[AAVSO-DIS] Status report on U Scorpii by Brad Schaefer
Matthew Templeton
matthewt at aavso.org
Wed Mar 3 09:36:00 EST 2010
Hello everyone,
Dr. Brad Schaefer has sent out a *very* comprehensive update on what data
has been collected during the U Sco outburst, and what we've learned so
far. This outburst of U Sco is by far it's best observed outburst in
history, and with the addition of multiwavelength data, U Sco 2010 may be
the best-observed nova outburst of any star in astronomical history.
For updates on the progress of the outburst, please keep an eye on the
AAVSO's news page on this event: http://www.aavso.org/news/usco.shtml
Thanks to all of the observers who have contributed, and to those who will
continue to. Please keep observing!
Clear skies,
Matthew
Date: Tue, 2 Mar 2010 23:57:15 -0600 (CST)
From: Brad Schaefer <schaefer at grb.phys.lsu.edu>
Subject: U Sco eruption; overview to Day D+33
Hi;
A lot has been happening with the U Sco eruption, and I think it
appropriate to give another update and overview of what is going on. The
idea is to let everyone know what sort of data is being taken and a lot of
the key points of the incoming data. In all, U Sco is looking to be the
best observed nova eruption of all time, I think surpassing even the 2006
eruption of RS Oph. My two prior broadcast reports were on Days D+4 (just
after the peak) and D+9 (just around the start of the plateau). And we
are now just at the very end (likely) of the plateau phase, with U Sco
poised to fade rapidly.
OVERVIEW:
U Sco went off when predicted (within the error bars quoted in
2005), and because of this we had many people around the world looking for
the eruption. The discoverers (Barbara Harris and Shawn Dvorak) had been
checking U Sco nightly since it came out from behind the Sun, so their
discoveries were intentional and the result of a lot of hard work. The
rise started a few hours earlier but was missed due to the narrow range of
longitudes with the short morning visibility lying over oceans. Once
discovered, the world was alerted fast, and many telescopes (in space and
on the ground) spun around swiftly. In the first day, we got X-ray, UV,
optical, and IR spectroscopy and photometry.
U Sco is the all-time fastest declining nova, for this eruption it
declined by 2 magnitudes in about a day and a half (t_2~1.5 days) and by 3
magnitudes in just over three days (t_3~3.1 days). During this fast
decline, U Sco was showing optical flickering with timescales of half an
hour and amplitudes of half a magnitude. This flickering was well
observed and characterized by the fast time series photometry provided by
small telescopes around the world. Flickering in the transition region
would not surprise nova experts (there has always large scatter in light
curves, and some novae show huge flares and oscillations), perhaps even in
an apparently smooth light curve. But, to my knowledge, this is the first
time that any nova has been intensely observed with fast time series
photometry. I know of no theoretical understanding how the expanding
shell can have such fast and large amplitude variations.
The character of the nova underwent a fast change around Day D+11.
(A) The nova suddenly became bright as a super-soft X-ray source (SSS).
(B) Eclipses suddenly started to be visible in the X-ray, UV, and optical.
(C) The fast brightness decline in UV/opt/IR stopped, and the light curve
went roughly flat from Day D+11 to around D+30. This is the plateau
phase. (D) The optical flickering largely stopped.
The interpretation of all these changes presumably is simply that
the expanding shell (with the resultant geometrical dilution of the column
density) got sufficiently optically thin that we can see back to the
binary in the middle. Another way of saying this is that the photosphere
receded back to the envelope surrounding the white dwarf. With the shell
becoming suddenly optically thin, we can see the eclipses going on in the
central binary. With the shell thinning out, we can see the soft X-rays
produced by the envelope of ongoing hydrogen burning around the white
dwarf (causing the turn-on of the SSS). The Hachisu & Kato model for
plateaus in recurrent novae has the plateau caused by the SSS flux being
reprocessed on the reformed accretion disk and other material in the
system.
A unique and wonderful feature of the U Sco system is that it has
total eclipses (in quiescence) with a 1.23 day period and a known
ephemeris. With the advance knowledge, we have arranged for photometric
and spectroscopic monitoring of the eclipses during the plateau in the
X-ray, UV, and optical bands. This allows us to 'eclipse map' the size
and brightness and temperature distributions of both the continuum and the
line-emission regions, and do this as a function of time throughout the
eruption! Wow, we could not have fantasized a better arrangement.
