Photometric follow up, 140 candidates, 182 nights

Astronomy research poster example

Six months of photometry on 140 candidate variable stars, laid out as an A0 portrait showcase board with two figures set into a three column grid. The board answers the question a small observatory asks each season: what fraction of somebody else catalogue can a 0.4 metre telescope actually resolve.

Create a poster with OneCraftA0 Portrait, printed at 841 by 1189 millimetres

The whole board

The poster at full size, exactly as it prints. Every number, citation and caption on it was written for this example, so the layout is being judged on real content.

One hundred and eleven confirmed: six months of photometry on 140 candidate variable stars

T. Halvorsen ¹, R. Mbeki ¹, S. Ivanova ¹ ², P. Corrigan ²

1 Department of Astronomy, Wrenfield University · 2 Wrenfield Observatory, Marrick Ridge

Abstract

A survey catalogue flagged 140 stars in one field as candidate variables on the basis of two epochs. We followed all 140 for six months on a 0.4 m telescope, collecting 11,300 usable frames over 182 nights, and classified each star from its own light curve rather than from a catalogue flag. One hundred and eleven are variable: 51 pulsating, 38 eclipsing and 22 irregular. The remaining 29 are constant to the precision of this data, which is 8 millimagnitudes at the survey limit. A small telescope with a long baseline therefore recovers a class for four candidates in five. Twelve of the 29 constant stars have a brighter neighbour within 30 arcseconds and six were flagged on epochs under an hour apart, which leaves eleven flagged stars this data cannot explain.

Why follow a catalogue flag

Two epoch surveys flag variability by difference, which cannot distinguish a genuine variable from a bad pixel, a cosmic ray or a night of poor seeing. The flags are useful precisely because they are cheap, and they are worth following up only if the follow up is cheap too. A 0.4 m instrument with a 182 night baseline is about as cheap as follow up gets, and this study asks what fraction of a flagged list it can resolve. The answer bears on the practical question a small observatory faces each season: whether to spend the time on a catalogue that somebody else produced. The field was chosen for its moderate galactic latitude and for having 18 usable comparison stars inside the same frame, which is what sets the photometric floor on an instrument this size. Nothing in the target list was selected on brightness, so the 140 span four magnitudes.

METHODS

Targets. All 140 candidate variables flagged in one 1.2 square degree field by a two epoch survey, with no further selection on brightness, so the list spans four magnitudes.

Instrument. A 0.4 m reflector with a cooled CMOS camera, one broad filter and 60 second exposures, run unattended on a scripted sequence each clear night.

Coverage. One hundred and eighty two nights across six months. Of 14,600 exposures taken, 11,300 survived a cut for cloud, seeing over 4 arcseconds and tracking faults.

Photometry. Differential photometry against 18 comparison stars inside the same frame, giving a precision of 8 millimagnitudes at the survey limit.

Classification. A period search followed by visual inspection of the folded light curve by two of the authors independently, with disagreements resolved by a third.

Figure 1. Confirmed variables by class: Ecl eclipsing 38, Pul pulsating 51, Irr irregular 22.

Source: 11,300 frames, 182 nights, 0.4 m telescope.

Results

Of the 140 candidates, 111 show variability above the noise floor across the six month baseline. Pulsating stars are the largest class at 51, with periods between 0.18 and 11.4 days; 34 of the 51 sit in a narrow band of period and amplitude that is consistent with a single type. Eclipsing systems number 38, of which 31 gave a period good to better than a part in a thousand and 7 were caught with only one eclipse in the whole run, so their periods are lower limits. The 22 irregular stars show variability with no period recoverable from this cadence, which for most of them means either a timescale longer than six months or an amplitude that wanders. The 29 constant stars are the interesting residue. Twelve of them sit within 30 arcseconds of a brighter neighbour, which is the classic two epoch false positive, and a further six were flagged on epochs separated by less than an hour, where the survey difference is most sensitive to seeing. That leaves eleven flagged stars for which this data can offer no explanation at all, and which are the ones worth putting on a larger instrument. Period recovery was limited by the observing pattern rather than by the photometry. With one site and a six month window, periods near one day and near half a day alias against the night, and four of the 38 eclipsing systems have a second period solution that fits the data almost as well. Frame yield was 11,300 of 14,600 exposures taken, with the rest cut for cloud, seeing over 4 arcseconds or a tracking fault, and the cut was applied before any photometry was run so it cannot have selected for variability.

Figure 2. Brightest eclipsing system folded on a 3.21 day period. Axis is phase, series is magnitude (m).

Source: 418 frames folded on the fitted period.

111 of 140

candidates confirmed variable

79%, with 29 constant to 8 millimagnitudes over 182 nights

What a small telescope can and cannot settle

The headline number, four candidates in five resolved, is the one an observatory time allocation committee needs, and it is high enough to justify the season. The classification split is less transferable: this field sits at a moderate galactic latitude and a field closer to the plane would return a different mix with more irregulars and more blends. What the data cannot do is measure anything at the faint end. Precision falls to 8 millimagnitudes at the survey limit, and a pulsating star with an amplitude of 5 millimagnitudes is invisible here regardless of how many frames are stacked, so the 29 constant stars should be read as constant at this precision rather than constant. Two practical lessons came out of the run. The first is that the comparison ensemble matters more than the aperture: moving from 6 to 18 comparison stars cut the scatter by a third at no cost in telescope time. The second is that nights are worth more than frames. Adding frames to a night improves a single point; adding nights across a six month window is what breaks the aliasing, and the four ambiguous eclipsing periods would all have been settled by ten more nights in the last month rather than by a thousand more frames. The season that follows is planned on that basis: fewer frames a night, more nights, and the eleven unexplained constants moved to a larger instrument rather than kept in the queue here.

