Anticipating STS viewing conditions

Walter Nissen (dk058@cleveland.Freenet.Edu)
Sun, 30 Jun 1996 10:05:28 -0400

This is the Columbia TLE in Ted Molczan's latest: 
STS 78 
1 23931U 96036A   96179.58333333  .00001083  54364-5  46366-5 0   265 
2 23931  39.0149 323.2952 0006902 319.0054 325.9736 16.01983738  1115 
 
Here are two other typical TLEs, also from OIG, from the file sts78.tle: 
STS78 
1 23931U 96036A   96173.58333333  .00000786  54136-5  44616-5 0    70 
2 23931  39.0132   3.5128 0007725 282.2939 280.5779 16.00670877   159 
STS78 
1 23931U 96036A   96180.58333333  .00001017  54423-5  44356-5 0   289 
2 23931  39.0156 316.5857 0006804 325.1376 336.4631 16.02338754  1279 
 
Here is one from another source: 
http://shuttle.nasa.gov/sts-78/orbit/orbiter/sighting/shuttle.html 
don.j.pearson1@jsc.nasa.gov 
This vector is valid for prediction purposes until 182:17: 0:  .0 UTC 
STS-78 
1 23931U 96036A   96180.71807878  .00096478  00000-0  16223-3 0  9117 
2 23931  39.0150 315.6828 0006893 328.8907  31.1662 16.02406423  1316 
 
And another: 
http://spacelink.msfc.nasa.gov/html/Spacelink.Hot.Topics.html 
David H. Ransom, Jr.   Email:rans7500@spacelink.msfc.nasa.gov 
STS-78 
1 23931U 96036A   96181.50325579  .00089828  33585-7  14206-3 0   301 
2 23931  39.0143 310.4050 0006652 331.3657 245.5070 16.02705157  1435 
 
>From the MMs in the 96173.583 and 96180.583 elsets from OIG, the implied 
ndot2 is .00119.  The ndot2 might be expected to increase slightly as 
Columbia flies lower into the atmosphere. 
 
Here is an elset which I cobbled together to obtain a more realistic 
expectation: 
STS-78 mine 
1 23931U 96036A   96181.50325579  .00120000  33585-7  17000-3 0    0x 
2 23931  39.0143 310.4050 0006652 331.3657 245.5070 16.02705157  1435 
 
Here is QuickSat output for these six elsets for Washington, DC: 
  39.000  77.000    0.    DC <----------------- 1950 13.5  4 F F F F F 
23931 STS78                             .1 
 ***  1996 July  6  *** Times are UT ***   133  850 
 H  M  S  TIM AL AZI C   U  MAG   REVS  HGT SHD  RNG  EW PHS  R A   DEC 
 7 51 59  1.0 19 107    72   .9  110.1  267  17  706 1.0 103  157   -.8 
 7 53 15   .8 21 109    72   .7  110.1  269  17  678 1.1 101  151   -.9 
 7 54  1  1.1 21 109    72   .6  122.1  268  18  665 1.1 101  150   -.8 
 7 59 25   .1 27 116    72   .0  125.1  273  18  557 1.6  94  120  -1.1 
 8  2  3   .1 30 120    72  -.3  141.1  273  17  511 1.9  90  1 3  -1.2 
 8 16 12   .3 49 165 C  72 -1.6  237.1  276  22  360 3.6  58 2244  -1.3 
 
You can connect the various output lines to the various elsets by noting 
the REVS parameter is larger for elsets with earlier EPOCHs.  The two 
latest elsets are distinguished by the time (of the emergence from the 
shadow of the Earth); the earlier time shown is for mine (because its 
ndot2 is higher). 
 
 
Comments: 
Obtaining a very recent elset is one of the very best ways to obtain 
acceptable accuracy in producing an indication of visibility.  But that 
won't work if you are trying to anticipate what viewing conditions will be 
like a few days later.  As you can see in the output above, the times for 
all six elsets differ by a fair amount, about a minute, or more; with the 
week-old elset being off by more than 24 minutes.  Worse, all the elsets 
distributed thru the various NASA channels "predict" visibility AFTER I 
anticipate it (as given in the first line above).  This means many people 
will go out looking for it only after Columbia has already passed by. 
This pattern is long-established for STS flights.  In this particular 
case, the fact that the real appearance will be poorer in quality, i.e., 
at a lower altitude, compounds the problem.  The week-old elset "predicts" 
a passage easily observed by the general public, with little assistance 
required, just the time accurate to a minute or so (which it isn't) and a 
general direction to look.  Actually, it appears the passage will be 
difficult for all urban and suburban viewers because it will be so low. 
 
Even after pumping up the ndot2 value to obtain an earlier, and more 
realistic, time for the passage, I find it very desirable to fudge the 
output by an additional 1 to 3 minutes (earlier) to accomodate the usual 
inaccuracy of peoples' watches (and the likelihood that the typical viewer 
exiting a typical home will encounter a ringing phone, a nagging spouse, a 
dog which needs to be fed, an unexpected street light, or other diversion, 
not to mention dark adaptation). 
 
I'm not sure why NASA seems to go to so much trouble to distribute 
immensely verbose state vectors and other verbose forms of elements with 
inaccurate drag factors which make it difficult to anticipate passages 
prior to newspaper deadlines, when what the general public needs is a few 
little TLEs released in a timely fashion with reasonably good values.  Two 
significant figures in the drag factors would be a huge improvement.  We 
need only a small improvement. 
 
(On the size issue, the TLE, which appears so concise when compared with 
these other bloated forms, is itself mostly useless ballast.  Of the 
typical 170 bytes in a TLE only 30, at the most, are information bearing. 
The rest are spaces, leading zeroes, INsignificant digits, and other 
fluff.  Actual information theoretic content is often around 15 bytes.) 
 
I'm also not sure why the many people who work at NASA whose future pay 
depends on public support wouldn't be interested in correcting this 
situation ASAP.  Connecting the public to their space program would seem, 
plausibly, to be a priority in maintaining political support for it. 
 
If there are reasons why NASA doesn't want to do this on an official 
basis, I would think that there would be those willing to act as conduits 
for good info, including me, tho my health might make me a less than 
ideal choice. 
 
I welcome comments, information and contacts for better information. 
 
Cheers. 
 
Walter Nissen                   dk058@cleveland.freenet.edu 
 
--- 

The night sky is your universe.