november stars

From: Joe Patterson <*jop_at_astro.columbia.edu*> Date: Tue, 21 Nov 2006 23:00:27 -0500 (EST) · jop_at_astro.columbia.edu

Dear CBAers,

I attach a short semi-popular paper I just finished on the new variable in 
Cassiopeia (Var Cas 06, as it is currently known).  It was a big CBA 
effort, with contributions from Arto Oksanen, Pierre de Ponthiere, David 
Boyd, Dave Messier, Donn Starkey, and Carole Haswell (along with several 
students from the Open University)... but the principal data (certainly in 
the CBA, and I think in the world also) were from Tom Krajci and Bob Koff.
Their precise data on the very first night after the announcement was the 
critical element in certifying the microlens fit.

So... you can read about it.  I'm pretty sure it's the first such event 
in history: a high-magnification event for a nearby star.  I'll put up the 
figures on the website in a couple days.

Time for new favorite stars!  I've been waiting for November, to promote 
two fascinating stars now transiting near local midnight.  RX0354-16 is a 
strange novalike variable with an enormous proper motion... a very 
tempting target, as it is likely to teach us something new about the 
local population of CVs.  Porb is still unknown, and there is some 
indication it might be as low as 40 minutes.  The magnitude is listed as 
16.0-18.4 ("Eri" in the Downes et al. catalog), although actual dwarf-nova 
eruptions are probably not present.  A great southern - and marginally 
equatorial - object if you can handle the faintness!

Likewise for SDSS0407-06, also "Eri" in Downes et al.  Alon Retter and 
Alex Liu wrote a nice paper on this star last year.  Our coverage verified 
the many periods they found, and I expect this star to be so rich in 
periodic content that it will be a prime target for two months or more.
Stated as 15-17, and probably a dwarf nova.  Good target regardless of 
brightness.

There's a new eruptive object in Leo, too,and Tom's photometry suggests an 
80 minute period.  I'm going to wait another day or two before signing on 
to this one, though.

Northern-only targets.  RX0636+35 ("Aur") is a shiny new DQ Her 
star, about 15.8 and a likely collection of strict periods.  Very nice 
target, and new to us.  Also an old friend: V405 Aur, another DQ Her star
and a great target for lousy conditions.

Most of the other targets from last month can be retired.  Exceptions are 
certainly AO Psc and FO Aqr: 1-3 hour observations of these stars in the 
evening sky continue to be quite useful.

Happy observing!  Please let me, and all of us, know what you find in 
these various new targets...

joe

          THE HALLOWEEN TRANSIENT OF 2006: A NEARBY MICROLENS?

    On 31 October 2006, the IAU's Bureau of Astronomical Telegrams sent

out a strange announcement: an 11th magnitude star in Cassiopeia had

suddenly jumped to 7th magnitude.  I remember feeling a bit skeptical;

I knew Dan Green was not generally fond of pranks... but it *was*

Halloween... and his office *was* known as the BATroom... and I

remembered a past announcement of Santa's reindeer.  Still, the e-mailed

telegram was not quite strange enough for a good prank; and the discovery

was credited to Akihiko Tago, one of the world's most famous discoverers

of comets and novae.  Then a second IAU email a few hours later provided

further details, and confirmation: GSC 3656-1328 had really erupted.

    All over the world, programs of visual and photographic observation

began, with results shard through internet news-groups.  Two members of

the Center for Backyard Astrophysics network (Bob Koff and Tom Krajci)

immediately started time-series photometry of the transient.  Their light

curves showed a blue star at V=8.9 and fading rapidly at 1.5 magnitudes

per day.  Initially, the greatest oddity was the apparent association

with a seemingly normal A star; this did not seem to resemble any known

class of variable star.  The oddities quickly grew.  On 3 November, we

had a busy day.

1. We obtained a target-of-opportunity X-ray observation with SWIFT,

which showed no detectable flux.

