Recurrent Novae

With the recent outburst of RS Ophiuchi still underway, now is a great time
to discuss recurrent novae. These cataclysmic variables grab our attention
and spark our imaginations because of the incredible amplitude of their
outbursts, typically 8-12 magnitudes, and the rarity of these spectacular
events. Many of these outbursts are once-in-a-lifetime events. Like an
apparition of Halley's comet, witnessing an outburst of T CrB twice in a
lifetime would be a matter of uncommon luck, longevity or both.

In the General Catalog of Variable Stars (GCVS) recurrent novae are included
in the same category as novae, with the main distinction being the features
of their light curves.

"According to the features of their light variations, novae are subdivided
into fast (NA), slow (NB), very slow (NC), and recurrent (NR) categories.

NR Recurrent novae, which differ from typical novae by the fact that two or
more outbursts (instead of a single one) separated by 10-80 years have been
observed (T CrB)."

This implies that the outburst mechanism, orbital periods, spectra and the
nature of the components of these close binaries are the same or very
similar. To understand recurrent novae we need to understand novae first,
and then make distinctions.

Novae are close binary systems with orbital periods from 0.05 to 230 days.
The primary of the system is a hot white dwarf star while the cooler
secondary components may be giants, subgiants, or dwarfs of K-M type.

Although few novae have been caught in the very act of rising to eruption,
it is generally accepted that the time it takes to go from restless
quiescence to full max is 1- 3 days. The same is probably true for recurrent
novae.

The cause of a nova eruption is a thermonuclear reaction on the surface of
the white dwarf. After years of mass exchange between the binary pair,
temperature and pressure at the surface of the white dwarf build
sufficiently to cause the layer of accreted material to explode like a
hydrogen bomb. This bomb, however, can have the mass of 30 Earths! Once the
temperature becomes high enough, this layer begins to expand. Minutes into
the process the shell can be radiating at 100,000 solar luminosities and
expanding outwards at 3000 km/s. Eventually the shell envelopes the entire
binary and the orbital motion of the pair acts like a propeller to whip
things up. After 1000 days or so the envelope expands to the point it can be
seen as nebulosity surrounding the pair. Over hundreds of years the shell
dissipates into the interstellar medium.

Most novae probably erupt more than once in their lifetime, with the mass of
the white dwarf determining the amount of accreted material that needs to
accumulate before triggering on outburst. Systems with a white dwarf of 0.6
solar masses might take as long as 5 million years between eruptions. A
system with a 1.3 solar mass white dwarf might only take 30,000 years
between eruptions.

So are recurrent novae simply the same type systems with even more massive
white dwarfs? The accretion rate of a system with a 1.4 solar mass white
dwarf could have a recurrence time of less than 100 years. T Pyx may be one
such system, but it is unclear at present if the outburst mechanism for all
recurrent novae is the same as novae, or if some are the result of accretion
by Roche-lobe overflow or stellar winds, or a result of disc instabilities.

Even more interesting is the possibility that recurrent novae may actually
be progenitors of Type Ia supernovae. Observations of novae eruptions and
the resulting nebulae indicate the mixing of the accreted layer with the
outer layers of the white dwarf may cause the white dwarfs to lose mass over
time and repeated eruptions. The heaviest white dwarfs, with their higher
accretion rates, may actually gain mass over time! Although a large part of
the envelope mass is blown away in the wind, these primaries may retain a
substantial part of the envelope mass after hydrogen burning ends. The white
dwarfs in some recurrent novae have now grown up to near the Chandrasekhar
mass limit and might soon explode as a Type Ia supernova.

With so few known examples and the rarity of these events it is no wonder
that recurrent novae eruptions are extremely interesting to astronomers.
Monitoring these stars for outbursts over decades of relative inactivity is
still one of the extremely valuable contributions visual observers can
provide to science.

Finding the stamina and determination to follow such stars is no small task.
Even Leslie Peltier, one of the greatest AAVSO observers of all time, had an
"unhappy affair" with T CrB that can serve as a lesson to us all. In
'Starlight Nights' he writes:

"From 1920 on I watched it closely at every opportunity. For more than
twenty-five years I looked in on it from night to night as it tossed and
turned in fitful slumber. Then one night in February 1946 it stirred, slowly
opened its eyes, then quickly threw aside the draperies of its couch and
rose!
Full eighty years had passed since the star had shattered the symmetry of
the Northern Crown. And where was I, its self-appointed guardian on that
once-in-a-lifetime night when it awoke? I was asleep!"

Peltier had set the alarm for 2:30 AM to observe morning variables. When he
got up the sky was clear and the stars were shining, but feeling he might
have a cold coming on he decided to go back to bed. He goes on to describe
his personal relationship with the star, one that many of us feel for our
favorite variables, and how it changed after that.

"I alone am to blame for being remiss in my duties, nevertheless, I still
have the feeling that T could have shown me more consideration. We had been
friends for many years; on thousands of nights I had watched over it as it
slept and then, it arose in my hour of weakness as I nodded at my post. I
still am watching it, but now it is with wary eye. There is no warmth
between us any more."

In more recent times, CI Aql had been suspected of being a recurrent nova
even though only one recorded outburst had occurred in 1917. As such it was
included in the BAAVSS Recurrent Objects Programme for many years. For
reasons he still will not discuss with even the best of friends, Gary
Poyner, coordinator of the program, decided to drop CI Aql from the list in
2000; literally weeks before it erupted again for the first time in over 80
years! Sorry Gary, but its just too good a story not to recount.

Below is a table of known recurrent novae.
Try not to sleep through the next eruption of any of these unpredictable
stars.

Name
RA/Dec (2000)
Magnitude range
Years of known outbursts
Chart availability

T Pyx
09 04 41.53 ?32 22 47.2
6.5 v - 15.3 v
Outbursts in 1890, 1902, 1920, 1944 and 1966
AAVSO charts

IM Nor
15 39 26.47 ?52 19 18.0
7.8 V - 22.0 j
Outbursts in 1920 and 2002
AAVSO charts

T CrB
15 59 30.19 +25 55 12.1
2.0 p - 11.3 p
Outbursts in 1866 and 1946
AAVSO charts

U Sco
16 22 30.80 ?17 52 44.0
8.8 V - 19.5 V
Outbursts in 1863 and 1999
AAVSO charts

RS Oph
17 50 13.12 ?06 42 28.2
4.3 v - 12.5 v
Outbursts in 1898, 1933, 1958, 1967, 1985, 2006
AAVSO charts

V745 Sco
17 55 22.27 ?33 14 58.5
11.2 p - 21 j
Outbursts in 1937 and 1989
No AAVSO charts

V394 CrA
18 00 25.97 ?39 00 35.1
7.2 V - 18.8 V
Outbursts in 1949 and 1987
No AAVSO charts

V3890 Sgr
18 30 43.32 ?24 01 08.6
8.4 p - 17.2 p
Outbursts in 1962 and 1990
Very poor AAVSO chart; lettered sequence only

CI Aql
18 52 03.57 ?01 28 39.4
8.8 V - 15.6 p
Outbursts in 1917 and 2000
AAVSO charts