Requiem for Relativity

Barbarossa & Frey images found on 1997 optical IR sky survey

The online optical infrared sky survey from 1997 (my "D" plate) has dots at 11:21:39.99,-8:34:09.6, and 11:21:01.25,-8:50:20.4. There's nothing, or almost nothing, there on either the 1987 or 1954 Red sky surveys. Neither the IRAS nor the Hensley infrared catalogs list anything within an arcminute of either spot. The NASA "2MASS" maps show nothing for J, H, or K bands at either spot.

If these are Barbarossa and Frey, resp., then the c.o.m. is only 9" away from the prediction based on the above computer program (near-circular orbit through 1954, 1986 & 1987 c.o.m.). This 1997 Frey, now on the west (back) side of the binary orbit, is almost halfway around the binary orbit (see above graphical construction) from 1987. The line between the 1954 & 1997 Freys is almost parallel to the line through the 1986 & 1987 Freys; this parallelism improves greatly with parallax correction (to account for the solar orbital motion).

Frey is 80" southeast of the position which would have given a perfect c.o.m. Again, this displacement is about parallel to the 1986-1987 (or 1954-1997) line. This also is consistent with my theory that "Frey" is really "Freyprime", a satellite of the main massive moon, Frey.

Both these dots lack nebulosity and closely resemble the images of nearby catalog stars both having "I" mag = 18.3. If the Red, and Optical IR, albedos were equal, then because of the sun's color magnitudes, both dots would have Red mag = 18.6, consistent with other identifications.

Update March 11, 2009: I had seen a third "disappearing dot", about equal to the above two, at 11:20:59.96, -8:48:15.5 on the 1997 Optical IR SERC plate scan, but though it wasn't on the Red sky surveys, I hadn't mentioned it because there is some nebulosity on the IR plate. Today I discovered the "DENIS" CCD Optical IR survey, available through SIMBAD! Its effective wavelength is 793nm vs. 807nm for the SERC, practically the same, though DENIS seems slightly less sensitive & its limiting magnitude is said to be 18.5.

None of these three disappearing dots on the SERC I (I = Optical IR) plate, appear at all on the DENIS I image! Now it seems that one or the other of the two more southerly disappearing dots, is Frey, and the other is a companion to Frey as seen in 1954, 1986, and perhaps (very close, merged image) in 1987. It seems that Frey's companion(s) are < 2' from Frey, usually almost straight N or S. With the 40::1 mass ratio, the choice of which dot is Frey, affects the c.o.m. position < 2'/40 = 3".

Unfortunately these small DENIS photos miss Barbarossa's theoretical position. The eastern photo is too far east and the western photo, corrected for Barbarossa's orbital motion and for the big Earth parallax in the DENIS photos (they weren't taken near opposition) too far west.

Another U. of Iowa Photo

I have another photo, Feb. 22, 2009, approx. 0600 to 1100 UT, from the U. of Iowa's 15 inch robotic telescope in Arizona. It is a twin set of photos, one set centered on Barbarossa's predicted position and one on Frey's. Each set is ten two-minute exposures with clear filter. The exposures in each set span five hours (but total only 20 min). I've ordered another such twin set of photos for tonight, i.e., early Feb. 23, UT.

So far I've only found out how to access the photo featured on the U. of Iowa website, which is the last of the ten Barbarossa photos. (There are two contradictory lists indicating 19, or 20, photos total). It's labeled "Barbarossa"! Hooray! The star field near Barbarossa's predicted position, shows that Blue magnitude is a better predictor than Red magnitude, of whether a star is seen on the photo. The Blue cutoff magnitude is about 20.3.