The eclipses have been visible during the plateau phase in the
X-ray, UV, and optical. (I know of no IR fast time series.) The eclipses
have been variable in depth, especially in the X-rays, but typical depths
have been ~0.7 mag (corresponding to a factor of 2X drop in brightness).
The spectra (in all wavebands) do not change substantially from inside to
outside of eclipse, indicating that there is no central hot spot. The
duration of the eclipse is roughly 8.6 hours (0.29 in phase). This is
longer than the duration in quiescence, and implies that the photosphere
is ~4 R_sun in size. The depth of eclipse then implies that the
photosphere is substantially brighter in the center than the edges. The
eclipses are definitely offset in time as compared to the pre-eruption
ephemeris in quiescence by up to 20 minutes. This same effect was poorly
seen in prior eruptions. It means that the center of light has shifted
position from the plateau phase (with the light presumable centered on the
white dwarf) to the quiescent phase (with the light presumably centered on
the offset hot spot).
As the all-time fastest nova, and having a very low-mass burning
envelope, the SSS will soon burn out. This presumably will be when the
light curve drops (as the SSS light is no longer being reprocessed).
>From all the past eruptions, only one eruption has any useable number of
magnitudes in the light curve after Day D+30, and that one has a very
well-sampled B-band drop on Day D+33, where the brightness falls by 2-3
mag over a few days. But we also have 3 isolate V-band points after D+33
with no apparent sign of a sharp drop. So, is there variability from
eruption-to-eruption, or do the colors behave greatly differently? I
don't know, but we'll find out in the next week!
The 2010 eruption has exhaustive coverage in the X-ray, UV,
optical, and IR; both spectroscopy and photometry. A higher level goal is
to have the all-time best data set to present to theorists and modelers.
I expect that they can use this fantastic data set to test and refine
their models as never before possible. Aspects of the U Sco eruption that
are unique and very valuable are the eclipse mapping, the intensive fast
photometry, and the polarimetry. Challenges to theorists include the fast
flickering of the shell, the 'Batman cowl' triple line profiles in the
early spectra, and how the orbital period can possibly *decrease* across
the eruption. At a higher level, this eruption should provide a sure
answer as to whether U Sco (and by extension the other long-period
recurrent novae) has a white dwarf gaining mass, hence making recurrent
novae a progenitor of Type Ia supernovae. The progenitor problem is now
uber-important in astrophysics, and U Sco has a lot to say for it.
X-RAY OBSERVATIONS:
The U Sco eruption has caused the fast mobilization of all the
X-ray detectors up in the sky. Here is a summary of what I've heard
about:
Swift D+0 ongoing 1000s each orbit, for most orbits since D+0 (!!!)
XMM D+22 60ks, great spectrum, one eclipse mapped
XMM D+35 64ks, upcoming, 3/4 Mar UT, eclipse mapping
SUZAKU D+7 to D+17 100ks, five epochs
Chandra D+17 23ks (ATel2451), spectra mainly N lines
INTEGRAL D+0 to D+13 No detection in hard X-rays or Gamma-rays
RXTE ongoing No detection in hard X-rays
Fermi GBM ongoing No occultation detection, hard X-rays only
MAXI ongoing No detection, confusion with nearby Sco X-1
The X-ray story is that we had three pointing satellites on target
in the first day, but U Sco was initially invisible in the hard and soft
X-rays. Swift first saw U Sco on Day D+3 as a relatively hard (~5 keV)
and very faint (0.0025 ct/s, 4-sigma) source with luminosity of 1.7x10^33
erg/s (ATel2419). After a Moon gap, on Day D+11, Swift saw the fast rise
of a relatively bright and definitely soft (kT=28+-8 eV) source, with the
hard/faint source still present. This super-soft X-ray source (SSS) is
from the very hot region around the white dwarf where hydrogen burning is
still ongoing. After the initial SSS turn-on, U Sco slowly brightened
from ~0.4 ct/s on D+11, to ~0.8 ct/s on D+20, to ~1.2 ct/s on D+26, to
~2.0 ct/s on D+31. The kT has held constant throughout this time, with a
column density of ~5x10^21 cm^-2. Throughout, we see 'flickering' in the
X-rays, with this being easily apparent in the single stares with Chandra
and XMM. Also, from Day D+14 or perhaps earlier, some 'ratty' X-ray
eclipses are visible. It is hard to define the eclipse shape, partly
because of the flickering, and partly because the light curve appears to
vary.
