SINFONI pipeline:
Science files and recipes

SINFONI SCIENCE data have been processed by the pipeline with the best available calibration data. Please note that ESO is not assuming any responsibility in respect to the usefulness of the reduced data. The adopted reduction strategy may not be suitable for the scientific purpose of the observations.

Notes. The SINFONI science recipe sinfo_rec_jitter provides several ways to subtract the sky spectrum and to coadd individual cubes (PRO.CATG = OBS_OBJ) to a coadded cube (PRO.CATG = COADD_OBJ). The strategy does not depend on the used template but on the number of raw SCIENCE input frames and the number of frames with DPR.TYPE = SKY_NODDING.

pipeline version 2.0.8:

In case the science raw file stack contains SKY frames, the closest in times observed SKY frame is extracted from each DPR.TYPE = OBJECT frame. The SKY subtracted OBJECT frames are coadded, according to the information given in the CUMOFFSET fits header keys. The offsets in pixel units are PixelX = -2 * CUMOFFSETX; PixelY = 2 * CUMOFFSETY (this convention may change from Oct 1, P76 on).

In case the science raw file stack does not contain any SKY frames, several strategies are provided:

1. MEDIAN_STACKING: The parameter sinfo.objnod.autojitter_method=2 in the configuration file sinfo_rec_jitter. The median of each pixel is considered to be SKY and is subtracted from each OBS_OBJ frame.
   The following example of four OBS_OBJ cubes (here the median collapsed in z is shown with 99%-cuts)
   

   frame 1 minus frame 2	frame 2 minus frame 1
   frame 3 minus frame 2	frame 4 minus frame 3

   
   The final coadded product (again median collapsed) looks like this:
   
   The first sky subtracted cube appears in the center of the final COADD_OBJ cube. The three other cubes are coadded accordingly.
   
   
2. SKY_IS_NEXT: The parameter sinfo.objnod.autojitter_method=1 in the configuration file sinfo_rec_jitter. Pairs of consecutive OBS_OBJ frames are generated, of which one is subtracted from the other as sky. Sources in the subtracted frame appear as sources with negative flux in the final COADD_OBJ product.
   The following example of four OBS_OBJ cubes (here the median collapsed in z is shown with 99%-cuts)
   

   frame 1 minus frame 2	frame 2 minus frame 1
   frame 3 minus frame 2	frame 4 minus frame 3

   The left negative source is twice as strong, since it is frame 3 (subtracted from frame 4) and frame 2 (subtracted from frame 1). The closest in time frame of the stack is used as sky to be subtrcated. It can be the previous or the next frame of the science stack.
   The final coadded product (again median collapsed) looks like this:
   
   
   
3. The sinfo.objnod.autojitter_method=0 in the configuration file sinfo_rec_jitter. No sky is subtracted.

pipeline version 2.0.8 (used for date from ~2009-10-10 on), as for 1.0.6, SKY frames are subtracted, in case they are part of the raw frame stack. If there is no SKY frame given, then:

1. sinfo.objnod.autojitter_method=0 in the configuration file sinfo_rec_jitter : No sky subtraction. This method is also applied in case a single OBJECT raw frame is submitted (no SKY).
2. sinfo.objnod.autojitter_method=1 in the configuration file sinfo_rec_jitter :
   use SKY_IS_NEXT method
   subtract the dummy sky from the OBJECT
   each processed OBJECT frame is cube reconstructed (PRO.CATG = OBS_OBJ):
   jitter_int = 1 : coadded cube is generated
   scales_sky = TRUE : spatial median subtracted from each plane of the reconstructed cube to take out temporal residuals of the sky
   ks_clip = TRUE : kappa-sigma clipping applied to each pixel after cube aligning (= in the set of co-aligned OBJECT frames), to mask cosmics and hot pixels.
3. sinfo.objnod.autojitter_method=2 in the configuration file sinfo_rec_jitter: SKY_IS_NEXT method
4. sinfo.objnod.autojitter_method=3 in the configuration file sinfo_rec_jitter: MEDIAN_STACKING method

Defaults: sinfo.objnod.autojitter_method=1 scales_sky = FALSE , ks_clip = FALSE, jitter_int = 1

