re: Iridium flare and flasher observed, TiPS, synodic anomaly

Walter Nissen (
Sat, 20 Jun 1998 20:49:08 -0400 (EDT)

Who should read this message: 
Frank Reed 
Patrick Wils 
anyone interested in the term "synodic anomaly", especially its history 
anyone interested in the term "synodic effect", especially its definition 
anyone interested in the term "synodic", in relation to satellites 
anyone who cares to see (the first?) actual use of the proper term 
  "Iridial glint" (instead of the rogue term "Iridium flare") 
anyone looking for the address of the 
  Visual Satellite Observer's Home Page 
anyone interested in the history of SeeSat-L 
anyone interested in search engines, especially the relationship to 
  neologistic behavior 
some other people of eclectic interests 
Sorry if that doesn't include you. 
Anyone who has a hard time comprehending that list should consult 
Figure 70 on page 370 of Douglas Hofstadter's Pulitzer Prize-winning 
"G"odel, Escher, Bach: an Eternal Golden Braid", sometimes cited as the 
greatest book ever written.  That figure exhibits, as its caption relates, 
`A tiny portion of the author's "semantic network"', where "author" could 
be either Hofstadter or the present writer. 
Frank Reed writes: 
> Yes, that's exactly what I meant.  In electrical engineering, we use the 
> term "phase" to denote the time offset between similar periodic waveforms. 
Well, if you want to talk engineering, throw in a few j's, phi's and 
theta's and let's do it.  Perhaps I should apologize for neglecting your 
engineering background. 
> If the common usage for the term "synodic 
> anomaly" is as Walter defines it above, that's fine with me.  However, it 
> doesn't necessarily imply (to me anyway) a _time_ shift in the cyclic 
> behavior, nor does it quite fit the definitions in my old (1950) college 
> Webster: 
"synodic anomaly" is a phrase I coined, somewhat like a neologism, to describe 
the different-length flash intervals which you and I describe, and which 
Bart and Patrick Wils have shown can be used, even with data from only a 
single OBS, to calculate the rotation axis of a spinning satellite, 
particularly a reflective (truncated) cylinder.  I believe there can be 
either one or 2 anomalous intervals separating the uniform intervals. 
Perhaps Bart would like to give some of these recent OBS a try, using his 
innovative technique. 
The term has appeared a number of times here in SeeSat-L.  Try 
"synodic anomal*" using the quotes as specified, in Compaq's AltaVista or 
HotBot or whatever.  For the time being, you won't be able to use the 
links they provide, but they provide copious clues to help you find the 
messages in the SeeSat archives, in its new location, but still below the 
VSOHP,, the 2 in that 
address being optional. 
I suppose all terminology has many horrible weaknesses, so your analysis 
doesn't really surprise me.  However, in fact, "synodic anomaly" is an 
excellent fit to earlier usages of both synodic and anomaly, besides 
making a rather precise characterization of the phenomenon. 
In describing satellite phenomena, and particularly flashing phenomena, a 
distinction is made between intrinsic phenomena, which occur because of 
some characteristic, e.g., rotation, or angular momentum about an internal 
axis, of the satellite itself, on the one hand, and, on the other hand, 
synodic phenomena, which occur because of changing circumstances during 
the progress of the OBS, particularly the changing phase angle at which 
the object is observed, due to its angular momentum about an axis through 
the center of Earth. 
E.g., the flashing of Ir 18 = Ir 20 = 24871 = 97- 34C is apparently an 
intrinsic phenomenon and this is pretty strongly suggested to an 
experienced observer by its observed appearance.  In contrast, the nightly 
Iridial glints produced by the other Iridia in working attitude in working 
orbits are synodic phenomena, which is similarly suggested to the 
Anomaly is simply the nounal form in its usual usage.  You can observe a 
hundred or more passes of tumblers before detecting a synodic anomaly, 
which spoils your period determination from multiple flashes of that 
object.  But these detections have recently been greatly assisted by the 
spread of inexpensive stopwatches with 30, 50 or 100-split capability. 
