Chunk 4 Target Selection Release Notes
December 23, 1999
Chunk 4 plates were targeted and designed in Dec 1999 for use in Feb,
March, and April 2000. A few Chunk 3 tiles were also added to the mix
to give a good range of hour angles. Note that the objects in chunk 3
are targeted from an older set of target parameters (ran in October).
These plates have been included to cover an RA range in late January
added by Steve Kent
Here is a summary of things that may be of interest.
Target and tiling
-----------------
Targets were tiled for all OK_SCANLINE objects (primaries and secondaries).
Below are the object counts for the chunk 4 plates which cover runs 752/756.
The first columns show the 'objType' as indicated in the plPlugMapP
file and the number of targeted objects with that main type assigned a
fiber. All 640 fibers on 50 plates are accounted for
once in this list.
The third and fourth columns show the breakout of the
primTarget and secTarget bit flags. Here there is 'multiple counting'
in many cases, as an object may be both a GALAXY and a GALAXY_RED or
both a QSO and a STAR_BHB for example.
In general the numbers look good, however to get 'uniform' sky
coverage, we had to go with a higher number of tiles than we should
for this part of the sky. This area is 227.5 square degrees covered
by 50 tiles = 4.55 square degrees per tile, where as a tile is
1.49*1.49*3.141=7 square degrees in area, indicating we have 54%
overlap by one method of counting. For the actual survey we were expecting
to have quite a bit less overlap. By this accounting method,
a hexagonal tiling of unit circles has area 3.141/circle inscribed around
hexagons with area 3*sqrt(3)/2 = 2.59 or 21% overlap.
Some of this problem should be eliminated when
we tile 'fuller rectangles of sky'. Up till now we have only been tiling
one 2.5 degree wide stripe. At a minimum we should tile 3*2.5=7.5
degree wide three stripe segments. There may also need to be adjustments
to the black box tiling algorithm if we cannot get the overlap
down. The main problem noticed in tiling was that if the
required efficiency was too low that the tiled plates left 'bands' of open
untiled galaxies (which would require severe masking of the
the Large Scale Structure window function).
See the file
http://www-sdss.fnal.gov:8000/~sdssdp/target/chunk4/tile/tile88/p113.ps
for example.
Culling
-------
A number of fields had an inordinately high number of QSO candidates, and
typically also a high number of star and serendip candidates. Most of
the fields have particularly poor seeing, which likely caused the
higher target rates. As mentioned by others, it would be better to put
seeing criteria into the target selection routines, rather than culling
the lists "by hand" after target selection has run.
It was decided that non-galaxy QSO candidates in fields with 5 or more
optically selected candidates would be culled. This resulted in the following
list of fields for which QSOs, SERENDIP_RED, _BLUE, _DISTANT,
STAR_SUB_DWARF, _WHITE_DWARF, and _CARBON were removed. All are in run
752 (none in 756).
Col 1:
47 52 64 85 87 114 175 196 198 202 204 217 231 233 234 238 245 256 257 474
475 479 480 481 482 483
Col 2:
202 474 475 480 482 483} {
Col 3:
58 474 476 480 482} {
Col 4:
47 173 474 480 481 482 483} {
Col 5:
39 47 181 196 236 247 256 474 480 481 482 483 607} {
Col 6:
476 480 482 483
We did not cull objc_cols 1025 in col 5 (where X. Fan has noticed problems)
as we didn't see an excess of targeted objects there. (However this
may reoccur, so watch for it.) We did not put in any checks for the QSO
targets and the other types listed above which were also targeted as some
other non-tiled target type. So, for example, a QSO target in one of the
bad fields which was also targeted as a STAR_CATY_VAR would still have been
culled.
Completetile and Plate
----------------------
Changed complete tile and plate, the main additions were
to allow even distribution of 32 sky fibers (1 per bundle)
across the plate rather than 50 skys. Also the spectrophoto
stds were raised to 8 per plate and were evenly distributed.
Tiling used parameters: 92% efficiency (92% of all tilable stars,
qsos, hot_stds and brown_dwarfs [still being tiled] get a fiber).
This is less than the nominal 95% since we are tiling only one 2.5
wide stripe on these 3 degree wide plates. Also a plate radius of
1.49 degrees was used and a maximum allowed number of 588
tiled objects per plate (640 - 32 skies - 8 spectro_photo stds
- 8 reddening stds - 4 min star/seren/rosat)
For many plates all 588 were taken by galaxies/QSOs, leaving only 4 fibers for
rosat/stars/serendip. For many plates however, of order 100 fibers were
left over for rosat/stars/serendip. So while the distribution of galaxies and
quasars is quite uniform, the distribution of the other types
can be quite clumped. This is as expected, though more extreme
than it will be for the actual survey.
Dan Vanden Berk
Brian Yanny
Bruce Greenawalt
=============================================================
50 plates, covers runs 752/756, RA 145-236
(227.5 sq degrees 91x2.5).
ObjType Number PrimTarget/SecTarget bit set
=============================================================
GALAXY 23867 105/square degree, 478/plate
GALAXY 21607
GALAXY_RED 5340
GALAXY_BIG 45
GALAXY_BRIGHT_CORE 2
GALAXY_RED_II 362
QSO 3932 17.3/square degree, 79/plate
QSO_SKIRT 3274
QSO_HIZ 786
QSO_FIRST_SKIRT 192
QSO_FAINT 600
HOT_STD 131
HOT_STD 132
STAR_BHB 616
STAR_BHB 836
STAR_BROWN_DWARF 19
STAR_BROWN_DWARF 19
STAR_CARBON 14
STAR_CARBON 18
STAR_CATY_VAR 5
STAR_CATY_VAR 11
STAR_SUB_DWARF 7
STAR_SUB_DWARF 9
STAR_WHITE_DWARF 125
STAR_WHITE_DWARF 154
REDDEN_STD 400
REDDEN_STD 716
ROSAT_D 164
ROSAT_A 39
ROSAT_B 106
ROSAT_C 108
ROSAT_D 294
SERENDIPITY_BLUE 169
SERENDIP_BLUE 2059
SERENDIPITY_FIRST 148
SERENDIP_FIRST 388
SERENDIPITY_DISTANT 136
SERENDIP_DISTANT 375
SERENDIPITY_RED 74
SERENDIP_RED 85
SERENDIPITY_MANUAL 19
SERENDIP_MANUAL 21
SPECTROPHOTO_STD 400
SPECTROPHOTO_STD 415
SKY 1600
SKY 1600
QA 174
======================================================
Sum 32000 (=50*640)
======================================================
NA (Other) 1824 (Light Trap + Quality + Alignment + Guide Star)
GUIDE_STAR 577