The hard/faint component is likely from the fast expanding gas in
the shell, perhaps from internal shocks. (Unlike RS Oph, there is no
ambient gas from a pre-existing wind to form a natural 'target'.) The
bright/soft component first became visible rather suddenly on Day D+11,
and this is certainly caused by the fast-expanding thin shell of ejecta
suffering geometrical dilution so that the photosphere recedes back to
close to the white dwarf. That is, the shell has thinned out enough so
that the soft X-rays from the hydrogen burning on the white dwarf can
suddenly get out.
I don't understand the X-ray eclipses. I would have expected the
X-rays to come from a region near the white dwarf, and suffer steady and
deep eclipses. Rather, the eclipses appear to come and go, and there is
flickering even in the middle of the eclipse. And the eclipses appear to
last 30-50% of the orbital period [1.23 days], especially in the last few
days. I guess that this is simply pointing to the X-ray emitting region
being larger than the companion star (~2.3 R_sun) and larger than the UV
emitting region.
The X-ray flickering is hard to understand. It appears even
during middle eclipse, so apparently it must not come from near the white
dwarf. The time scale for the flickers is typically 20 minutes, with a
typical amplitude of 30%. That is a lot of energy coming from a small
volume. The simultaneous *optical* light curves show no flickering.
Now that we are coming to the end of the plateau phase, I expect
(with only moderate confidence) that the X-ray will suddenly turn off.
We know the sharp drop-off in the optical light curve at Day D+33 (for the
1945 eruption), and the Hachisu & Kato model for the plateaus in recurrent
novae is that these are powered by the SSS being reprocessed by the
accretion disk, so a drop in the optical corresponds to the SSS turn-off.
The timing for the turn-off also must be very fast (after all, U Sco is
the all-time fastest known nova eruption), although the model given by
Hachisu & Kato (2010, ApJ, 709, 680) cannot be extrapolated to such short
values as t_3=2.6days.
ULTRAVIOLET:
The UVOT on Swift has been getting awesome spectra and time
coverage from the first day on. Much of the UVOT data comes from the UV
grism, which gives nice spectra from ~1800A to the visual band. The XMM
Optical Monitor (OM) has a nearly identical instrument as the UVOT, and it
was taking UV grism data during the entire 60ks run on Day D+22 (an
eclipse mapping epoch!).
These spectra allow for a detailed record of the continuum flux
throughout the eruption. This flux has declined by a factor of 12X from
Day D+2 to D+12, followed by a fairly flat light curve out to D+26.
Inexplicably, from D+26 to D+31, the UV continuum flux has dropped by
about a factor of 3X. This *drop* in UV flux corresponds to the same time
that the X-rays are brightening by 2X, the optical light is essentially
flat, and the IR is dropping (in a complicated way).
The UV light carries a lot of the luminosity of the eruption.
Quantifying the total amount of energy radiated by the nova has suddenly
become important because Mike Shara has put out a paper on ArXiv just two
weeks ago (Shara et al. 2010, ApJLett in press, arXiv:1002.2303) that
demonstrates that the mass ejected in novae is simply proportional to the
total radiative energy, with the constant of proportionality being known.
The total radiative energy is what you get when you integrate the nova
light over all wavelengths and all time. With the ejected mass for
recurrent novae being so important for the Type Ia supernova progenitor
problem, Shara's result is vital as a new way to get the M_ejecta for U
Sco. (I now know four ways to do this from the current eruption, so we
can get a good measure independent of the many big uncertainties inherent
in the old traditional method.) This is a strong reason to keep the
X/UV/opt/IR light curves for the whole eruption.
The UV spectra shows many lines. I have not seen any real
analysis, for example, line IDs, for the UVOT spectra.
The UVOT light curves show blatant eclipses, starting around Day
D+14 or possibly earlier (ATel2442). They have continued throughout the
plateau phase, even up to this morning. The total duration of the
eclipses looks to be ~0.25 in phase. Early in the plateau, the eclipses
had depths 0.5-0.7 mag in all bands (from ~1800A to the U-band), while the
last few days have had depths like 1.5 mag. (The optical eclipse depths
have also been deepening.) This points to the radius of the UV
photosphere being ~2 R_sun.