[Details: The catalog star at 11:26:39.5,-9:19:48 according to the 1987 Red sky survey, is barely visible in this U. of Iowa two minute exposure; its mags are R2=19.2, R1=19.6, B2=20.1. The star at 11:26:36.8,-9:19:56, has no trace I can see in the U. of Iowa exposure; its mags are R2=18.5, R1=19.0, B2=20.5. From my recent experience correlating R2 & R1 mags with recent photos, I know that R2 is more reliable, so, the true Red mags of these stars might be 19.3 & 18.6, resp. Though USNO-B Blue mags are said to be unreliable if >21, nonetheless, a star with Red mag 18.6 & Blue mag 20.5 isn't seen, but a star with Red mag 19.3 & Blue mag 20.1 is seen.]

What's the Blue magnitude of Barbarossa or Frey? They're about the same, but I have the most information handy for Frey. My best comparison estimates from the three Red sky surveys are 18.3, 18.55, & 19.4. From the stacked Dec. 22 U. of Iowa photo, using textbook color corrections, I estimated Red mag 19.0 (assuming 0.1 mag brighter than the comparison star). From the 1997 Optical IR sky survey, I estimate I=18.3, which I can color correct to R=18.6. So my best estimate of Frey's Red mag, is 18.77. Yet Frey's Red mag should be 1.2 (effective Wickramasinghe & Hoyle estimate for B-R for reddish KBOs) + 1.0 (Sun's B-R) = 2.2 brighter than its Blue mag. Thus Frey's Blue mag is 21.0. This is 0.7 mag dimmer than can be seen on one two-minute U. of Iowa exposure.

I don't know the mathematical details of the "median stacking" that was done, but it's a good guess that median stacking (not summation) increases sensitivity as sqrt(N). That is, the ten stacked photos of Dec. 22 would be 1.25 mag more sensitive, i.e., 20.3+1.25=21.55 cutoff. I saw Frey, because 21.55 > 21.0.

Update Feb. 23: another twin batch (another 19 or 20 photos, again two contradictory lists) was taken last night, spanning about the same interval. Again, the last taken of Barbarossa, is featured on the website and I was able to look at it. The B2 = 20.5 star was visible this time, but anything less, wouldn't have been distinguishable from the background.

Sky Surveys Alone are Proof Beyond Reasonable Doubt

I now have (see above) four Barbarossa-Frey pairs on Red and Optical Infrared sky surveys (online scans of photographic plates). All these are obvious, fairly starlike non-nebulous "disappearing dots" of comparison Red magnitude 18 to 19.5. I corrected an error in my program posted above (ambiguous variable names in the 1954 Earth position module). Now I find that when I choose three solar orbital parameters (the mass ratio, radius, and dr/dt) appropriately, I can make the unexplained acceleration disappear, either for the first three pairs, or for the last three pairs. This alone might not be surprising, because often three parameters can be chosen to make three functions (in this case, the components of the unexplained acceleration) disappear.

The proof of my theory, is that if these dots are random, there is no reason for the mass ratios required, to be the same, or for dr/dt to be the same, or for r to be better than roughly the same (I did use rough estimates of r, to choose which fields to study). The results prove that the dots are not random. For the set of the first three pairs, and also for the set of the last three pairs, the mass ratio, found by searching "parameter space", which minimizes the unexplained force (reducing it to a small fraction of solar gravity, and with a perfectly coplanar orbit), is 0.975::0.025, to the nearest 0.005. For the first three vs. the last three, the best dr/dt differs less than 20% and the best radius differs less than 0.5% (but see paradox below).

Where are you now, "scientists" who (willfully ignorant of 99% of my argument) proclaimed that this is "pseudoscience", and that these dots are "hot pixels"? "Scientists" who then censored me from their messageboards (Dr. Van Flandern's is the only major astronomy messageboard which did not censor me)? Would any of those who own the ALPO messageboards that censored me, like to place some money with a third party, to bet on the correctness of their dismissive opinions? Where are you hiding now, journal editors, none of whom even condescended to send my letter for peer review?