"synodic anomaly" is a fairly direct derivitive of the earlier term 
"synodic effect" which is used to describe the difference between "flash 
period" as measured by a stationary observer and "rotation rate" in 
inertial space.  That difference is between the 2 types of angular 
momentum just mentioned above.  In other words, flashes don't come at 
exactly the same rate as the satellite's rotation, because the changing 
Sun-satellite-observer angle "places the observer out-of-position" as 
successive flashes occur, preventing him from accurately measuring the 
rotation rate. 
There are many possible words yet to be assigned meanings.  We could call 
"synodic anomalies" by any number of possible terms, such as "tempoburps", 
"synodomals" or "achronicities".  But many people show resistance to such 
neologisms, while allowing "neologistic" phrases which then produce 
massive confusion in search engines.  As an aside, there are now two 
totally distinct literatures extant concerning "microsatellites", one in 
space technology and one in genetic technology.  Not too nice.  I am 
extremely interested in "friendly numbers", but, on the Web, most of the 
references are a) coincidental (e.g., "...system is user-friendly. Numbers 
of..."), b) jazz tunes, c) diet choices, d) crop reports or e) 
mischaracterized references to "amicable numbers", leaving f) the real 
references as needles in a haystack.  Replacing the general term "number" 
with the more specific "integer" will pick up two essentially distinct 
literatures, one in discrete mathematics and one in computer programming. 
Oh, well ... 
Bart writes: 
> The term 'synodic anomaly' was coined by either Patrick Wils or myself, I 
> can't remember right now, when describing Walter Nissen's observations of 
> 91- 29 B. 
It was you, Bart, who played a keystone role in the "adoption" of the term 
"synodic anomaly", but it was one of approval, not of coinage. 
On Fri, 19960412, SeeSat-L's machine went down.  My coinage of the term 
appeared, coincidently, in a message sent to SeeSat-L the next day while 
it was down and Bart was handling SeeSat "by hand".  (The protocol was to 
send SeeSat messages to Bart's personal e-mail address using a Subject: 
line which began "SeeSat:", which would cause the long-suffering Bart to 
forward it by hand to his own (slightly dated) list of subscribers). 
Maybe that message is therefore not in any of the online archives. 
Here it is, and please bear with me as I reproduce it and Bart's response 
in full, thus placing it within reach of the search engines: 
>> Date: Sat Apr 13 12:13:08 1996 
>> From: dk058@cleveland.Freenet.Edu (Walter Nissen) 
>> Subject: SeeSat: OBS for PPAS, also general interest 
>> To:, dk058@po.CWRU.Edu 
>> Reply-To: dk058@cleveland.Freenet.Edu (Walter Nissen) 
>> Walter I. Nissen, Jr., CDP,, 55 Barrett RD #808, 
>> Berea, OH 44017-1657, USA, 216-243-4980, -81d 51.823', 41d 22.413', 256m, 7x35 
>> 87- 49 B 96-03-04 10:33:17.3 WN   29.2 1.    4  7.3   cmplx, dtm, C* 1850 r 
>> Prior to the interval given for C* 1850 r, other period(s) may have been 
>> present (I have more data if anyone wants it). 
>> 88- 20 A 96-03-11  1:23:36.8 WN  149.6  .5   8 18.70  cmplx, C* 1933 
>> 95-  8 B 96-03-12 10:34:48.1 WN   81.6 2.    3 27.    M'M', C* 2306 r 
>> 88- 50 A 96-03-12 10:44:58.3 WN   92.9  .6  14  6.64  A'A', C* 1953 
>> 88- 20 A 96-03-19  0:27:38.2 WN   57.3  .4   6  9.545 F'F' w/in cmplx, C* 1933 
>> 94- 24 B 96-03-30  0:50:53.7 WN   69.4 1.5   2 35.    A'A', C* 2279 r 
>> 89- 17 B 96-03-30  1:17:28.2 WN   77.9  .5   4 19.5   A'A', C* 2004 r 
>> 92- 36 B 96-03-31  0:51:49.9 WN   66.1  .5  17  3.89  A'f'a'f'A', C* 2195 r 
>>                                       both A'f'a'f'A' and A'F'a'f'A' were seen 
>> 92- 36 B 96-04-02  1:41:12.6 WN   63.1 1.   16  3.94  F'F', C* 2195 r 
>> 92- 36 B 96-04-03  0:46:54.5 WN   63.6  .5  16  3.97  F'F', C* 2195 r 
>> 71-120 B 96-04-04  1:38:50   WN                       S, M* 1-10 r 
>> 94- 74 B 96-04-04  1:44:37.3 WN   25.2  .8   3  8.4   A'A', Resurs 1-3 r 
>> 92- 80 A 96-04-08  9:52:24.8 WN  114.2  .6   1 114.2  F'F', C* 2221 
>> 94- 72 A 96-04-11  0:48:29.0 WN   21.0 1.    2 10.5   A_'A_', C* 2293 
>> For C* 2293, I measured the dips, which were sharper, and tried to convey 
>> that by recording "A_'A_'". 