OPTICAL:
We have a wonderful large database of ~12,000 magnitudes from fast
time series photometry from small telescopes operated by 16 AAVSO and CBA
observers spread all around the world. (I expect this to triple in size
by the time it is all over and I've collected from more observers.)
Roughly, U Sco has fast photometry for half of its time since eruption.
(This is awesome, especially given that any one observer can pick up U Sco
for only ~2-3 hours in the morning sky before dawn stops the run. So
we've been patching all the observers together into a single consistent
string of photometry.) This work has revealed fast flares ('flickering')
for the time before the start of the plateau. The flickers go up to half
a magnitude on times scales of half an hour. During this time, the shell
was optically thick (e.g., we saw no SSS nor any eclipses), so the source
of the flickering must be the expanding nova shell. But how can a shell
~5 light-hours in size suddenly brighten by 60% on a time scale of 0.5
hours?
This wonderful fast photometry is also critical for measuring the
light curves of eclipses in the optical. The eclipses started at the
start of the plateau. The eclipses started out with depth 0.6 mag, and
8.6 hours in duration (0.29 in phase), see ATel 2452 for light curves.
(This is shallower and longer than optical eclipses during quiescence.)
The optical photosphere was somewhat larger than the companion star, say
around 3 R_sun. By Day D+21, the eclipses had deepened to 0.8 mag and
broadened significantly. By Day D+30, the eclipses had further deepened
to 1.2 mag, and the total duration of the eclipse was 0.33 in phase. I
have no understanding of why the optical photosphere should be expanding
as time gets later in the plateau.
The fast photometry also produces great time-resolved folded light
curves throughout the eruption. And this has mysteries of its own.
First off, the V-band light curve shows something that looks like a
secondary minimum at 0.5 phase with depth of 0.25 mag. (In quiescence, U
Sco shows a secondary eclipse only in I-band, as the accretion disk covers
the companion star.) So there must be some source of light from the
neighborhood of the companion star that is eclipsed by the envelope around
the white dwarf. I can only think that the nova light is heavily
irradiating the companion star so as to make it greatly brighter than
usual, hence allowing secondary eclipses. The second mystery is that U
Sco is brighter at phase 0.25 than at phase 0.75 by 0.11 mag. This
implies some asymmetry in the system, and the only thing that I can think
of is the accretion stream. The two other known recurrent novae most like
U Sco (V394 CrA and CI Aql) both show this same asymmetry during
quiescence.
Several groups have been following U Sco with many filters on a
night-by-night basis for long runs. Ashley Pagnotta and myself are using
the SMARTS 1.3m telescope in Chile to get BVRIJHK from the first day until
31 July. Aaron LaCluyze ('Cluze') and Dan Reichart have been using their
PROMPT telescopes in Chile to get nightly UBVRI time series for the entire
time U Sco is up since the first day. [The Chilean data, including that
of Arto Oksanen with a scope in the Atacama, have all had no problem with
the Chilean earthquake!] Gerald Handler had a 25 night run on the SAAO
0.5m that started just before the eruption (what perfect luck!), and he
has gotten nightly UBVRI *plus* Stromgren b and y. Hannah Worters has
also been taking a lot of UBVRIJHK plus Stromgren y independently from
SAAO. Chile and South Africa have had great weather with few clouds, so
we have wonderful coverage in bands from U to K. With the likely sudden
drop in brightness around Day D+33 (now) with confused color changes, the
UBVRIJHK over the next week will be very interesting to watch as it comes
in.
One science task that Ashley Pagnotta has started is to construct
spectral energy distributions (SEDs) from X-ray to UV to optical to IR for
every day of the eruption. Intergrating under the SEDs over all time and
wavelength will give the total radiative energy put out by U Sco. And
Shara et al. have shown how to use this number to get M_ejecta.
Another science task is to use the light curves to test the
'Universal Decline Law' for novae of Hachisu & Kato (2006, ApJSupp, 167,
59). They predict that the *continuum* light from novae will decline
after the peak initially with a slope of -4.4 (in magnitudes per
logarithmic time since the eruption) and decline at late times with a
slope of -7.5 (in the same units). A practical problem is that the usual
V-band light curve also includes substantial flux from lines, which will
throw off this prediction. They point to a solution of using the
Stromgren 'y' band, which is like a narrow V-band that avoid most of the
emission line flux. So here is a prediction that we can test.