2. Our colleague Ron Remillard (MIT) searched the RXTE All-Sky Monitor

data for any sign of an X-ray transient; there was no detectable signal

on the days of outburst, nor on any other timescale over the mission's

10-year baseline.

3. We studied available data on the color of the outburst light.  To

within 0.05 mag, the outburst had the same color as the A star in

quiescence (B-V=0.20).

4. We received a fascinating new telegram (CBET 718) which deepened

the mystery.  The spectra reported by Ulisse Munari (Padova Observatory)

indicated a fairly normal A-star spectrum as early as November 1.1, with

no emission components and no evidence of rapid rotation.  And a search

by Sergei Antipin (Sternberg Astronomical Institute) of 400 photographic

plates during 1964-94 revealed no variation from a mean mpg=11.8, the

same brightness seen in modern surveys (Tycho, USNO, TASS, etc.)

    By the end of November 3, all the easy-to-imagine theories for the

outburst (prank, X-ray transient, cataclysmic variable, rapidly rotating

shell star) appeared to be ruled out.  A constant star had jumped 4 

magnitudes in a week, and now was returning to normalcy in another week.

No X-rays, no flickering, no evident change in color or spectrum.

Amazing.  Stellar zoology didn't seem to have a place for this event.

    Actually, there is a place; it just has a very low *a priori*

likelihood.  Add 10 magnitudes, and this would describe a common

microlensing event in the Magellanic Clouds.  And the assembled light

curve was looking cuspy and symmetric, the characteristic shape of

microlensing events.  The words were first pronounced in print by

Maciej Mikolajewski (Nicolaus Copernicus University).  Although a

microlens is a mere accident of geometry, it's a new toy in astrophysics;

and as every parent (or dog-owner) knows, new toys are vastly more

interesting than the old ones.

    By November 4, hundreds of astronomers were discussing this in

seminars, corridors, and internet groups.  More variable-star observers

took notice, and the star's fade to quiescence was thoroughly covered.

Calls went out for images of Cassiopeia during the sparsely covered rise

and maximum, and digital photos were retrieved by British amateurs Keith

Geary and Mike Collins.  Careful analysis of Geary's image by Michael

Richmond (Rochester Institute of Technology) showed how these unfiltered

images from short-focus lenses in crowded star fields could yield

calibrated magnitudes.  And most importantly, images were found in the

test runs of the new northern station of the All-Sky Automated Survey

(ASAS) at Haleakala, Hawaii.

    ASAS is an automated array of telephoto lenses which image the entire

visible sky every night through V and I filters.  The images flow through

an analysis pipeline and produce beautiful long-term light curves of

variable stars.  A collaboration between Bohdan Paczynski (Princeton) and

Grzegorz Pojmanski (University of Warsaw), ASAS-South has been in operation

at Las Campanas Observatory in Chile since 2000, and has become a

spectacularly useful tool.  In 2006, a northern station began test runs

at Haleakala.  Pojmanski, the modern virtuoso of variable-star research,

quickly retrieved and analyzed many images of the star for us, with 5

covering rise and maximum light.  These were especially critical, since

they yielded calibrated magnitudes through standard filters.

    Our discussions with microlens experts at Ohio State (Subo Dong, Scott

Gaudi, Andy Gould) and Notre Dame (Dave Bennett) taught us the importance

of finding data acquired during the rise and maximum.  A foreground unseen

star or planet crossing the line of sight to a distant star should produce

a *symmetrical* disturbance in the image due to microlensing.  The light

curve has to be achromatic, symmetric, and follow a specific mathematical

shape.  This is a very exacting requirement.  After Dong's careful

splicing of early magnitudes from small cameras, the ASAS magnitudes, and

the extensively observed decline portions, it was evident that the entire

light curve really did conform in detail to the prediction based on

microlensing.  The accompanying figure shows excellent agreement with the

theory: at about 2200 UT on Halloween (local midnight in Transylvania), an

unknown massive object apparently crossed in front of GSC 3656-1328.  