>> 95- 32 B 96-04-11  1:51:18.0 WN   15.8  .8   1 15.8   A'A', C* 2315 r 
>> 95- 32 B 96-04-11  1:51:33.8 WN    9.4  .8   1  9.4   A'A', C* 2315 r 
>> 95- 32 B 96-04-11  1:51:43.2 WN   13.1  .8   1 13.1   A'A', C* 2315 r 
>> 95- 32 B 96-04-11  1:51:56.2 WN   12.6  .8   1 12.6   A'A', C* 2315 r 
>> 95- 32 B 96-04-11  1:51: 8.8 WN   14.6  .8   1 14.6   A'A', C* 2315 r 
>> Despite the paucity of data, the large scatter in the "uniform" intervals, 
>> and the lack of real-time recognition of an anomaly, my impression is that 
>> the timings of C* 2315 r are pretty accurate and most likely display one 
>> of those synodic anomalies that Bart and Patrick Wils have explored. 
>> 94- 74 B 96-04-11  1:55:48.3 WN   35.8  .8   4  8.9   A'A', Resurs 1-3 r 
>> Cheers. 
I should perhaps mention that I made something of a bad habit of reporting 
spurious minute values in early reports of synodic anomalies.  I would 
fail to increment the minute in successive lines.  The intervals reported 
are reliable and the later minute values should be incremented to make the 
minute consistent.  I.e., there is a "51" above which should be a "52". 
Bart's response: 
> Date: Sun Apr 14 12:45:00 1996 
> From: (Bart De Pontieu) 
> Subject: SeeSat: Rotation Axis of 95- 32 B 
> To: (Bart De Pontieu) 
> Walter Nissen writes: 
> >Despite the paucity of data, the large scatter in the "uniform" intervals, 
> >and the lack of real-time recognition of an anomaly, my impression is that 
> >the timings of C* 2315 r are pretty accurate and most likely display one 
> >of those synodic anomalies that Bart and Patrick Wils have explored. 
> I just looked into the data and analyzed it with the software I 
> developed last year to look (amongst other things) at the synodic anomaly 
> you observed of 91- 29 B (and Mike McCants' obs of 92- 38 B). For those of you 
> who have never read Flash, here's a good opportunity. Check out this URL for 
> details on my method to determine the rotation axis of flashing satellites: 
> Figure 6 of that article refers to Walter's observations of 91- 29 B. 
> And check out this URL for details of 92- 38 B's rotation axis: 
> First a few remarks about the observations: 
> >95- 32 B 96-04-11  1:51:43.2 WN   13.1  .8   1 13.1   A'A', C* 2315 r 
>                                                 ^^^^ 
> Shouldn't this read '13.0', viewed the time difference between the two 
> consecutive times: 1:51:43.2 and 1:51:56.2 ? 
> Not that a difference of 0.1 will not make a big difference on the results 
> found below, but just for the record. 
> >95- 32 B 96-04-11  1:51: 8.8 WN   14.6  .8   1 14.6   A'A', C* 2315 r 
>                       ^^ 
> I'm assuming this is 1:52: 8.8. Please correct me if I'm wrong, Walter, and 
> disregard what follows if it really is 1:51: 8.8. 
