Requiem for Relativity

Initial Analysis of Robert Turner's March 19, 2008 Photo of Frey


This is the only photo taken so far in 2008, of the Barbarossa/Frey system. It is the only color photo (not counting Red filtered sky surveys) ever taken of the Barbarossa/Frey system. It shows Frey as a bright green dot in the NE quadrant of the frame. (As this photo is shown here on Dr. Van Flandern's messageboard, S is up and W is to the right, so, the frame is upside down, but not reversed; the usual presentation is, N up and W right.) It was taken with a 14-inch reflector on Tenerife, remote-controlled under the direction of amateur astronomer Robert Turner of England, observing at my predicted coordinates for the Barbarossa/Frey center of mass.

I haven't rechecked my calculation yet, but to the best of my ability to calculate Frey's position based on Joan Genebriera's 2007 Barbarossa photo, Barbarossa's 1954 and 1986 positions, and my theory of Frey's precessing orbit, the position of the bright green dot noted by Turner in his photo, differs only 16" from expectation. This could be caused by a plausible, 3% error in my estimate of Frey's orbit. The theoretical position of Barbarossa, is off the photo, slightly E of the margin. I knew of four Barbarossa/Frey candidates on the photo before calculating the position; the random chance that one or more of these four would have been within 16" of Frey's theoretical position (Barbarossa's theoretical position is off the frame) is 4 * (pi*16"^2) / (17.1' * 17.6'), i.e., p = 0.003.

After Turner found the first disappearing dot, I found, basically, three more in the photo. These are not near the theoretical positions of Barbarossa, Frey or Freya. Near the S edge, is a medium-bright disappearing green dot (i.e., not found on the 1987 SERC-ER survey); this is W of a disappearing double red dot. In the SW quadrant is a faint disappearing red dot; its faintness, its color, and its starlike broad smooth intensity contour, make it especially likely to be a real astronomical object.

As I noted on Steve Riley's photos last year, these electronic photos, near their detection limit, give highly variable images. Stars confirmed on sky surveys, of brightness near the electronic photo's detection limit, can have a classic appearance, or a small bright or irregular appearance, or be totally absent; all on the same photo. An unusual photon distribution should be especially likely to have an unusual appearance, especially near the detection limit.

Barbarossa and Frey theoretically have about 1" diameters. A star might be approximated as a point source hopping randomly on a 1" disk, staying in each place for ~ 1/30 sec. The light from Barbarossa or Frey, composed of many point sources more-or-less evenly distributed on a 1" disk, presents the electronic camera with a drastically different object, for which light intensities over the blur circle are much more constant. This involves the subject of "reciprocity" of photographic film. Characteristics of electronic detectors, loosely analogous to reciprocity, seem little investigated. I'm wary of dismissive, unsupported claims about how an image of Barbarossa (especially, near the detection limit) should appear.

Another factor, is the now known (see my posts above) transient generalized extinction (dimming) of stars within a few degrees of the Barbarossa/Frey system. Such an interposed cloud would further alter these light sources, further vitiating any assumptions about the appearance of Barbarossa/Frey images.

The paired blink photos (posted by "marsrocks" on another thread of this messageboard) do show the abovementioned bright green disappearing dot (not in color)(in the lower left, i.e., NE corner, above & left of a small equilateral triangle of stars) disappearing as expected. The background lightness of the other photo almost obscures a nearby star of similar brightness. The blink pair might give more false positives, than comparing Turner's March 19, 2008 photo to a sky survey, without blink. Still, the blink pair gives us a dramatic demonstration.

Requiem for Relativity

Such a large mass as Barbarossa, so close to the positive "Cosmic" Microwave Background dipole, would refute the theory that the dipole is a Doppler effect. It also would refute the theory that the "Cosmic" Microwave Background is cosmic.

Likely, the CMB would be found to originate from the sun's gravitational field, and the dipole from Barbarossa's gravitational field. This would demolish present relativistic ideas about gravity and light.

That is why Barbarossa is in the thread entitled "Requiem for Relativity".