For convenience, the Barbarossa/Frey pairs I found on sky surveys are reposted here (for some of them I have, in my latest version of the above program, slightly more accurate coordinates, from my recent re-reading of the sky surveys, from files with built-in coordinates):

Plate A, 1954, Red
Barbarossa (object A) 11:03:12.4, -5:58:9
Frey (object A2) 11:02:25.47, -5:56:12

Plate B, 1986, Red
Barbarossa (object 11:16:55.78, -7:55:14
Frey (object B3) 11:16:51.67, -7:49:40.4

Plate C, 1987, Red
Barbarossa (object C11) 11:18:00.41, -8:1:57.7
Frey (object C) 11:18:03.18, -7:58:46.1

Plate D, 1997, Optical Infrared
Barbarossa (object D4) 11:21:39.99, -8:34:09.6
Frey (object D5) 11:21:01.25, -8:50:20.4

One paradox must be resolved here. Though the mass ratio and dr/dt are the same, for the (overlapping) first and last solar orbit segments (i.e., 1954-1986-1987 and 1986-1987-1997) the radius fitting the beginning of the last segment, doesn't match the end of the first segment. Instead, it breaks and starts over at the same radius at which the first segment started. The reason for this, is that the solar orbit really is nearly circular, but not without unexplained force. I was adjusting dr/dt to provide a fictitious explanation for a secular change in angular speed. (I also tried adjusting d2r/dt2, but found that it was small and of negligible effect.)

The angular speed changed *as though* dr/dt were 64 & 74 AU per (2800yr/(2*pi)) for the first & last segments, resp. Though consistent with a believable eccentricity (> 0.3) this 15% difference, in so short a time, and with a best fitting d2r/dt2 of nearly zero, is more evidence that the angular speed change isn't due to eccentricity. It's due to the acceleration of Barbarossa & Frey by an undiscovered distant second moon, Freya.

If Freya lies near Barbarossa's orbital plane, then acceleration perpendicular to Barbarossa's orbit is small. The observed acceleration, equal to what would be caused by dr/dt = 1/3 radius per radian, equals 2/3 of the Sun's. The most conservative Jacobi-like estimate of maximum obligately stable binary orbital radius for Freya, is (0.01/81)^(1/3)*200 AU = 10 AU (Prof. Mutel of the U. of Iowa forwarded this formula to me). So, if at the maximum stable radius, Freya needs only 2/3/20^2 = 0.0017 solar mass.

The A, B, C & D pairs are enough to support my graphical construction above for the Barbarossa/Frey binary orbit, according to which a=1.44 AU, P=22 yr. The Freya binary orbit, with a<10, would have P<400 yr. Suppose P=240 yr. The phase would change 30deg in 20yr; if it starts at maximum acceleration, that would change 1/7 in 20 yr. So the 15% change in angular acceleration, in 20yr, from midpoint 1954-1987 to midpoint 1986-1997, is plausible.

Now that the foregoing paradox is resolved, I'm able to predict positions. The midrange radii for the first & last segments, were 211.9 & 210.9, so let's use r = 211.4 AU, and dr/dt = 0. The angular acceleration increased from 64 to 74 of my units (see above) which I can extrapolate to about 79 units for the interval 1997-2007 or 2009. This leads to the prediction:

0h Feb. 23, 2009: Barbarossa/Frey c.o.m. (geocentric)
11:26:33.95, -9:15:17.9

With only (typically) 10 arcmin apparent difference between Frey & Barbarossa, and a mass ratio 40::1, Barbarossa typically would be only 15" from the c.o.m.

This further westward and northward change in the estimate, causes almost all the prospective photos taken so far, to be obsolete. I'll request more photos from the U. of Iowa Rigel telescope, today.

Results of two photo searches: telescopes 16 inch & larger

<blockquote id="quote"><font size="2" face="Verdana, Arial, Helvetica" id="quote">quote:<hr height="1" noshade id="quote"><i>Originally posted by Joe Keller</i>
<br /> Sky Surveys Alone are Proof Beyond Reasonable Doubt

I now have (see above) four Barbarossa-Frey pairs on Red and Optical Infrared sky surveys...