Science products of service mode packages use :

• sinfo.objnod.autojitter_method=1
• scales_sky = TRUE
• ks_clip = TRUE
• jitter_int = 1

In case science stacks contain embeded SKY frames, it is assumed there are reasons for that observing strategy (e.g. filamenteous background). In this case the SKY frames itself are used to correct OBJECT frames.

pipeline version 1.0.9 (used for date from ~2005-07-06 on)

1. sinfo.objnod.autojitter_method=0 in the configuration file sinfo_rec_jitter : No sky subtraction. This method is also applied in case a single OBJECT raw frame is submitted (no SKY).
2. sinfo.objnod.autojitter_method=1 in the configuration file sinfo_rec_jitter :
   use SKY_IS_NEXT method
   subtract the dummy sky from the OBJECT
3. sinfo.objnod.autojitter_method=2 in the configuration file sinfo_rec_jitter: MEDIAN_STACKING method

Defaults: sinfo.objnod.autojitter_method=1 scales_sky = FALSE , ks_clip = FALSE, jitter_int = 1

Science products of service mode packages use :

• sinfo.objnod.autojitter_method=0 : The individual object cubes are not sky subtracted. SKY is subtracted as part of the co-adding.
• scales_sky = TRUE
• ks_clip = TRUE
• jitter_int = 1

Input

Products.

QC parameters

QC Reports for SCIENCE Frames.

To aid in the assessment of the quality of a given science OB a number of QC plots are created. These plots are included in the Service Mode distribution.

QC Report 1                                                                                            QC Report 2

QC Report 1:
UL: the OBJECT_NODDING_STACKED full detector intermediate science image with the number of saturated pixels listed.
UR: the MED_COADD_OBJ median cube collapsed science image. A number of paramters are listed: the Program ID, the Obs ID, the target name, and for the brightest source detected:
source center (pixels), the integrated flux (within 3 sigma), the isophotal area (3 sigma), the net flux per pixel, the image FWHM (arcsec), the DIMM seeing corrected for the same airmass and filter used by this standard star, the status of the AO loop (OPEN or CLOSED), the position angle of the brightest source (degrees CCW from the x-axis), and its elongation and ellipticity.
The source center, its position angle, and its FWHM are marked on the image. NOTE: The conversion from pixels to arcsec for the image FWHM takes into account the fact that the reconstructed SINFONI cube has rectangular pixels. The reconstructed pixels are such that on-sky Delta_X = 0.5*Delta_Y. The correct conversion, in terms of the source position angle (PA), is:

pixel_scale is given by the keyword INS.OPTI1.NAME.
LL: the (scaled) difference between two cube planes. Plane 100 - Plane 2072. The wavelengths of these planes are listed to the right of this image.
LR: a horizontal (blue line) and vertical (red line) cut through the center of the brightest detected source of the MED_COADD_OBJ median cube collapsed science image. The respective averages through this center are shown as dotted lines.

QC Report 2:
UL: three single column traces of the current product (OBJECT_NODDING_STACKED) through: x = 500 (dark grey), x = 1000 (middle grey), x = 1500 (light grey) and an averaged sum of all columns (blue).
UR: three single row traces of the current product (OBJECT_NODDING_STACKED) through: x = 500 (dark grey), x = 1000 (middle grey), x = 1500 (light grey) and an averaged sum of all rows (blue).
LL: the background-subtracted spectrum of the brightest science source. This spectrum is computed by the post-pipeline script QCsinfo_sci.py. The green line is the peak flux extracted from each cube plane at the position of the brightest source, whereas the blue line is the average flux over the FWHM measured for the brightest source. The red line is an estimate of the flux rms.
LR: a logarithmic histogram of the OBJECT_NODDING_STACKED full detector image is shown as a dark blue line, while the reflection of the non-positive flux histogram is shown in light blue.

QC Report 3 (only created if an over-flux is detected!)

QC Report 3
TOP: a logarithmic histogram of the OBJECT_NODDING_STACKED full detector science image with the persistence critical region shown as a red dotted line.
LL: the OBJECT_NODDING_STACKED full detector intermediate product image.
LR: an estimate of the location of possible persistence critical pixels.