Solwind astrometry

Using three stars

Usually there are no stars visible in the Solwind images. Three stars are needed to solve for the plate constants with Turner's equations. Fortunately there are images with (at least) three stars on November 27. These stars are:

star mag R.A. 2000.0 decl 2000.0
beta1 Sco 2.60 16 05 26.2 -19 48 19
nu Sco 4.35 16 11 59.7 -19 27 38
omega1 Sco 3.94 16 06 48.4 -20 40 09

With these stars it is possible to derive the transformations between R.A./decl and the plate coordinates (x,y). With this knowledge it is easy to compute the plate scale, the orientation of the image with respect to celestial north and the position of the Sun's center in the image. I have done this with three images:

Date Time vertical scale horizontal scale PA of celestial north suncenter x suncenter y
1980 Nov 27 10:06:05 69.1 69.9 6.22 114 125
1981 Nov 27 13:45:39 70.8 70.7 5.90 117 123
1981 Nov 27 14:15:01 69.8 69.6 5.10 115 123

Time is UTC, scale arcseconds per pixel, PA in degree, suncenter with (0,0) in upper left corner .

There is a small image shift between the last two images. Compensating for this, the suncenter of the last image is at (116,123).
The plate scale is close to 70 arcseconds per pixel (and not 75 as other sources suggest).
The top of the images is closer to celestial north than to solar north (as I have assumed earlier)

On December 13 xi and theta Oph are visible in the Solwind images. In 1981 the planet Mercury is also visible in the field of view. It's mag -1.1 but not saturated. I have measured two images:

star mag R.A. 2000.0 decl 2000.0
xi Oph 4.39 17 21 00.0 -21 06 43
theta Oph 3.24 17 22 00.6 -24 59 58
Mercury image1 -1.1 17 27 52.9 -24 36 26
Mercury image2 -1.1 17 28 49.3 -24 37 51
Date Time vertical scale horizontal scale PA of celestial north suncenter x suncenter y
1981 Dec 13 03:25:06 71.0 70.1 6.17 117 122
1981 Dec 13 06:41:25 70.9 70.1 6.17 117.5 121

The plate scale is close to 70.5 arcseconds per pixel.

Using only two stars

The closest (visible) star to the new comet is delta Sco (mag 2.3) on 1981 Nov 23. Beta1 Sco (mag 2.6) is also visible in these images though behind the pylon of the occulter. The pylon seems to distort the images. I have measured plate scales of 66.3/66.2/67.0/68.3 arcseconds per pixel on Nov 23 (the distance in pixel between the two stars is slowly decreasing as if the pylon magnifies that part of the image). On Nov 24, beta1 has cleared the pylon and I have used the image of 10:37:12 UTC:

Date Time scale PA of celestial north suncenter x suncenter y
1981 Nov 24 10:37:12 70.2 7.10 120 122

I have used this result to convert plate x,y to R.A. , decl in the images of 1981 Nov 20.

Measuring the comet

I have used images three times enlarged and measured to the nearest picxel.

Time comet x comet y hot spot x hot spot y dx dy new x new y R-A. 2000.0 decl 2000.0
01:24:58 566 677 495 356 0 0 566 677 15 37.9 -21 52
01:29:52 566 678 496 356 -1 0 565 678 15 37.9 -21 52
01:34:46 562 677 494 358 +1 -2 563 675 15 38.0 -21 51
01:39:39 557 675 491 359 +4 -3 561 672 15 38.0 -21 50
01:44:33 552 675 489 362 +6 -6 558 669 15 38.1 -21 49
01:49:26 548 674 488 362 +7 -6 555 668 15 38.2 -21 48
01:54:20 548 674 487 359 +8 -3 556 671 15 38.2 -21 49
01:59:14 549 668 491 358 +4 -2 553 666 15 38.3 -21 47
02:04:07 551 665 492 356 +3 0 554 665 15 38.3 -21 47
02:09:01 551 662 495 356 0 0 551 662 15 38.4 -21 46
06:21:49 461 581 496 357 -1 -1 460 580 15 41.3 -21 12
06:31:36 450 581 488 360 +7 -4 457 457 15 41.4 -21 11
06:36:30 446 578 488 362 +7 -6 453 453 15 41.5 -21 09
06:41:24 444 577 488 359 +7 -3 451 451 15.41.6 -21 09

The comet's x,y are measured from the upper left corner of the images (with the pylon on the left side of the images).
The measurements of the hot spot are used to compensate for the image shift.

Rainer Kracht, 2005 July 11