Type Ia supernovae and the ESSENCE supernova survey

Type Ia supernovae and the ESSENCE supernova survey
Kevin Krisciunas
There are two basic kinds of supernovae. Roughly 1 out
of 1000 stars that forms has a mass greater than 8 Msun.
Shortly after such a star develops an iron core, it explodes
as a Type II SN.
Type II SNe
show atomic
lines in
Common photometric bands in astronomical photometry:
mean wavelength (m)
Optical and
light curves
of the Type II-P
SN 2003hn.
Krisciunas et al. (2009)
When a single massive star blows up as a Type II SN:
1) elements heavier than iron (up to uranium) are
produced; and
2) a compact remnant is left behind (either a rapidly
rotating neutron star or a black hole)
3) 99 percent of the energy of the explosion is radiated
away in the form of neutrinos
The most common end state for a star is a
white dwarf.
S. Chandrasekhar (1910-1995)
From a consideration of the equation of state of
relativistic degenerate matter, Chandrasekhar
discovered in 1930 that more massive white dwarf
stars were smaller than less massive ones. If a
white dwarf star had a mass approaching 1.4 Msun its
radius ---> 0.
Binary stars are very common in the universe. Some
are close enough that they influence each other's evolution.
It is generally believed that a Type Ia SN is a
carbon-oxygen white dwarf that acquires mass from
a nearby donor star. When the mass of the WD exceeds
1.4 Msun the WD completely obliterates itself.
The spectra of Type Ia supernovae are characterized by
having no hydrogen emisssion. The prime signature
is a blue-shifted absorption line of singly ionized silicon
observed at ~6150 Angstroms.
Two Type Ia
supernovae at
about the same
number of days
after explosion.
The spectra are
almost identical.
Krisciunas et al. (2007)
Light curves of SN 2004S in the optical and
near-IR bands.
According to Kasen (2006), the secondary hump in
the near-IR light curves of Type Ia supernovae is due
to a transition from doubly ionized atoms to singly
ionized ones.
The astronomical magnitude system was originated
by Hipparchus in the 2nd century BC. The brightest
stars in the sky are said to be “of the first magnitude”.
The faintest stars visible to the unaided eye are 6th
magnitude. For two stars of intensity I1 and I2 their
apparent magnitudes are related as follows:
m2 – m1 = log (I2/I1)
Thus, if we receive 100 times as many photons per
second from star 1 than from star 2, star 1 is 5 magnitudes
brighter than star 2.
B-band light
curves of two
Type Ia
The “decline
rate” is the
number of
magnitudes it
gets fainter
in the first
15 days after
The distance at which the radius of the Earth's orbit
subtends an angle of 1 second of arc = 1 parsec
(3.086 X 1013 km). Astronomers needed a standard
distance at which to compare the apparent magnitudes
of stars in order to be able to compare their instrinsic
brightness. They chose a standard distance of 10 pc.
The apparent magnitude a star would have if it were at
a distance of 10 pc is called the absolute magnitude (M).
M = m + 5 – 5 log (dpc)
mM = 5 log (d) – is a measure of distance, and is
called the distance modulus.
At optical wavelengths, the absolute
magnitudes of Type
Ia supernovae at
maximum brightness
are related to the
decline rate parameter.
These supernovae are
candles. The brighter
ones are 4 billion
times brighter than
the Sun!
Garnavich et al. (2004)
Two independent groups
of astronomers announced
in 1998 and 1999 that
the universe was not just
expanding, but it was
accelerating as it expanded.
Science magazine hailed
this as the “breakthrough
of the year”. How can
we understand this
deduction made from
observations of Type Ia
For nearby galaxies the velocity of recession is related to
the distance (in Megaparsecs) according to Hubble's Law:
V (km/sec) = H0 DMpc ,
where H0 ~ 72 km/sec/Mpc is the Hubble constant.
The redshift z =  (~ v/c for low redshift).
Beyond a redshift of ~0.1 we must must also
consider the curvature of space-time, which is
related to the mass density parameter M = crit .
After much grinding away, we can obtain an expression
for the effective distance (aka proper motion distance):
If the density of the universe is very small compared to
the critical density, then
The arrival rate of photons from a distant galaxy in an
expanding universe is diminished by (1 + z). The
frequency of the photons is likewise diminished by
(1 + z). The flux is then related to the luminosity by:
We can then define a luminosity distance to be:
For extragalactic astronomy, where distances are measured
in Mpc instead of parsecs, the formula for distance modulus
then becomes:
The curvature of the universe depends on the
mass density parameter M = crit , where
crit = 3 H02 / [ 8 G ] .
