COTM: Serpens Caput

Albert Brakel

Serpens the snake is the only constellation that consists of two discrete parts: the head, called Serpens Caput, is separated from the tail, Serpens Cauda, by the constellation of Ophiuchus. Here we consider only the head region; the tail will follow next month. The concept of Serpens probably dates from before the dawn of history, with the snake always being portrayed as being grasped by the hands of Ophiuchus. In ancient times, the serpent and serpent-bearer were commonly considered as a single constellation.

Unlike the tail of Serpens, the head of the snake is far from the Milky Way, so apart from a single globular cluster, the objects of interest are double stars and faint galaxies. I observed from suburban Downer, with a 20-cm SCT.

Globular cluster

M5 (15h 18.6m, +02° 05’), a large and bright globular cluster, is the showpiece object of Serpens Caput, and is readily visible through a finderscope or binoculars. It has a well-condensed, brightly glowing core set in a dimmer halo of stars. Even 50x power resolved the stars in the halo. Tiny stars are resolved across the entire face of the cluster with 77x (which did not, however, resolve every star in the core). While not in quite the same class as the magnificent globulars NGC6397 in Ara, M22 or M13, it is nevertheless very impressive and worth revisiting from time to time. Just 20’ SE of M5 is the double star 5 Serpentis.

Double stars

5 Serpentis (15h 19.2m, +01° 46’): magnitudes 5.1 and 9.7, separation 11”, position angle 036° -- despite the large magnitude difference, the separation is great enough to make out the secondary without too much trouble.

6 Ser (15h 21.0m, +00° 43’): mags. 5.5 and 9.5, sep. 3.1”, PA 020° — quite a difficult one to resolve, because the brightness of the golden yellow primary overwhelms the much fainter companion. I only glimpsed the companion intermittently, and without complete confidence.

Delta Ser (15h 34.8m, +10° 32’): mags. 4.3 and 5.2, sep. 4.2”, PA 176° — far and away the best binary in Serpens Caput. Both stars are a lovely yellow, and aligned N-S, with the brighter primary at the N end. This is a good one for a small scope — the pair was easily split with 77x, and was a delight at higher powers.

Burnham 619 (15h 43.2m, +13° 40’): mags. 6.9 and 7.4, sep. 0.7”, PA 007° — a really tough customer, with very close yellow components. Although 0.7” is technically resolvable with 20 cm aperture at high powers (I have managed such a feat in the past), this requires perfect conditions, which did not prevail on this occasion.

Struve 1987 (15h 57.2m, +03° 24’): mags. 7 and 8½, sep. 10.4”, PA 322° — a pretty pair consisting of a white primary contrasting with the brownish secondary. The unequal components have a wide separation, and were readily resolvable even with 77x power.

Struve 1985 (15h 55.9m, -02° 10’): mags. 7 and 8, sep. 5.8”, PA 347° — a dainty, pale yellow, unequal double, fairly easy to split.

Galaxies

The galaxies in this part of Serpens are all faint ones, so I didn’t try for them from suburbia. If you feel you just have to look at one, go for NGC 5921 (15h 21.9m, +05° 04’), which at mag. 10.8 is the brightest galaxy in Serpens Caput. It is a barred spiral measuring 4.9’ x 4.2’ across, and is said to resemble a small, roundish blotch of light with a bright centre. If you intend to try it, good luck, and may your sky be dark.

HJ4404 - Neglected No Longer

Michael Curtotti

Not so long ago I took the plunge and bought a Meade LX90 8 inch telescope. The scope, purchased at Bintel in Sydney, came packaged with a Meade Lunar Planetary Imager and is all set up to take quite reasonable pictures of targets such as well - the Moon and planets.

The telescope, for those who are interested, is great in every respect, except the optics (don't laugh). Objectively, the optics only rate a "good". A slight astigmatism is noticeable just outside focus (this could however be a collimation issue). Something Meade could also have done better is not use a prism for the star diagonal. For bright targets (e.g.Sirius) this introduces visible concentric defocused ghost images. (Useful for centering an image but a problem if you are interested in photometry, as well as introducing a lot of noise for planetary images). However I would be conveying a misleading impression if I didn't say that I feel like it is a work of art every time I use the telescope. The mount, the aperture, the go to and accurate tracking make it a pleasure to use. Its optics are also far better than what I was used to (see below).

The last article I wrote for Southern Cross delved into my interest in binaries and adventures imaging with a 4 inch Newtonian and a webcam. Imaging any double other than Alpha Centauri presented insuperable practical problems because of a mount which at CCD scales vibrated at a scale bigger than the field of view and an imager which only worked to magnitude 5.

The Meade LPI imager (based on new CMOS technology) images, in real time, when used with an 8 inch SCT, comfortably down to 11th magnitude and at a stretch to 12th, (the imager is limited to a 15 second exposure). With a 3.3 focal reducer the imager images to 14th magnitude at a stretch. Armed with this and a suite of freeware software I was in business and returned to my natural interest — measuring double stars.