The midrange radii for the first & last segments, were 211.9 & 210.9, so let's use r = 211.4 AU, and dr/dt = 0. The angular acceleration increased from 64 to 74 of my units (see above) which I can extrapolate to about 79 units for the interval 1997-2007 or 2009. This leads to the prediction:

0h Feb. 23, 2009: Barbarossa/Frey c.o.m. (geocentric)
11:26:33.95, -9:15:17.9 ...
<hr height="1" noshade id="quote"></blockquote id="quote"></font id="quote">

To get a box, I changed two of the assumptions above. For each of the segments 1954-1987 and 1986-1997, I found, that constant radius which, with the 0.975::0.025 mass ratio, gave a perfectly straight arc; these radii were 214.5 & 210.1 AU, resp. Rather than average them, I extrapolated them, getting 207.9 AU to use for my 1986-1997-2009 prediction. This alternative treatment of the radius estimate, changed the predicted 2009 position only a few arcseconds; i.e. the equations form a poorly conditioned system, if solved for radius.

The other change in assumptions, was that I averaged the *apparent* dr/dt (i.e. secular deceleration due to distant third body) rather than extrapolated it. This changed the prediction more. Both alternative assumptions together give a new prediction

0h Feb. 23, 2009: Barbarossa/Frey c.o.m. (geocentric)
11:26:35.37, -9:15:29.8

These predictions differ only 0.4 arcminute, in 12 yr.

I'm still working on stacking the sets of U. of Iowa photos. Aside from these, I have two photos from other sources. The first is from an unknown source at (? midpoint 7h UT) 2/16/2009, 400+ stacked exposures taken during a four hour interval with a telescope &gt; 16 inches and Red filter. The second is from Joan Genebriera at 1h 2/23/2009, a single unstacked two minute exposure with a 16 inch telescope and Red filter, from Tacande Observatory on Tenerife.

Both these photos were aimed at my older coordinates, calculated before I found the error in my computer program. The predicted position of the c.o.m. is just beyond the north edge of the 2/16 photo and just within the edge of the 2/23 photo. However, objects likely to be Barbarossa & Frey were found in the 2/16 photo. (I found Barbarossa; the man who gave me the photo, already had found Frey, which I rediscovered independently.) The new prediction is slightly nearer what I found; this prediction is

7h Feb. 16, 2009: Barbarossa/Frey c.o.m. (geocentric)
11:26:41.67, -9:15:58.3

On the 2/16 photo, I found a density, comparable to that formed by stars of Barbarossa's expected Red magnitude, at

11:26:40.28, -9:16:59.9

(Neither photo has built in coordinates, so I measured by eye and with a ruler, comparing to the 1987 Red sky survey.) This density is near the breakup of the photo at its north edge and therefore is questionable. From the Frey position below, and the 40::1 mass ratio, I expect Barbarossa to be 0.43s W & 4.2" N of the c.o.m.; the measured position therefore is 0.96s W & 65.8" S of predicted.

However, the "Frey" (or "Freyprime") in the 2/16 photo, falls within 10" of the line through the 1986 & 1987 Freys, each referred to its own Barbarossa as origin. This is consistent with a binary orbit viewed nearly edge-on. Frey falls on this line, when the observed, not predicted, Barbarossa is used as origin; and with the solar orbital parallax correction. This is evidence that both the Frey and Barbarossa detections are real.

The observed 2/16 Frey is at

11:26:59.0, -9:18:47.5

It is elongated to about 6 arcsec, like the 1987 sky survey Frey, but whereas in 1987 it was horizontal EW, on 2/16/2009 it was sloped 60deg downward NE-SW. These objects would move almost 3" in 4hr, so Frey and Barbarossa likely really are as bright as nearby stars of Red mag ~19 which appear slightly brighter.