In Barbarossa's Cavern There Are No Stars (Part IV)

The Other Hubble Relation


Hubble, ApJ 56:162+, 400+, 1922, discovered that with a 60 or 100 inch telescope, the apparent photographic size of stellar "reflection nebulae" is what would be expected assuming total scattering of the central star's light at the outermost detectable edge of the nebula, where the nebula is just barely detectably brighter than background (p. 410). Hubble's empirical relation (p. 411, eq. () is:

apparent magnitude + 4.90 * log10 (radius in arcmin) = 11.0.

If all light is scattered at the effective surface of the nebula, then the star itself should be invisible. Yet Hubble was able to use the apparent magnitude of the star, to estimate how much light was available for scattering at the surface of the nebula. There's twice too much light in his model. Hubble (p. 401) even cites Hertzsprung's finding (Astronomische Nachrichten 195:449, 1913) that the effective nebular surface scatters only 1-2% of the light.

In Hubble's model, wouldn't the cloud of nebular material often be bigger than given by Hubble's relation, and continue to scatter 100% of the light? If so, then the contrast at the edge of the nebula, would be too low to detect. Suppose this were the case for many stars (after all, most stars lack detectable nebulae). Then, likely also for many stars, the cloud would be smaller than Hubble's relation, giving a smaller reflection nebula. Yet, according to Hubble's Fig. 2, p. 411, this seldom occurs.

Maybe the radius of the cloud of nebular material coincidentally (but not accidentally) happens to be proportional to (stellar mass)^1.5. For stars from 1 to 30 times solar mass (MNRAS 382:1073+, 2007, Table 6, p. 1078) there is an accurate quadratic polynomial for log(luminosity) as a function of log(mass). Using Wikipedia's 18 solar mass figure for Type B stars (more accurate would be to average Hubble's spectral type sample) the luminosity-mass relation becomes a power law with exponent 2.91. That is, Hubble's relation for the typical Type B star, is equivalent to:

nebula radius = const. * (stellar mass)^(2.91/(4.90/2.500)) = (stellar mass)^1.48

with such proportionality constant as to give that radius at which the total luminance of the star, divided by the surface area of the nebula, just exceeds background intensity.

Hubble's nebulae almost all were around stars of Spectral Type B (though one of the 33 listed in his Table III, p. 181, was K8). Other, presumably older spectral types, seem generally to lack such nebulae. Maybe a failed cold brown dwarf also would have such a cloud, never completely expelled by radiation pressure.

Assuming luminosity were proportional to mass^2.91 (typical in the Type B range), the Barbarossa system's absolute magnitude, if it could behave as a Type B star, would be roughly 4.83 - 2.5 * log(20,000 * (0.0103/18)^2.91) = +17.67. At 10pc, Barbarossa's nebula would, according to Hubble's relation, subtend a radius of 0.0435'. At Barbarossa's actual distance, that's 454' = 26 AU. This predicted angular diameter of Barbarossa's cloud, 15 deg, agrees well with the observations of transient dimming in sky surveys, discussed in Parts (II) and (III) above.

In Barbarossa's Cavern There Are No Stars (Part V)

According to Skuppin et al, A&A 177:228+, 1987, absorption lines in the spectra of fast-rotating Type B stars, are one of the best ways to quantitate local interstellar matter. They discuss three such stars, one of which is Theta Crateris. The data for Theta Crt and for one of the other stars, were obtained by the Intl. UV Explorer (IUV) satellite, Jan. 1985.

The density of interstellar material between us and Theta Crt, measured by spectral lines of four different elements, was consistently greater than for either of the other two stars; the factor generally was ~10. Also, only Theta Crt lacked the MgI line; this signifies that only Theta Crt lacked hot gas along the line of sight. So, Theta Crt has more intervening material, but in contrast to the other two stars, all the intervening material is cold.