> Here is what the time series I fed my program: 
>   0.00  0 
>  15.80  1 
>  25.20  2 
>  38.20  3 
>  50.80  4 
>  65.40  5 
> And the orbital elements used: 
> 1 23604U 95032B   96103.16183644 +.00000015 +00000-0 +00000-0 0 01024 
> 2 23604 082.9027 148.4072 0025263 199.0133 161.0075 13.73606201038713 
> My method basically takes each subsequent couple of timings and calculates 
> the rotation period from those two times, for *all* possible directions of 
> the rotation axis. I do this for each adjacent couple of timings, and thus 
> come up with 5 different rotation period values for each possible direction 
> of the rotation axis. Obviously, the rocket only has one rotation period. 
> So one would expect the 5 different rotation period values to become equal 
> for that direction that coincides with the actual rotation axis. In reality, 
> because of measurement limitations (0.8 accuracy!?) there isn't one 
> specific direction for which the 5 values coincide. There can however be a 
> set of directions for which the dispersion about an average rotation period 
> becomes minimal. In fact this usually happens for any set of observations 
> of any object. However, only for a few special passes during which a 
> 'synodic anomaly' (like that term, Walter :-) happens will the minimum be 
> significant. That is what I found for Walter's 91- 29 B observations last year. 
> And now also for the above observations of 95- 32 B. I found 
> one direction for the rotation axis that minimizes the dispersion around 
> the rotation period to a minimum in a quite drastic way. 
> For a rotation axis pointing to almost any direction in the sky, the 
> dispersion around the average rotation period is 4.5 seconds (considerably 
> larger than e.g. the 0.8 second accuracy Walter reports). 
> For one specific region with right ascension 194 degrees and declination 
> about 18.5, this dispersion drops to as low as 1.8 seconds. This much better 
> fit occurs only for a region of about 0.5 degree diameter about the above 
> mentioned coordinates. A better expression for the difference between 
> predicted times of flashes (from my program) and the observed times, is the 
> average absolute value of (P_rot - P_rot_average). This gives 1.3 seconds 
> for the region mentioned above, and is on average 3.8 seconds. I'd say that 
> the 1.3 seconds average difference is not in contradiction to the 0.8 sec 
> accuracy Walter reports. 
> More specificially, the best-fit-region is actually double in that two 
> directions are possible: 
>     Right ascension     Declination    Rotation Period 
>     194.0               18.2           26.25 s 
>     194.0               19.0           26.45 s 
> Mark that the rotation periods given here are probably only accurate to 
> +- 0.5 seconds. 
> Mark also that a rotation period of 26.3 seconds will usually give flashes 
> about every 13 seconds. And mark finally that Walter reports seeing flashes 
> sometimes every 9 seconds, sometimes every 15 seconds. This can only happen 
> if you see a 'synodic anomaly' because the rotation axis then coincides with 
> the bisectrix between the lines sat-sun and sat-obs. 
> Concluding, I think it's safe to say that the rotation axis of 92- 36 B 
> was pointing at alpha=194 and delta=18.5 on April 11. 
> Generally speaking it is possible to also determine the rotation axis of 
> satellites that do not show a 'synodic anomaly', given quasi-simultaneous 
> observations of that object. That is what 'DRAP' (Determination of Rotation 
> Axis Project) of the BWGS is all about. It will be revived soon, with an 
> emphasis on the accelerating satellites. DRAP will probably be expanded with 
> a 'video' side, since video-observations are more accurate. More in a few 
> weeks (hopefully). 
> Cheers, 
>   Bart De Pontieu <> 
You will note that in there Bart writes: 
> However, only for a few special passes during which a 
> 'synodic anomaly' (like that term, Walter :-) happens will the minimum be 
> significant. That is what I found for Walter's 91- 29 B observations last year. 
The observations he refers to were made 19940414, see page 93 of Flash 
81-82.  (I guess it was his analysis which occurred the next year; i.e., 
"last year" modifies "found").  This was prior to the time, 199411, that 
Bart acted (a 1000 Thank You!'s, Bart) upon my suggestion to create 
SeeSat-L.  For some details of the creation of SeeSat-L, see the second 
anniversary messages in 199611. 
For more history of synodic anomalies, see Flash's #81-82, 90, 91, and 
Walter Nissen         
-81.8637, 41.3735, 256m elevation