Relatively few observatories can work in the 'y' band. So it was
great good luck that Handler and Worters at SAAO can and did follow U Sco
in the 'y' band. With Handler's data, I find an slope over the first 22
days as -4.6. This is in nice agreement with the Hachisu & Kato
prediction. But Handler's run is now over, so we need to keep following U
Sco in the Stromgren 'y' filter so as to test the second part of the
theoretical prediction. Arlo Landolt and James Klem will fill this gap,
as between them they have many week long runs at KPNO and CTIO for the
next three months. (They both have already come back with data from runs
each, that start just after Handler's run ended.)
Optical spectra are being taken regularly by Fred Walter with the
SMARTS 1.5m on Cerro Tololo, Michael Bode and Matt Darnley with the 2.0m
robotic Liverpool Telescope in the Canary Islands, and John Menke with his
18-inch telescope in Maryland. At least some more optical spectra are
being taken by A. Arai et al., Ramya et al., and Thijs Kouwenhoven &
Hannah Worters. (If you know of more optical spectroscopy, please let me
know.)
One of the weird features of the optical spectra was the early
appearance of a triple feature lines, with the profile looking like a
frontal view of Batman's cowl with high pointy ears. We have wondered
whether this might be produced by a jet, but there is no plausible path to
jet formation, and we are seeing the system edge-on so jets wouldn't have
any significant radial velocity. We have wondered whether this might be
related to the disk, but this can't be as the shell was optically thick
when the triple peak was visible. Fred Walter points to previous cases he
has found (YY Dor, N LMC 2009, V2672 Oph, KT Eri, and now U Sco), with
each of these being a known or suspected recurrent nova. See
http://www.astro.sunysb.edu/fwalter/SMARTS/NovaAtlas/ for Fred's awesome
atlas and the data with the 'Batman' features. Fred has been asking me
about these 'Batman' features for at least 8 years, and I've never known
the answer. So what about recurrent-nova-hood (like a high mass white
dwarf, or a low mass envelope) makes for the 'Batman cowls'?
I have not heard of any detailed analyses of the optical spectra.
But this is much needed. And in combination with the X/UV/opt/IR spectra
from many epochs, the U Sco data set should be perfect for a complete
analysis at many epochs. (I am collecting such spectra for this use.) A
primary science would be to derive the mass of the ejecta. Other primary
science would be determining the abundances, for example to look for
dredge-up material and to see if the matter is rich in CNO material or if
it is hydrogen poor.
INFRARED:
For JHK photometry, we have Ashley Pagnotta and myself taking
nightly JHK images with ANDICAM on the SMARTS 1.3m at Cerro Tololo, Hannah
Worters taking a long running series of JHK images at SAAO, and R. K. Das
and coworkers have JHK photometry on the 1.2m at Mt. Abu.
For infrared spectroscopy, we have Howie Marion getting a peak
spectrum with the IRTF, David Lynch getting a 0.8-2.4 micron spectrum on
11 February with the IRTF, Mark Rushton using the SOFI spectrograph on
ESO's 3.6m NTT at La SIlla, and N. M. Ashok, R. K. Das, and coworkers on
the 1.2m at Mt. Abu. Mark says that he was going to try for Gemini NIRI
time, but I have not heard the outcome. I do not know of any IR
spectroscopy coverage for the next week or month. I have had two
senior people suggesting that the IR spectra are very valuable, perhaps
much more valuable than optical spectra, and so I'd encourage anyone with
IR spectral capabilities to push hard.
POLARIZATION:
Our collaboration's preplanned polarization capability is not
available for this eruption (the polarimeter is off the telescope). So in
my last report, I had asked for polarization measures from anyone. G. C.
Anupama now tells me that A. N. Ramaprakash has obtained multi-band
optical imaging polarimetry on the nights of D+4 and D+5 with the 2m IUCAA
Girawali Telescope in India. I have not heard the results of this work.
It seems these results are critical for being the only way to measure the
non-circularity of the ejecta. We are looking at the U Sco binary
edge-on, so any bipolar shape might be apparent with the polarization
data.
CIRCULARS:
A variety of IAUCircs, CBETs, and ATels have come out. Here is a
summary of what is out there:
IAUC9111 Schaefer Discovery by Harris, confirmed by me, Dvorak disc.