NGC3532 HJ4404 near centre

I found a convenient target in NGC3532 — a beautiful and crowded southern cluster sporting an attractive double trapezium marked by a red giant close to its centre. This field contains a double not shown in the Tycho2 catalogue. A little research established that this double is HJ4404, listed by the folks who run the Washington double star catalog as a neglected southern double last measured in 1927. But how to establish the field orientation and image scale for this double?

The drift method proved prone to significant error as a means of establishing image scale. The fundamental cause of this error arises from a combination of small CCD size and the fact that Microsoft Windows 'time' is only calculated to the nearest second. Put another way if we turn off tracking and take two images 10 seconds apart there could be up to a 10% error in scale due simply to the 'time' value windows allocates the picture.

Three different trials of the system provided arc second per pixel scale of 0.80, 0.83 and 0.85 arc seconds per pixel respectively with data from 30 separate images. A plot of the average arc second per pixel separation at different time points for one of these trials illustrates the timing problem.

Plan B was to work out scale from catalogued stellar positions. This turned out to be much better. Cartes du Ciel, a freeware star chart, provides Tycho2 stellar positions.

Taking 84 separate distance measurements of stars within two clusters (NGC 3532 and NGC 3293) and comparing with pixel separations from the images the image scale was established to be 0.8185 arc seconds per pixel with a 99% confidence interval of 0.0002 “/pixel i.e. 0.8185 +/- 0.0002 arc seconds per pixel.

Finally I was able to measure my double with a result of position angle 301.68 +/- 0.47 and separation of 8.43 +/- 0.06 (the additional error is from the actual measurements). (The drift method finally did come in useful for establishing position angle).

The above measures compare to the 1927 measurement of 298 degrees and 8.3 arc seconds, respectively for position angle and separation.

My fun with this double was not exhausted however. Observation done, but is this double a true binary? We know it appears in a cluster. But are the two stars members of the cluster? The WEBDA website is just what is needed to find out more about this pair. The WEBDA site collects information from various studies of clusters. It has information on position, star type and lots more, including cluster membership.

Can we tell anything about the period of these stars – if in fact they constitute a binary system?

Our pair is not a non-member of the star cluster. This helps us establish a probable known distance (486 parsecs). Given information about the apparent magnitude (apparent V magnitudes are 8.91 and 9.85 respectively (Webda data)) and the distance we can form an estimate of absolute magnitude (0.36 and 1.3 V) and from absolute magnitude we can form a reasonable guesstimate of mass (around 3 and 2.8 solar masses respectively ) – from the luminosity mass relationship of most stars – in this case taken from a graph in an introductory astronomy text. We can also state the apparent separation in astronomic units (arc” x parsec distance) or 4096 au.

In respect of binaries, Kepler’s law can be expressed in the form P²=a³/(m1+m2) where p is the orbital period, a is the semi-major axis in astronomic units and m1 and m2 are the masses of the two stars measured in solar masses.

While we cannot know the period from one observation can we make a statement of the form “if this double is a binary its orbital period is greater than x”?

While there are a few steps involved, it would appear possible. Firstly let’s assume that we are observing the stars at their greatest apparent separation (if this is not the case the period will be bigger). If the stars are binary with little eccentricity the apparent separation will no greater than the semi-major axis (let’s also assume that we are looking at the orbit face on for the moment).

If the binary is highly eccentric (i.e the primary will be close to one side of the secondary orbit) then we can only say that the apparent separation will be no greater than the major axis (i.e. twice the semi-major axis) or to put it another way half the separation will be no greater than the semi-major axis.

If the orbit happens to be inclined to us this does not affect this relationship as the apparent separation will then (except where the axis of inclination coincides with the position angle of the two stars) be smaller than the separation that would be observed if the orbit was face on. (Note that in the case of an inclined orbit the greatest apparent separation (which will be smaller than the greatest apparent separation if the orbit were face on) will appear at another point in the orbit than the true major axis or aphelion of the two stars (this is counter-intuitive but can easily be demonstrated by rotating a piece of paper with an ellipse and a semi-major axis drawn on it).

In any case we should be able to state:

In the case of HJ4404 we can say that the orbital period must be greater than around 38 500 years. If our double is a binary at all, it is not going to do anything soon! If the necessary information is available we should be able to apply the foregoing to exclude observation of doubles whose orbits, if they exist, are just a little too time consuming to bother with.

HJ4404 is one of thousands of neglected southern doubles within reach of amateur equipment.

Photometry Footnote:

More recently I have attempted photometry with the LPI imager. Given its budget character, its use for photometry has to be a dubious exercise. So far, testing I have done indicates some difficulty in calibrating against any standard system. The imager however does a reasonable job with differential photometry.