Theta Crt is 66pc distant, so having 10x the interstellar material, gives extra Visual extinction equivalent to 600pc, i.e. roughly 0.6 mag using the usual approximate 1 mag / kpc value. If the distribution of UV extinction resembles that of Blue extinction in Part (III), then Theta Crt is where the extinction was, in 1985, increasing fastest, as the front of Barbarossa's cloud moved in. Where its derivative is greatest, the normal distribution is at 0.6 times its peak. So, the peak extinction of Barbarossa's cloud is roughly 1 mag in Visual.

According to Skillen et al, MNRAS 265:301+, 1993, the RR (type RRab) Lyrae star, W Crt (RA 11:26:30, Decl -17:55)(near Gamma Crt) underwent dimming of a portion of its V and B light curves, between 1986 & 1988. The authors say they have "no explanation for the discrepancy" (p. 303). Their variable star light curves were, as usual, calibrated against nearby reference stars. If the light of W Crt becomes polarized during this phase of its cycle, and the Barbarossa cloud preferentially extinguishes light of that polarization, the dimming might be explained. Strohmeier's text, "Variable Stars", says some stars of another variable class, the semiregular variables (Type M) do show weak polarization of their visible light.

In Barbarossa's Cavern There Are No Stars (Part VI)

Fruscione et al, ApJSupp 94:127+, 1994, Table 1, list 594 stars toward which the ("integrated") column density of neutral hydrogen, log( N (H I) ) has been determined, either by direct measurement, or by estimation from the logN determined for other atom(s). Only two of these 594 really rival Theta Crater, as evidence of nearby high extinction.

Almost 1/3, i.e. almost 200, of Fruscione's stars are <= 100pc distant (Theta Crt is 66pc). Of these, only four exceed Theta Crt's value log(N(HI)) = 20.50 +/- 0.10 (a few stars had ambiguous values such as " < 21.00"). One of these four exceeds Theta Crt slightly, but because its distance is greater, the average HI density along the path to it is less, though the sum, i.e. ("integrated") column density, is more.

The values of all four of these rivals, are less reliable than Theta Crt's. Theta Crt's value was estimated from more than one (four) other kinds of atoms. One of the others, 34 Cygni, was measured directly from LymanAlpha, which normally would be the gold standard, except that only "earlier than B2-B3" spectral types are suitable for this method (Fruscione, p. 128) and Fruscione lists 34 Cyg as B2. The other three rivals all were estimated from NaI alone; this can overestimate HI by 0.5 log units (p. 129).

Thus the only two stars (among the almost 200 at distance <= 100pc) toward which average HI density seems really higher than toward Theta Crt, are 34 Cyg and Lambda Per. 34 Cyg lies before "a heavily obscured region of the galaxy" (Burnham); it's less than a degree from the open cluster M29 (5 to 8' diam.). Lambda Per, according to Norton's Atlas, lies between an open cluster and a diffuse nebula, a degree to either side of it. Theta Crt, by contrast, lies in an almost empty field (only a small galaxy, almost 3 deg away).

Another of Barbarossa's champions is 69 Leonis (RA 168, Decl 0, distance 80pc). Welsh et al, ApJ 437:638+, 1994, Tables 2 & 4, reporting a study of interstellar NaI density, list, inter alia, 28 other stars at distances 70 to 90pc inclusive. Six of the 28 exceed 69 Leonis' value of log(N(NaI)) (none were ambiguous). Three of these six were toward the galactic center; these three also were by an investigator who performed only 19 of the 167 total determinations used. All eleven of Welsh et. al.'s own determinations, in that distance range, were at least 1.15 log units less, than for 69 Leo.

Welsh cites Albert, ApJ 272:509+, 1983, as his source for 69 Leo. Albert tested only three stars at distance <= 100pc. For TiII and NaI absorption, differences between the three, corrected for path length, did not exceed error bars. However 69 Leonis, at 80pc, showed log(N(CaII)) = 3.0 +/- 0.5, vs. (1.1 or 1.0) +/- (0.6 or 0.4) for the other two stars, at 100pc or 90pc, resp. Correction for distance, gives 4.5x the average density, along the path to 69 Leonis.