A search for faint asteroids

 tl;dr - A teacher and two teenagers drive up to an observatory with a 16-inch telescope at 1,600 m in South Tyrol on a clear December night. The plan is to discover faint asteroids with the so-called “blink” method. Despite several obstacles, as a snowstorm of hotpixels and a merciless ski-slope light dome, they come away with ~ 3x15 faint asteroids (and one interesting variable star). Here’s the full story!

On 28 December 2024 the sky was clad in a perfect winter blue, and two of my students — Emma (15) and Miriam (17) — despite being in the midst of winter holidays, agreed to an observing night at our school's rooftop roll-off observatory.

Ready to hunt down asteroids. Our 16-inch RC telescope.

The evening started promisingly: While we set up our equipment the air smelled cold and clear. Then however, the village below started their night-skiing and a gruesome light dome devoured all faint stars in the southeast. We had no other option but to postpone plan A (our asteroid hunt) in favour of Plan B: taking H-alpha pictures of the hydrogen clouds in the "hidden" galaxy IC 342 — certainly less scientific, but nonetheless pretty. The delay wasn't that bad in the end: we had gained plenty of time to plan our asteroid assault.

The light dome toward Southeast was a big nuisance for our task. 

The inspiration for this latest endeavour of our students asteroid group came straight out of Mike Brown’s book "How I Killed Pluto And Why He Had It Coming". The detection method he describes is quite simple: take several exposures of a sky region and “blink” them in sequence — stars stay put, asteroids jump. 

We used the free software Stellarium to pick a promising patch of sky: the border between Taurus and Gemini, high enough to reduce air mass and roughly opposite the Sun to exploit the opposition (Seeliger) effect. It states that asteroids with dusty surfaces brighten when you look at them with the Sun at your back. Our field of choice also happened to be right in the Milky Way, so we expected to see plenty of stars.

A key choice was exposure duration. To find the optimum we worked out the asteroid's orbital speeds with the Titius-Bode law, Kepler’s third law and a little bit of trigonometry. With our typical seeing value of ~3 arcseconds in mind, we settled for 5 minutes. That would keep the moving asteroids close enough to point-like sources.

By 22:00 the ski-slope went dark at last, and the value of our sky quality meter dropped to 20.8 mag/arcsec² — not our best, but fair. We removed the H-alpha filter we used for IC342, left the UV/IR blocker out (to gather as much light as possible), used the autofocus routine in N.I.N.A. to get our stars sharp, and were ready to feed starlight to our CMOS sensor. Miriam picked a half a square degree wide field centered on the star HD48641 (V≈7), 36′ from the ecliptic. Emma and I chose two neighbouring fields.

Miriam's first frame showed a beautiful, overwhelming star soup — just as expected. One single image is worthless for the search for asteroids, though. We needed at least two for motion perception. So we patiently waited for the end of the second exposure of Miriam's field and while the exposure run went on automatically, we registered the two images in PixInsight.

Blinking was chaos at first: the uncalibrated raw frames were full of hotpixels. Looking at the laptop screen was like gazing out from a window into a furious blizzard. Those faulty, bright pixels hopped around in our blink loop and got us totally dizzy. But suddenly we saw it: a slightly elongated smudge jumping a little between frames. It clearly was too large for a hotpixel. Neither was it a cosmic. An asteroid! Emma grinned; Miriam, who was engaged in a call with her worried mom, explaining she was merely out, discovering moving objects on a laptop at midnight, grinned too. 

On that night we collected nine science frames, then hurriedly gathered a few bias frames but skipped the darks and flats in the rush. That choice would haunt us later.

Miriam's star field

Two weeks later, we met in the school computer room for a careful data analysis. Because of the missing darks and flats we had do deal with vignetting (darkened image corners), and the already-familiar hotpixel storm. However, once the images were cleaned and prepared, we projected Miriam’s field onto the largest screen in the school — a 98-inches-diagonal monster — and divided the 4800×3211 pixels large image into nine search tiles. Heads down we looked up (🤪) at the blinking star field, marking suspected movers with yellow Post-its.

Magdalena and Robert marking asteroids with Post-its.

Identification of the unknown objects was a straightforward procedure: we used the Minor Planet Checker (a free tool by the Minor Planet Center)  and entered observation time (as fractional day), our measured RA/Dec, a 2′ search radius and a 21 mag limit. In nearly every case the top entry in the resulting output list matched our object. For some objects we cross-checked NASA JPL’s Small-Body Database Lookup for diameters and discovery histories — learning that 1997 NN, for example, is a ~2.7 km object with a very slow ~53 h rotation, discovered at Kitt Peak observatory.

We even found something unexpected: in Emma’s field one star seemed to change slightly in apparent diameter between frames. WikiSky showed a 14-mag, reddish object there. VSX (AAVSO’s variable star index) identified it as V0449 Gem, an eclipsing contact binary of the W UMa type. I knew this type of variable very well from a previous project — what a great opportunity for a spontaneous lesson about stellar intimacy. 💕

#	RA (J2000)	Dec (J2000)	V mag	asteroid designation
1	06 46 07	+23 50 13	19.2	2004 FF30
2	06 45 55	+23 49 41	19.7	2007 RY83
3	06 45 48	+23 49 49	18.8	2000 DY93
4	06 45 53	+23 56 27	18.9	2008 CU209
5	06 46 23	+24 00 18	18.3	1997 NN
6	06 45 48	+23 55 06	19.2	2003 SC165
7	06 45 35	+23 43 52	19.7	2008 RJ66
8	06 45 23	+23 26 40	19.9	2008 VA80
9	06 45 25	+23 23 22	20.4	2004 BQ66
10	06 45 11	+23 29 09	17.9	Velasco
11	06 44 49	+23 37 12	20.1	2008 EF31
12	06 44 57	+23 47 10	18.2	1998 VY3
13	06 44 37	+23 42 23	20.5	2003 WY210
14	06 46 22	+23 20 52	20.7	2002 AN205
15	06 46 41	+23 45 51	20.9	2006 QQ162

Table 1: Asteroids detected in our blink search in Miriam's field. The faintest objects reached about V ≈ 21, impressively dim — roughly a million times fainter than the faintest naked-eye stars.

After the great success I decided to write about our little adventure in the German journal "Sterne und Weltraum". The article was published in the September 2025 issue. 

Feeling curious? The three blinking images from Miriam’s field are online — see if you can catch an asteroid or two!

The article in "Sterne und Weltraum 09/25".

Christof Wiedemair