If we scale Einstein's cosmological constant
as follows:
 c2 / [3 H02 ] ,
this is the (dimensionless) scaled Dark Energy
parameter. The universe has flat geometry if
M + 
Loci of distance
modulus vs. redshift
fan out in the Hubble
diagram, depending
on what the mass
density parameter
M and the Dark
Energy density are.
It is common to pick one of the loci in the previous diagram
as a reference and to plot a differential Hubble diagram.
The expectation prior to 1998 was that Type Ia supernovae
would fall along the “open” line in this diagram.
The gravitational attraction of all the matter in the universe
is not enough to show a deceleration of the universe.
The SN points fall along a curve that is most easily
interpreted as evidence for a positive cosmological
Riess et al. (2004)
Instead of using the strategically chosen bins to provide
the medians shown in the previous diagram, a graph of the
individual points (along with some ESSENCE objects) is
not nearly as convincing.
Krisciunas et al. (2005)
Also, how do we know that distant supernovae are
just like nearby ones?
Previous SN survey work had relied on photometry
taken with a variety of telescopes, cameras, and filters.
Even with knowledge of the effective transmission
functions of different filters, it is difficult to unscramble
differences in the photometry owing to features of
the hardware and the intrinsic differences amongst
What was needed was a systematic search carried out
on the same telescope, with the same camera, and with
the same filters.
ESSENCE = Equation of State.
SupErNovae trace Cosmic Expansion
Objects discovered with CTIO 4-m + prime focus camera
● Almost all photometry from CTIO 4-m
● Oct-Nov observing every other moonless night 2002-7
● 32 principal fields (each 0.6 deg X 0.6 deg)
● Imagery obtained in R, I bands (rest frame U, B, V)
● Spectra from Keck, VLT, Gemini, Magellan
● Some optical and IR imagery from HST and Spitzer
We observed on 191 nights with the CTIO 4-m.
5458 R- and I-band images
2000 transients detected by pipeline
400 spectra obtained
~220 Type Ia supernovae identified
Pipeline helped
identify flux
variable objects
and provided
light curves.
Histogram of redshifts of ESSENCE Type Ia supernovae
discovered from 2002 (bottom) to 2007 (top).
objects that
were also
with HST.
We found a
lot of blue
slow decliners.
The appropriate formula for the luminosity distance
given flat geometry and a non-zero Dark Energy
density that is not necessarily the same as Einstein's
cosmological constant is:
Here w = P / (c2) is the equation of state parameter.
If w = -1, then the Dark Energy is equivalent to
Einstein's cosmological constant.
Our differential Hubble diagram can provide a value
of w, but we need to control systematics in the photometry
to the level of a few hundredths of a magnitude.
w = 1 if
Dark Energy
is the same as
The goal of the ESSENCE project was to try to measure
the equation of state parameter to +/- 10 percent.
We're still working on the data, but we have a pretty
good idea what our final result will be.
Differential Hubble diagram of nearby Type Ia supernovae
(red), ESSENCE objects (black) and 1st year of Legacy
SN search (blue).
SN data combined with data from baryon acoustic oscillations
(BAO) give M +  = 1 at the 68 percent confidence level.
Also, within 1- the equation of state parameter is
equal to , indicating that Einstein (1917) might
have been right after all.
Our preliminary answer is w = 
If w is less than , then eventually the universe will
be accelerating so fast that everything will be ripped
apart. This is called the Big Rip.
If w = 0.98 or so, then various string theory models
could hold.
We live at
a curious
time in
when the
matter and
Dark Energy
are comparable.
What's next?
Dark Energy Survey (w/ refurbished CTIO 4-m)
Large Synoptic Survey Telescope (LSST)
Joint Dark Energy Mission (JDEM)
pushing the rest-frame optical and infrared
Hubble diagrams to higher redshift
At near-IR
Type Ia supernovae are not
just standardizable
candles. The
slow decliners
and mid-range
decliners are
standard candles.
Krisciunas et al. (2004a)
Decline rate relations in the near-IR
In the near-IR
Type Ia SNe
that peak early
are standard
candles, even
if they are fast
decliners. The
fast decliners
are subluminous
in the optical

similar documents