CBET2152 Arai Opt. spect (Higashi-Hiroshima) FWZI=11000 km/s
CBET2153 Ashok IR spec (Mt Abu) FWZI=10000 km/s, HI,OI,HeI
CBET2157 Das IR spec & JHK (Mt Abu) triple peaked lines
ATEL2411 Anupama Opt. spectrum on Jan 29, FWHM~7600 km/s
ATEL2412 Manousakis INTEGRAL, no detection on first day
ATEL2419 Schlegel Swift detection of weak hard X-rays on D+3
IAUC9114 Worters Opt phot (SAAO 1.0m) flickering on D+8
ATEL2430 Schlegel SSS turn on D+11, kT=28+-8 keV, L~2x10^37 erg/s
ATEL2442 Osborne UV eclipses (~0.6mag deep) start D+14, X-ray also?
ATEL2451 Orio Chandra D+17, T~510,000K, flux~3x10^-11 erg/cm2/s
ATEL2452 Schaefer Deep optical eclipses start D+14, duration 8.6h
U SCO FACTS FOR EASY REFERENCE:
Here is a collection of values measured for U Sco for easy
reference. The source is mainly my huge summary of all recurrent
nova photometry (Schaefer, 2010, ApJSupp, in press, arXiv:0912.4426), but
also the light curve data I've been collecting for this eruption:
T0 = eruption start 2010 Jan 27.8 = JD 2455224.3
Peak V=7.5 on 2010 Jan 28.1 = JD 2455224.6
Discovery Barbara Harris at V=7.8, Shawn Dvorak
Discovery time 2010 Jan 28.4385 = JD 2455224.9385
Distance 12 +- 2 kpc
Extinction E(B-V)=0.2 mag
Period 1.2305631 days
Secondary star G5 IV (5300-6000 degrees K), mass=~1 M_sun
White dwarf mass 1.35 +- 0.05 M_sun
Companion star radius ~2.3 R_sun
Binary seperation ~6.4 R_sun
Eclipse contact phases -0.0857, -0.0132, +0.0121, +0.0915
Duration total eclipse 45 minutes (in quiescence)
Overall eclipse dur. 5.2 hours (in quiescence)
Vpeak 7.5
B-V peak 0.3 mag
M(V) peak -8.5 mag
Vmin 17.6
M(V) quiescence 2.0 mag
Quiescent colors U-B=-0.3, B-V=0.4, V-R=0.4, V-I=0.9, V-J=1.05
t_2 1.2 days
t_3 2.6 days
Eruptions 1863,1906,1917,1936,1945,1969,1979,1987,1999,2010
Recurrence timescale 10.3 years
Galactic Latitude 22.5 degrees [so little extinction, low density]
LOOKING FORWARD:
I expect that U Sco will start turning off fast, starting about
*now* (i.e., Day D+33). But the details of this are unclear, because we
only have a light curve for after Day D+30 in *one* prior eruption (the
one I found on the Harvard plates in 1945) with this showing a sharp drop
of ~2 mag in ~4 days (see Figure 14 of Schaefer 2010, arXiv:0912.4426),
while the V-band only has 3 scattered points from different eruptions that
apparently do *not* show the sharp drop off. Such differences between B
and V seem unphysical, but here we are at Day D+33 and the X/UV/opt/IR
bands are all behaving whacko differently from their nearest neighbors.
One implication is that I don't really know what to expect in detail or
with confidence. A further implication is that it becomes vital to follow
the U Sco light curve in all bands over the next week.
The eclipse mapping over the next week will be particularly
interesting, as we can map out the flux-emitting and line-emitting
regions as they are changing fast.
Maybe, in ten days, U Sco will have faded down to about a
magnitude above its normal quiescent level. The time scale to fade the
last magnitude is unknown. The optical depth of the large expanding shell
is unknown, but possibly measurable from the depths of the eclipses.
Presumably, the eclipse times will transition from their eruption-shifted
times back to the quiescent ephemeris. But the quiescent ephemeris will
now have a different period due to the mass lost during the eruption.
I'll be keeping watch for the next decade to collect many exact eclipse
minima times and derive a post-eruption orbital period accurate to ~0.1
ppm. I'll also be keeping track of the accretion rate (from the B-band
flux) so that I can predict the next eruption time.
And sometime in ~2020, we will have another eruption of U Sco!
Cheers,
Brad
More information about the Aavso-discussion
mailing list