Next I looked for Barbarossa & Frey on Genebriera's 2/23 photo. This unstacked photo has more hot pixels and the like, so I studied densities morphologically resembling the faintest (Red mag ~19) sky survey stars that could be seen on the photo. I used my "new" prediction above to estimate the c.o.m. motion, which was so small in &lt; 7d, that almost any prediction would have been accurate enough. I also estimated Frey's binary orbital motion according to my previous graphical construction; Barbarossa's binary orbital motion would be 1/40 of that. The predictions for Genebriera's photo are

1h Feb. 23, 2009: Barbarossa (geocentric)
11:26:33.93, -9:16:30.8
Frey (geocentric)
11:26:53.6, -9:18:37.4

The predicted Barbarossa position, is ~10" within the north edge of Genebriera's photo. On this JPG image, I see starlike but very faint density there, possibly consistent with the more definite images of other Red mag ~19 stars farther from the photo's edge.

The predicted Frey position, corresponds to a starlike image, slightly fainter than a nearby star of Red mag ~19. The observed "Frey" is at

11:26:53.7, -9:18:43

so is only 1.5" E & 6" S of predicted. Even this small displacement is consistent with my graphical orbital theory, because it is along a slope of ~75deg, not far from the 55deg slope of this part of Frey's binary orbit. I've explained the failure of the binary orbit to obey Kepler's second law, as the orbit of a visible "Freyprime" around a massive "Frey". In this theory, Freyprime would move as much as 12" per week, mainly parallel to the binary orbital track.

The U. of Iowa photos (I now have most of these files: ~10 per night, two minute exposure, robotic 15 inch telescope, clear filter, southern Arizona) taken 2/22-2/25, used a more recent aim point, so the positions of interest will be well away from the edge. Studying them is next on my list.

Barbarossa likely appears on Steve Riley's March 29 (22:00-24:00 Pacific Daylight Time) 2007 Photo (Frey not within frame)

In 2007, Joan Genebriera (Tenerife, 16 inch) and Steve Riley (California, 11 inch) took photos at my coordinates. Genebriera produced four photos that were of excellent quality, and Riley six. According to my new orbital calculation, only one of these ten, which happens to be one of Riley's, is relevant now.

My newest method of predicting position, is slightly closer to the observed Barbarossa of the Feb. 16, 2009 photo, so I used it to predict geocentric c.o.m. coords. for Riley's observation time. Then I used the observed Barbarossa & Frey candidate positions on the 2/16/09 photo, to find the observed c.o.m. assuming the 39::1 mass ratio, and found the difference between this observed c.o.m. and predicted for that time.

My extrapolation method implies a quadratic error growing from 1997, so the 2007 error should be only (10/12)^2 as big. Subtracting this error from the prediction, gives a corrected c.o.m. prediction (essentially an interpolation between 1954, 1986, 1987, 1997 and 2009) of

11:25:12.00, -9:06:05.0 for Riley's 3/29/07 (midpoint 23h PDT) photo

I have only a JPG version, lacking built-in coords., of this stacked, clear filter photo. There is a star visible at 11:25:11.75, -9:06:18.8; its USNO-B catalog magnitudes are R2=19.22, B2=20.55. Studying two, two minute, 15 inch telescope, single clear filter CCD exposures from the U. of Iowa, I noticed that B2 mag is the best predictor of visibility (the cutoff was B2=20.3 on one photo and &gt;20.5 on the other). The detection I suppose to be Barbarossa, is slightly fainter than the star, and at

11:25:13.74, -9:06:37.85

Assuming its Blue magnitude is 20.6, its Red magnitude would be 20.6 - 1.2 (implied B-R from Wickramasinghe & Hoyle KBO color) - 0.3 (Sun color) = 19.1, consistent with observed sky survey Barbarossa & Frey Red mags. The star is, like other nearby objects, stretched one or two arcsec NS, presumably due to Declination error, but Barbarossa is symmetrical. Maybe the 1" EW stretch during Riley's 1.7hr observation interval, was enough to symmetrize Barbarossa's image.

