Astronomy Cast

Astronomy Cast takes a fact based journey through the cosmos as it offers listeners weekly discussions on astronomical topics ranging from planets to cosmology.

Sixtyseven years inbetween ...

Somehow I recently stumbled upon pictures of the Digitized Sky Survey and thought to myself: hmm, they look very similar to the pictures I take with the 14 inch telescope in my astrometric follow-up work. It should be really interesting to compare the two directly. The DSS-1 was created 1994 by digitising images of the POSS-1 - the Palomar Observatory Sky Survey – a photographic survey of the sky with the large Schmidt-Telescope of Mount Palomar Observatory. Those photographic plates were taken in the years 1953 to 1956. Today the data can be retrieved either from ESO at or from STSCI, the Space Telescope Science Institute at

The 'Big Schmidt' is a large telescope with an aperture of 122 cm (48 inch) and a focal length of 300 cm, built in 1939-48. After several upgrades and being equipped with CCD cameras it is also used for asteroid tracking. First for NEAT (Near-Earth Asteroid Tracking) from 1995 - 2007 and now for the ZTF program, the Zwicky Transient Facility, that is contributing to NEO discovery since 2018.

I wanted to compare my images with DSS-1, because the resolution of 1.7 arcseconds per pixel of those scans is almost identical to the 1.75"/pixel of my equipment. I picked one of my images from August 8, 2020. The limiting magnitude almost reached 21.0 mag (measured with the software 'Astrometrica' with reference stars from GAIA DR2 G-Band) after an exposure time of 60 minutes, despite an 80% illuminated moon but good seeing and negligible clouds. The CCD camera in use was my old SBIG ST8-XME, having its peak sensitivity in the red spectrum at 650 nm. In comparison POSS-1 red was exposed for 45 minutes on Kodak 103aE photographic plates reaching a limiting magnitude between 21.0 and 22.0 mag depending on the location in the sky. According to the FITS header of the DSS image, retrieved from ESO, it was taken on December 31, 1953 at 03:32 Universal Time. New Years Eve - what a surprise.

The center coordinates of the selected field are at RA 02:11:21.521, DEC +43:57:00.192 (J2000.0). My image of 2020 consists of 120 individual exposures of 30 seconds each and is calibrated with dark- and flat-frames with no further image processing. The minor planet 2012 KE25 (a Near Earth Object) is also part of the show, moving through the field at a speed of 6.6"/min and a magnitude of 19.2. However, its trace cannot be seen here. To make it visible, it would be nessecary to combine the 120 individual images using the "track-and-stack" method. Both images - mine and that from the DSS - are aligned to each other only by rotating, shifting and a bit of resizing. The result is amazing. How similar the two actually look! With only a little more exposure time the small 14 inch telescope can compete with the 'Big Schmidt' of Mount Palomar! By using a very sensitiv electronic camera instead of outdated and antiquated photographic plates of course. What development technology has taken. And there are also 67 years between the two recordings! And for that fact alone, it was worth taking a closer look at the picture.

Click on the image for a larger version

Almost instantly I realized: some of those dots are moving! As a minor planet observer I was immediately going: ohh, minor planets! Which is complete nonsense when you think just one second further. With such little movement after 67 years those dots must be very far away. Way beyond our solar system. That must be stars. But when stars are moving at the night sky, two reasons come to mind: parallax and proper motion. But the parallax of the closest star, Proxima Centauri, is only 0.772 arc seconds. The difference in position here is certainly multiple arc seconds large. In the online catalog 'Aladin' at one can find images of a large number of sky surveys in different wavelengths superimposed with catalog data containing all information needed.

I have marked three stars with A, B and C and searched for their data in the GAIA EDR3 catalog. Here is an excerpt from the GAIA data:

Star A
ra 32.87122666596
dec 43.84692308850
source_id 351679863593242624
parallax 8.5896 mas
proper motion 129.407 mas.yr**-1
magnitude G 17.421354

Star B
ra 32.88377621316
dec 43.85214332317
source_id 351773700038991616
parallax 6.5438 mas
proper motion 57.481 mas.yr**-1
magnitude G 9.304178

Star C
ra 33.02367303479
dec 43.95086652237
source_id 351775722966261248
parallax 6.2619 mas
proper motion 93.416 mas.yr**-1
magnitude G 17.900475

explanation: mas = milli-arcseconds

Now, if you multiply the proper motion for each star with 67,608 you get the following results:

Star 'A': 8.748 arcseconds, Star 'B': 3.886 arcseconds, Star 'C': 6.315 arcseconds motion over 67 years and 222 days. For comparison: the fastest star in the sky is Barnard's Star with 10.25 arc seconds per year! Those three stars where just randomly picked.

Now I measured star A and C in both images with the software package 'Astrometrica'. Unfortunately, star B is saturated and could not be measured.

                                                         FWHM Peak Fit
                 Day      RA           DEC                 "  SNR  RMS
A-A17   C2020 08 09.04758 02 11 29.173 +43 50 48.36      4.4 35.9 0.016
A-DSS1  C1953 12 31.16207 02 11 28.360 +43 50 49.49      3.9 29.3 0.033
C-A17   C2020 08 09.04758 02 12 05.765 +43 57 03.07      4.3 32.7 0.024
C-DSS1  C1953 12 31.16207 02 12 05.154 +43 57 01.35      3.6 32.5 0.039

comparison stars with proper motion < 10 mas/year
V1-A17  C2020 08 09.04758 02 11 31.49 +43 57 04.6
V1-DSS  C1953 12 31.16207 02 11 31.53 +43 57 04.6
V2-A17  C2020 08 09.04758 02 11 52.93 +44 00 44.0
V2-DSS  C1953 12 31.16207 02 11 52.98 +44 00 44.0
V3-A17  C2020 08 09.04758 02 10 33.70 +43 58 57.4
V3-DSS  C1953 12 31.16207 02 10 33.72 +43 58 57.5
V4-A17  C2020 08 09.04758 02 11 04.94 +44 00 20.0
V4-DSS  C1953 12 31.16207 02 11 04.96 +44 00 20.0

As reference catalog Gaia DR2 was used.

The astrometric error of DSS image solution: 172 of 234 reference stars used: dRA = 0.09", dDe = 0.10"
The astrometric error of A17 image solution: 265 of 342 reference stars used: dRA = 0.12", dDe = 0.12"

The angular separation is determined by the following relation:
theta = arccos(sin(dec1)*sin(dec2)+cos(dec1)*cos(dec2)*cos(ra1-ra2)).

astropy with the SkyCoord-method was used to calculate the angular separation between the two coordinates.

My result:

A-A17 / A-DSS1 = 8.867 ±0.10 arcseconds.
C-A17 / C-DSS1 = 6.818 ±0.10 arcseconds.

Not too bad, compared to the values based on the Gaia proper motion measurements.