointing, not the exposures taken when the group was executed. For example if an observing group was observed but not qualified, the set of exposures matching (taken for) that observing group would be bearing the "D" flag. However if the observing group is executed again later on and gets validated then its qualified flag applies to all exposures attempted for that observing group, including the ones taken previously, even the non validated exposures! This is why some non validated exposures end up with a qualified observing group status: it simply means that that group was repeated again and that given exposures was re-taken in better conditions.

In consequence, the catalogs proposed on these pages won't tell if a given exposure had its integration time deducted for the observing program. The QSO database has more information on these relative connections and can compute very precise statistics actually based on Observing Groups validation. See this page to consult the QSO based statistics.