A Stellar Photometric Catalog

What shall we do with `bright' time? The concepts of `bright' and `dark' time have a somewhat different meaning for the SDSS telescope than for most other optical telescopes, since the telescope is designed to make efficient use of partial nights. The incremental costs involved in doing something with the bright time are therefore less than they might be elsewhere, where the telescope would otherwise be shut down for part of each month. However, any such work would be subject to the very necessary constraint that it use exactly the same hardware and real-time software as does the survey. This appendix discusses how we could go about making a complete stellar photometric catalogue for the northern sky using some of the bright time which is not useful for the main goals of the survey. This project, suggested by B. Paczynski, falls outside the goals of the SDSS and would require additional funding; we include it as an example of the other uses to which the SDSS instrumentation might be put.

Table B.1: First bright pass, 0.5m effective aperture, bin 3x3.
STAR: Spectrum f_nu = constant. Sky(V)= 19.7 Time= 7.0 sec
Filter	u'	g'	r'	i'	z'
Star saturates at V=	8.1	9.7	9.8	9.4	7.9
Colors	0.00	0.00	0.00	0.00	0.00
Eff Sky Brt, mag sec^-2	20.12	19.77	19.15	18.19	16.55
Sky+bkg count/pxl	21.36	126.82	250.94	420.99	468.10

Filter	u'		g'		r'		i'		z'
m_v	count	S/N	count	S/N	count	S/N	count	S/N	count	S/N
12.0	26345	160.9	112888	334.3	126478	353.0	87259	290.5	21576	136.8
12.5	16622	127.2	71228	264.8	79802	279.2	55057	228.5	13614	104.7
13.0	10488	100.3	44942	209.4	50352	220.3	34738	178.8	8590	78.8
13.5	6617	78.7	28356	165.2	31770	173.1	21918	138.9	5420	58.1
14.0	4175	61.4	17892	129.8	20045	135.3	13830	106.6	3420	41.8
14.5	2634	47.5	11289	101.3	12648	104.8	8726	80.6	2158	29.2
15.0	1662	36.2	7123	78.5	7980	80.3	5506	59.7	1361	20.0
15.5	1049	27.1	4494	60.0	5035	60.4	3474	43.2	859	13.3
16.0	662	19.9	2836	45.1	3177	44.6	2192	30.5	542	8.8
16.5	418	14.2	1789	33.2	2005	32.1	1383	20.9	342	5.7
17.0	263	9.9	1129	23.8	1265	22.5	873	14.0	216	3.6
17.5	166	6.7	712	16.7	798	15.3	551	9.3	136	2.3
18.0	105	4.5	449	11.4	504	10.3	347	6.0	86	1.5
18.5	66	2.9	284	7.6	318	6.7	219	3.9	54	0.9
19.0	42	1.9	179	5.0	200	4.4	138	2.5	34	0.6
19.5	26	1.2	113	3.2	126	2.8	87	1.6	22	0.4
20.0	17	0.8	71	2.1	80	1.8	55	1.0	14	0.2

Table B.2: Second bright pass, 0.1m effective aperture, bin 5x5.
STAR: Spectrum f_nu = constant. Sky(V)= 17.7 Time= 3.5 sec
Filter	u'	g'	r'	i'	z'
Star saturates at V=	1.9	3.5>	3.6	3.2	1.7
Colors	0.00	0.00	0.00	0.00	0.00
Eff Sky Brt, mag sec^-2	18.12	17.77	17.15	16.19	14.55
Sky + bkg count/pxl	2.64	15.68	31.03	52.06	57.89