    In a microlens event, a foreground star has to accidentally cross

within ~20 microarcseconds of the source star.  This is extremely rare

("one in a million per year"), and therefore searches concentrate on

regions where there are many thousands of source stars (other galaxies),

or lines of sight containing thousands of potential lensing stars (near

the plane of the Milky Way).  The "optical depth" to microlensing

increases as the square of the distance, so all known microlensing

events are very distant.

    That's why all microlensed stars are quite faint, and why this event

was so surprising.  Dave Bennett estimated that an event like this --

high-magnification microlensing of a bright star -- should be seen from

Earth only once per 30 years, and even that rate assumes that we never

miss any.  Is it reasonable to associate this brightening with such an

unlikely event?

    Perhaps.  First, since the set of all possible unlikely events is

infinite, individual unlikely events happen frequently!  Second, it's

worth noting who made the discovery: a man who has been sweeping the

sky for novae and comets, and finding them, for... well, for 40 years

(his first discovery was in 1968).  No other event like this has been

seen by Tago, or by anyone else.  So maybe one per 30 years is about

right.

    But this estimate is for high-magnification microlensing of bright

stars.  Slightly fainter stars (say V=13) are much more numerous, and

lower-magnification events are more likely since they do not require

very close approaches to the line of sight.  So a targeted all-sky

search for such events with wide-field cameras might indeed find a decent

supply.  And this is an exciting prospect, because the accident of

microlensing enables us to discover fine details of invisible stars (the

lenses), including planets, even Earth-mass planets.  So powerful a

diagnostic exists because the lens passes within ~1 AU of the line of

sight, offering a transient opportunity to scan the lens's gravity

field at planetary distances.  About a dozen planets have been discovered

with this technique, all in fields well-monitored by the established

microlens search programs (Galactic bulge, LMC, etc.).  The Halloween

transient lures us with the prospect of learning more by extending this

to the entire sky.

    So... did the lensing star on Halloween have planets?  We don't know -

we managed to obtain enough light curve to certify the lens, but not

the nearly continuous coverage needed to see the small deviations from

symmetry that are the signature of planets.  We just weren't ready this

time.  But there will be a next time.  And it is likely that the heroes

of next time will be mainly amateur astronomers, who are globally

distributed and can pounce quickly on newly discovered transients,

obtaining the critical data in the narrow window around maximum light.

Even in the present case, when hundreds of professional astronomers were

excited by the event, the main heroes were amateurs (Tago, Krajci, Koff;

and the many astronomers sending observations to internet news-groups).

    While I was writing this paper, Andy Gould told me a fascinating

story about the involvement of an amateur scientist in the *theory* of

microlensing.  Einstein wrote his famous theory paper in 1936, where he

calculated that such a stellar event would probably never be observed.

There is evidence he worked out the theory in 1912, and then set it aside

as of little interest.  The Hungarian engineer Mandl pestered Einstein

about it, even travelling to Princeton when Einstein didn't answer his

mail.  Eventually Einstein gave in and published the result in 1936...

and then wrote a followup letter to the editor of *Science* saying that

the result, which "Mister Mandl squeezed out of me", was useless but

publishing it was good because "it makes the poor guy happy".  Other

professional astronomers remained similarly myopic, until Paczynski's

famous revival of the subject in 1986.

    Finally, do we know with certainty that the Halloween transient was

a microlensing event?  No; the case seems very strong, but certainty is

too lofty a standard.  I think we can say this: the event signifies

either a microlens, or something even more exotic: a unique variable star

that manages to mimick a microlensing light curve in fine detail, and not

offer any of the credentials of known variables.  Search programs in the

LMC have monitored nightly millions of A stars, and never found a light

curve like this -- or more correctly, "everything that *looked* like a

a microlens *was* a microlens" (Dave Bennett).  Both hypotheses are

worth exploring; but Ockham's Razor now seems to strongly favor

interpretation of the transient as a spectacular microlensing event, which

may kick off a new chapter in the history of professional-amateur

collaboration.