Conclusion

One hundred and eleven of 140 flagged candidates were confirmed variable and given a class; 29 were constant to 8 millimagnitudes, and eleven of those have no ready explanation. Four candidates in five were resolved on a 0.4 m telescope in one season.

111 of 140 candidates confirmed, 79% of the list

51 pulsating, 38 eclipsing, 22 irregular

11,300 frames over 182 nights on a 0.4 m telescope

18 comparison stars cut the scatter by a third

REFERENCES

1.Mbeki, R. and Halvorsen, T. (2025) Follow up yield for two epoch variability flags. Small Telescope Astronomy, 26(2), 88 to 109.

2.Ivanova, S. (2024) Comparison ensembles and differential precision at 0.4 metres. Observational Notes, 17(1), 12 to 27.

3.Corrigan, P. (2026) Aliasing in single site six month baselines. Wrenfield Observatory Reports 2026-04.

Block by block

What each block on the board is for, in the order a reader walks it.

Title band, four authors and a QR code: The title leads with the result: one hundred and eleven confirmed, six months of photometry on 140 candidate variable stars. Four authors carry generated superscripts over a department and an observatory, with the QR code at the right.
Abstract: A catalogue flagged 140 stars on two epochs. All 140 were followed for six months on a 0.4 metre telescope, 11,300 usable frames over 182 nights, and classified from their own light curves. One hundred and eleven are variable and 29 are constant to 8 millimagnitudes.
Why follow a catalogue flag: Two epoch surveys flag variability by difference, which cannot separate a real variable from a bad pixel or a night of poor seeing. The flags are cheap, so the follow up has to be cheap too, and this study asks what fraction a small instrument can resolve.
Methods as a labelled protocol: Five entries: 140 candidates in one 1.2 square degree field, a 0.4 metre reflector with a cooled camera and 60 second exposures, 182 nights and 11,300 usable frames, differential photometry against 18 comparison stars at 8 millimagnitude precision, and period search with visual inspection.
Class figure and folded curve: A bar chart gives the three classes as 38 eclipsing, 51 pulsating and 22 irregular, keyed in the caption. A second figure folds the brightest eclipsing system on a 3.21 day period, with phase on the axis and magnitude as the series.
The headline number and results: A key number block reads 111 of 140, at 79%, with 29 constant to 8 millimagnitudes over 182 nights. The results block gives the period ranges, the seven systems caught with a single eclipse, the twelve blends, the six short interval flags and the eleven unexplained.
What a small telescope can settle, and references: The discussion separates the transferable number, four candidates in five, from the classification split, which depends on galactic latitude. Two practical lessons follow on comparison ensembles and nights against frames, with four checked points and three references.

What makes this board work

The residue is explained, not ignored

Twenty nine constant stars are broken down: twelve with a brighter neighbour within 30 arcseconds, six flagged on epochs under an hour apart, and eleven with no explanation at all. Those eleven are the sentence a reviewer remembers.

Two figures do two jobs

One bar chart gives the class split across 111 confirmed variables; one folded light curve shows what a single classification looks like. Repeating the class split as a pie as well would have added colour and no information.

The practical lesson is quantified

Moving from six to eighteen comparison stars cut the scatter by a third at no cost in telescope time, and ten more nights would have settled four ambiguous periods. Both are numbers another small observatory can act on.

Questions people ask

What goes on an astronomy research poster?

The target list and where it came from, the instrument and filter, the cadence and the baseline, the frame count after cuts, the photometric precision, and the classification rule. Precision belongs on the board because it sets what the null results mean.

How do I show a light curve at poster size?

Fold it on the fitted period and plot a handful of phase points rather than every frame. This board folds 418 frames onto four phase points and gives the period in the caption, which is enough for a reader to see the eclipse depth from a metre away.

How do I keep chart axis labels readable?

Keep them to two or three characters and put the key in the caption, where the type is larger. This board labels classes as Ecl, Pul and Irr and spells them out in the caption, which also keeps the legend clear of the axis row.

What text size does an A0 astronomy poster need?

Body text on this board is near 24 pixels on a 1656 by 2340 canvas, about 12 millimetres printed, readable at a metre. Avoid dark surface themes for a board with charts, because the chart tick labels render in the page text colour rather than the theme axis colour.

Can I print this at 300 DPI?

The export runs at twice the working canvas, so A0 portrait leaves at 3312 by 4680 pixels, roughly 100 DPI at full size. That is the working resolution for a poster viewed from a metre, and no bleed, crop marks or CMYK conversion are produced.

How do I adapt this to my own observing run?

Retype the title, abstract and the five method entries, then edit the two figures directly. Each holds its own categories and series, so a different class breakdown or a different folded curve is a data change. Both can be switched between ten chart types.

Build your own in about a minute

The button below opens the generator with this use case already described. Change the wording to match your own, generate, then edit anything you like.

Make my astronomy research poster example

Sources

Written and checked by the OneCraft team. Last checked 12 September 2026.