The difference between this Barbarossa's position and the predicted c.o.m. (which would be accurate to within a few arcsec, if the 2/16/09 objects, used to correct the prediction, really are Barbarossa & Frey), together with the 39::1 mass ratio (found by searching steps corresponding to 20% of this value), predicts Frey's position (it's far off the photo's west edge). My graphical binary orbital model says that in March 2007, Frey was slightly east of the northwest end of the apparent binary orbital ellipse. The model thus implies that Barbarossa should be 34" +/- 3", from the center of mass; it is 42" from the predicted c.o.m. If I assume that Frey is exactly at the end of the ellipse, and proceed to construct its major axis, I should mildly overestimate the model's 54deg slope; indeed I find 58deg.

Barbarossa on Feb. 23, 2009 U. of Iowa Photo


[Aside. I recall three episodes of loss of signal on the radio:

1. The signal was lost near the climax of eyewitness testimony at a live U. S. Senate hearing about the Waco siege and fire. (St. Louis National Public Radio station.)

2. The signal was lost when a local Congressman, known in political circles but not by the general public, to be alcoholic, began to confess his alcoholism on a popular radio talk show. This Congressman soon afterwards was elected to a higher office. (St. Louis commercial station.)

3. The signal was lost last night on George Noory's talk show when a biographer of Tesla said, "The ether reduces the speed of light to 186,000 miles per second...". That was the last intelligible sound to come from the radio for about a minute. The station blamed it on the satellite. (WHO AM1040 in Des Moines.) ]


From the Barbarossa & Frey positions I found on the 2/16/09 photo, I got an arcminute correction to the extrapolation from 1954, 1986, 1987 & 1997 sky surveys. Simply adding this 2/16 correction, to the same extrapolation curve for 1h 2/23, gives the Barbarossa prediction I mention above, for Genebriera's photo:

11:26:33.93, -9:16:30.8

The 2/23 extrapolation curves (which include, of course, Earth parallax) move 0.98s W and 4.75" N per day. The U. of Iowa photos generally were taken between 6h & 11h UT. I add 7.5hr to get

11:26:33.62, -9:16:29.5

This includes a tiny refinement: to add, to the arcminute correction, also the estimated error that accrues between 2/16 & 2/23. From 1997 to 2009, Barbarossa's position deviated about 64" S & 20" W. Presumably this deviation is quadratic, therefore its derivative at present, is double the average derivative. So, the additional deviation in the last week should be 2*64" (resp. 20") * 1/(52*12) = 0.2" S (resp. 0.00s W).

Two days ago, the 2/23/09 U. of Iowa Barbarossa photos were stacked by Mauro Lacy. Lacy included only that majority of the photos, which seemed of very good quality. Lacy chose 2/23 because the photos of that date seem to be about the best. (Also, I had found that the last Barbarossa photo, the showcase photo, of 2/23, showed a nearby catalog star of B2 mag 20.5, which was absent on the last Barbarossa photo, the showcase photo, of 2/22.)

Estimating the center of light by eye, and using the coordinates built into the file, I find Barbarossa at

11:26:33.39, -9:16:32.8

only 3" W and 3" S of expected. This could be explained by error in locating Barbarossa on the 2/16 photo, which lacked built-in coords.

The Barbarossa image is streaked about 3 arcsec NW to SE, though this is statistical; by eye, it looks like a flying swallow. It's brighter than a nearby catalog star with B2=20.53, dimmer than one with B2=20.55 (slightly paradoxical due to the inaccuracy of the B2 mags) so I can estimate its B mag as 20.54. On the U. of Iowa photos (clear filter) I've found that USNO-B Blue magnitude is a better predictor than USNO-B Red, of whether something shows up, although Blue magnitudes in this range, 20-21, are near the "unreliability" limit, said to be 21. I haven't recently reassessed my sky survey Barbarossa mags, but it's about the same as Frey, and recently I reassessed those mags as R=18.7, s.d. 0.5, range 18.3-19.4. Because B-R=0.3 for the Sun, and assuming the "Wickramasinghe & Hoyle" reddish KBO color, one must add 1.5 to this to get B2=20.2, s.d. 0.5, range 19.8-20.9. So the position, shape (3" stretch over ~5hr) and magnitude all confirm that this is a detection of Barbarossa.