Filter	u'		g'		r'		i'		z'
m_v	count	S/N	count	S/N	count	S/N	count	S/N	count	S/N
7.0	51635	222.9	221261	467.7	247896	494.8	171027	408.8	42290	196.1
7.5	32580	175.1	139606	370.3	156412	391.6	107911	322.6	26683	151.8
8.0	20556	136.8	88085	292.6	98689	309.3	68087	253.7	16836	116.1
8.5	12970	105.9	55578	230.5	62269	243.5	42960	198.3	10623	87.3
9.0	8184	81.0	35067	180.8	39289	190.7	27106	153.8	6702	64.2
9.5	5164	60.9	22126	140.9	24790	148.3	17103	117.8	4229	46.1
10.0	3258	44.8	13961	108.7	15641	114.1	10791	88.8	2668	32.2
10.5	2056	32.2	8809	82.7	9869	86.5	6809	65.6	1684	22.0
11.0	1297	22.5	5558	61.8	6227	64.3	4296	47.3	1062	14.6
11.5	818	15.3	3507	45.1	3929	46.7	2711	33.2	670	9.6
12.0	516	10.2	2213	32.1	2479	33.1	1710	22.7	423	6.2
12.5	326	6.7	1396	22.3	1564	22.8	1079	15.2	267	4.0
13.0	206	4.4	881	15.1	987	15.4	681	10.0	168	2.5
13.5	130	2.8	556	10.0	623	10.1	430	6.5	106	1.6
14.0	82	1.8	351	6.5	393	6.6	271	4.2	67	1.0
14.5	52	1.1	221	4.2	248	4.3	171	2.7	42	0.6
15.0	33	0.7	140	2.7	156	2.7	108	1.7	27	0.4

Several of us have been concerned for some time that the survey photometric system is a `faint-object-only' system, since the survey saturates at about 14th magnitude, and nearly all the stars that anyone knows anything about in detail are brighter than that. A few days of bright time could be used to produce a complete photometric catalogue on our system of brighter objects by combinations of stopping the telescope down, scanning fast, and binning on the CCDs. We would degrade the image size somewhat to alleviate the saturation problem and also to obtain better sampling, and a good way to do that is to make up the aperture by using several small holes, separated by of the order or more than the Fried parameter; in this way we will make images with the diffraction limit of the small holes and can make good but larger images. If one looks in detail, it probably requires two passes, the first with the telescope diaphragmmed down to about 0.5 m aperture, binning 3x3, and scanning 8 times the sidereal rate, with 2 arcsecond images, the second with about 0.1 m effective aperture, binning 5x5 and scanning at 16 times the sidereal rate with about 5 arcsecond images. Tables B.1 and B.2, which are in the same format as the corresponding Table 8.1 for the full survey, tell the story. The main photometric survey catalogs stars between about 14^m and 23.5^m . Table B.1 shows the limits and signal-to-noise values for the first bright pass, which catalogs stars between 10^m and 18^m . Table B.2 summarizes the second bright pass, measuring stars between 3.5^m and 10^m . The first and second steps produce photometry for about 2 million and 80,000 stars, respectively, and the first step yields astrometric accuracy comparable to that for the survey itself. The stars are almost the same set as the `overlap' astrometric stars, and this is an independent check on the accuracy of that system.

The data rates are about the same (a bit smaller) than the survey rates, and the total data about 18 percent of the imaging survey. All one is interested in, however, is the star catalog for objects with a very high signal-to noise ratio (higher than about 100:1), so the processing will be a very simple subset of the survey procedure, and the catalog will require about 30 bytes per object, or about 60 MBytes total.

The first step requires about 12 photometric bright nights, the second about 6, to cover the survey region; it would, of course, make sense to do this over the whole available sky, which would increase the time requirements about a factor of 2.5.

First Bright Pass

This pass carries us to about 10th magnitude and has a signal-to-noise ratio of 100:1 at the saturation level of the survey ( 14.5^m ). The effective aperture is 0.5 m, made up of 150 4 cm holes; the average separation is 16 cm, which may be a little small (it should be larger than the Fried parameter). The diffraction limit is about 2 arcseconds; we bin 3x3 for 1.2" pixels and scan at 8 times the sidereal rate (the data rate is a bit slower than the survey rate). The calculations in Table B.1 were done with a sky at V = 19.7^m , two magnitudes brighter than the dark sky, but in fact the sky could be a bit brighter yet before any serious effect is felt in the signal-to-noise ratios. It requires about 12 bright photometric nights.

Second Bright Pass

This pass carries us to about 3.5th magnitude and has a signal-to-noise ratio of 100:1 at about 10th magnitude; the effective aperture is 0.1 m, made up of 40 1.6 cm holes, with 30 cm average spacing. The diffraction limit is about 5 arcseconds; we bin 5x5, and run at 16 times the survey rate. It requires about 6 bright photometric nights to finish. These calculations were done with a sky at 17.7V, four magnitudes brighter than dark, but again it could be much brighter yet before it made itself felt significantly. The catalog will contain about 80,000 stars. It requires about 6 bright photometric nights.

The Rest

There are only about 100 stars left after the second bright pass, and they could easily be measured by the Monitor telescope (itself stopped down) in a couple of moonlit nights.