XZ Carina here provided the target. It is a Cepheid variable with a delta magnitude of about 1 magnitude and a period of about 16 days. Using a modified dark subtraction and automatic stacking routine utilizing the Meade provided software I have so far obtained the results indicated in the graph. Confidence intervals were obtained by taking ten or more measurements for each combined image, with each image representing the mean of 100 stacked images.

Creating Mosaic Images of Comets

Vello Tabur

After the last monthly meeting, Ross Gould asked me to write an article describing how I created the LMC / NEAT mosaic image. I wanted to capture the comet and the LMC in a wide field shot and it would have been easiest to use a short focal length camera lens, say of 80mm focal length, attached to my CCD camera. Unfortunately I have overhead power lines near my observatory which would have ruined the image, so I chose to build a mosaic of smaller (but higher resolution) images.

I normally use a Nikon 80-200 mm zoom lens set to 140 mm, which gives a FOV of 5.7 x 3.6 degrees. The camera uses a Non-Anti Blooming Gate (NABG) detector. The advantage is that it’s more sensitive than an ABG chip so it picks up more faint nebulosity. The down-side is that bright objects will saturate fairly quickly and leave a blooming spike, the length of which depends on the degree of saturation. At my short focal length, it starts to become a problem on exposures longer than 30 sec, or when very bright objects are in the field. Luckily there were no really bright stars in the target area so I took ten 30 sec exposures of each field. It required a little planning to determine the correct field centres, but no great precision was necessary to slew to each field, as I allowed for some overlap.

The main priority was to image the segments that contained the comet so that they could be registered without the comet’s movement becoming apparent. The other constraint was that there is a fair bit of light pollution in my SW sky, so each image had to be taken at the greatest possible altitude but still be below the lowest power line (which is at an altitude of about 40 degrees). It took about 1.5 hours to capture all the raw images.

In addition to the light frames, I took 25 dark frames. Darks are normal images taken with the same exposure time and temperature as the lights, but with the shutter closed. Heat in the electronics of the detector generates photo-electrons in the same way that photons coming from the sky do. By taking darks, we can remove this source of noise from the images. The darks were combined into a master dark frame with my own software using a sigma-reject algorithm to remove chance cosmic ray hits.

Each raw image was dark-subtracted and flat-fielded. The latter removes the effects of vignetting, dust on the optics, and inter-pixel sensitivity differences. Each set of calibrated images were then stacked (co-added to improve signal the to noise ratio) using some commercial software (AIP4WIN). This yielded twelve master frames, each one being a calibrated 300 sec exposure. The stacking process is performed by selecting two reference points on each frame. These are usually well sampled stars that are diagonally opposite each other in the corners of the image. The software rotates and scales the slave images to the orientation of the master and adds them together.

One potential problem is that periodic error in the drive or an incorrect drive rate may produce a slight shift on some images so that not all of them contribute equally to the pixels on the outer edge of the image. For this reason, I usually crop the resulting image slightly to remove a couple of rows/columns around the outer edge.

Since the images were taken in a light polluted sky, each one contained an obvious gradient (brighter towards the horizon). Trying to combine them into a mosaic at that stage would have (and did) produce an unsightly mess, so the next step was to remove the gradient. Luckily I already had some software that could do this, although it had to be enhanced a bit. The basic technique is to place an imaginary grid (say 64 pixels square) over an image and to calculate the background intensity in each cell. This is done using pixel statistics.

Knowing the background for each grid, it is an easy matter to calculate the background intensity for each pixel using a bi-linear interpolation and subtracting an appropriate amount from each pixel. This results in a gradient corrected image; well almost. The gotcha here is that if a cell contains a lot of nebulosity, like a comet tail or within the LMC, the background value is overestimated and subtraction removes the nebulosity too! I changed my program to superimpose the grid over an image and used mouse clicks within selected cells to disable the background calculation for them. This way I could exclude cells that contained the comet’s coma and tail from the background calculation. All other cells were computed as before, and the cells that were excluded were interpolated using nearby cells.

This produced a nice result for most images, with the exception of the images that were wholly within the LMC, as they contained nebulosity in all the cells. I was stumped for a while but finally came to the realization that other images containing non-nebulous regions taken at a similar altitude and azimuth to the LMC shots recorded a similar sky glow. Applying the background grid from the most appropriate images to the LMC sections proved quite successful, although a little manual correction (in AIP4WIN) was required.

The next step was to arrange the images into a mosaic. Up to that point in time the largest mosaic I had produced was one of Comet LINEAR containing three panels. A lot of people use Photoshop for building mosaics as it allows one to manipulate the images independently in separate “layers” until they are properly registered and then merge them into a single image. However, Photoshop costs about $1000, so I downloaded GIMP (GNU Image Manipulation Program – see www.gimp.org) for nix. It performs almost all of the functions of Photoshop. As Comet LINEAR was imaged near the celestial equator, a rectangular projection suited the images and they were easy to align with up/down and left/right shifts. However, the NEAT images near the LMC were taken at high declinations where a conic projection was more appropriate. This meant that after aligning just two images manually, I found that all sorts of rotation and scaling operations were required to get the rest to overlap properly.

Not keen on trial and error, I chose to write a bit of software to fix the problem. First, an astrometric plate solution was calculated for each image. This allows one to translate between RA/Dec and x/y on each image. Then I created a 4500 x 3000 pixel canvas and using the astrometric solution of each image, determined where each pixel should be placed on the canvas using the solution (projection) of the central image. This does introduce a little distortion, but it’s not noticeable in practice, and I was after a pretty picture and not a science grade image after all.

The final (heartbreaking) step was to resample the canvas to a more manageable size. The original images showed pinpoint stars down to 16th magnitude with a wealth of detail. But for web publishing and display the canvas was resampled to 1280x1024 and compressed into a jpeg with a further 20% loss of detail. Even then, it still looks pretty good, but nothing like the original. Still, not bad for an ordinary telephoto lens used from a suburban backyard with a 36% sunlit Moon in the sky!

CAS Coastal Astro Star Party (CASCAS) at Kioloa, 12-14 July 2004

Nicole Kennedy

Kioloa. Pronounced the Anglo way of Ky-Ole-Ah. Those who were feeling tropical and wanted to pretend they were in Hawaii pronounced it Ky-A-Low-Ah.

The story begins with someone having a great idea of a CAS Starparty. Someone else – not to mention any names or point any fingers but her initials are Kim Rawlings – had contacts. And before you knew it, the trek to just south of Bawly Point on the southeast coast was set for the June long weekend.

About thirty of us – Canberra based members and four from Merimbula – ventured to the ANU's research facilities at Kioloa. I was expecting to have to rough it, i.e. not a power point or flushing toilet in sight, and was pleasantly surprised by the whole site.

The bunkhouses were roomy and very comfortable, the showers with their separate flushing toilets very clean. The mess hall was huge, with an industrial-sized kitchen, there were even table cloths!

Telescopes would be set up about 100 m from the sleeping accommodation, in the carpark next to the classroom. Viewing would be spectacular – if the blasted cloud cover would naff off!

And we had a very special guest, author and video astrophotographer Steve Massey. Steve endeared himself to all present, not by his informative, interesting and easy to understand talk on video astronomy, but because of his very firm statement that 'CAS is a lot more fun than the Sydney people – they take themselves very seriously'.

Anyway.

The rocket launch at approximately 16:30 hrs EST 13 July 2004 was a fabulous success. About 40 cm long, powered by not much more than a fireworks detonator, this thing took off like SpaceShipOne (another privatised space flight). The rocket landing was not quite as successful – the parachute and rocket separated over a power line and at time of printing may not long have been recovered by Country Energy and a long ladder.

Viewing prospects for the evening were grim; for all of Saturday, there was a fairly dense cloud cover that didn't look like budging. By 7 pm however, after we'd all eaten and talked our heads off in the communal mess, the clouds had rolled away and we had a magnificent panorama of cosmic beauty to get lost in.

This writer retired early – about 11 pm – but the Party went on. Our bunkmate dragged her butt in at 2 am, but led us to understand others continued on to even wee-er hours.

Sunday dawned bright and beautiful and gloriously warm. It was divinely idyllic. Those not affected by hangovers, tiredness or the megrims went for an adventure walk and fossil hunt to The Lighthouse [at Ulladulla], others strolled around the beaches or gazed at cows, I was interested in tea, biscuits and mahjong. A quiet day was had by all as we geared up for another great night.

Steve Massey treated us to an excellent talk on video astronomy at 5 pm. It's great to know that in our sphere of interest, with some very high-brow talkers and concepts waaaayyyy beyond this writers' ken, there are such personable people like Steve out there. He makes his topic of interest easy to understand and well within reach of everyone.

Again, we were troubled by persisting cloud, but decided to have dinner and see what the go was a little later. Sure enough, by 7:30 pm, the skies were about 90% clear. Thin cover scudded across every now and then in the high south, but we only had to wait a few minutes and bang, LMC back in view.

It was a night of meteors, and I kid you not when I write that at least every five minutes we saw streaks of light.

What more can I say? It was a top idea, great accommodation at our disposal, fantastic facilities, a really cool guest, and the funniest, most laid back Astronomical Society in ACT/NSW to share it all with. Our Merimbula members had a ball – how nice it was to meet them!

If you weren't able to make this year's CAS Starparty, fear not. We plan to have another next year, and while it may be bigger, possibly better, the June 2004 Kioloa version was excellent.

See you in the LMC!