PG_Lecture_Dec18_2008

Download Report

Transcript PG_Lecture_Dec18_2008

X-ray and UV
spectroscopy of the
Sun and Flare Stars
Ken Phillips
Postgraduate Lecture
December 18, 2008
Recommended reading
Introduction to Stellar Astrophysics, vol. 2 (Stellar
Atmospheres). E. Böhme-Vitense (CUP, 1989)
The Solar Transition Region. J. Mariska (CUP,
1992)
Ultraviolet and X-ray Spectroscopy of the Solar
Atmosphere. K.J.H. Phillips, U. Feldman, E. Landi
(CUP, 2008)
Space Science (eds L. Harra and K. Mason: Imp.
College Press, 2004) esp. chapter 8
Atomic Spectra. 2nd ed. H. G. Kuhn (Longman
1969)
Atomic Spectra and Atomic Structure. G. Herzberg
(tr. Spinks) (Dover 1944)
Atmospheres of “active” stars
and the Sun
Flare stars and the Sun have hot
atmospheres, usually a corona
(temperature ~ 106 K) plus a
chromosphere (~10,000 K) and “transition
region” (~105 K).
These temperatures are generally much
hotter than their surface temperatures.
E.g. The Sun has surface (photospheric) T ~
6000K but its corona has T~106K
Why are some stellar
atmospheres hot?
Some stellar atmospheres are hot because there
is a non-radiative energy source.
This is associated with a magnetic field present
in the convective zone of its interior.
Heating may occur by either dissipation of MHD
waves or numerous tiny flares.
The magnetic field continually re-created by an
α-ω dynamo action in the stellar interior –
differential rotation (ω) + convection (α).
HR diagram for nearby stars
RS, XC
NC
RS = RS CVn
binaries.
XC = X-ray
coronae
MS stars liable
to have
coronae
NC = no
coronae,
just cool
winds
Some preliminaries
In visible, UV and X-rays, nm is SI unit
of wavelength but in solar physics still
generally use wavelengths in
Ångströms: 1Å = 0.1nm
For X-rays, sometimes use energies in
keV rather than wavelengths: E(keV)
= 12.4 / λ (Å)
Spectral units
“Spectral flux” – spectral irradiance – measured in
erg cm-2 s-1 Å-1 (cgs) or W m-2 nm-1 (SI) or
W m-2 nm-1 Hz-1 (SI frequency units).
For X-rays or UV radiation, units often in photons
cm-2 s-1 Å-1 (cgs) or photons m-2 s-1nm-1 (SI)
“Spectral intensity” – spectral radiance – units:
erg cm-2 s-1 Å-1 sr-1 (cgs) or W m-2 nm-1 sr-1 (SI).
Example of a spectrum
-- but ideally spectral irradiance is in SI
units (W m-2 nm-1) and wavelength in nm.
Emission from solar and stellar
atmospheres
Chromospheres and transition regions: from
EUV (100-1000 Å) to UV (>1000Å)
Coronae: from EUV to soft X-rays (<100Å)
Active regions: from EUV to soft X-rays
Flares: from EUV to hard X-rays (<1Å)
Line emission
Line emission from abundant elements – H,
He, C, N, O, Ne, Mg, Al, Si, S, Ar, Ca, Fe.
Elements normally ionized, e.g. in stellar
coronae: C in form of C+3, C+4, C+5; Fe is
in form of Fe+9, Fe+10… Fe+16.
Often use iso-electronic series to describe
an ion, e.g. C+4 is “He-like” (2 electrons),
C+5 is “H-like” (1 electron).
Only ions with at least one electron can emit
lines – e.g. C+6 is fully stripped C, so
cannot emit lines.
Ion and spectrum notation
Atom or Ion
Neutral H
Neutral He
He+1
Neutral C
C+4 (He-like)
Fe+16 (Ne-like)
Fe+24 (He-like)
produces spectrum:
H I (first spectrum)
He I (1st spectrum)
He II (2nd spectrum)
CI
CV
Fe XVII
Fe XXV
Note: “H II regions” doesn’t make sense!
Electron configurations
Electrons in an atom have 4 quantum
numbers,
n (principle q.n. related to distance from
nucleus),
l (orbital or azimuthal q.n. related to angular
momentum and thus shape of orbit),
ml (orientation w.r.t. mag. field) of orbital
plane),
ms (orientation of electron spin).
No two electrons can have same set of 4 q.n.’s
(Pauli’s exclusion principle).
Notation for configurations
Principal quantum number n = 1, 2, 3,...
Orbital quantum number l for a given n can
have values 0, 1, ..., n - 1
Notation used is s (l =0), p (l =1), d (l =2),
f (l =3) ....
For given n, orbit with largest l is circular,
those with smaller l’’s progressively more
elliptical.
Electron configurations
H-like ion: one electron, configuration in
ground state is 1s (n=1, l=0, spin either
up or down)
He-like ion: 2 electrons, g.s. config. is 1s2
(spins up and down)
Li-like ion: 3 electrons, g.s. config. is 1s2 2s
Ne-like ion: 10 electrons, g.s. config. is
1s2 2s2 2p6 (2s e’s elliptical orbits, 2p e’s in
circular orbits)
Ar-like ion: 18 electrons, g.s. config. is
1s2 2s2 2p6 3s2 3p6
Pauli’s exclusion principle
explains “chemistry”
He, Ne, Ar, Kr have closed subshells (1s2,
2p6, 3p6 etc.) – inert (“noble”) gases
Na, K have single outer electron (3s, 4s) –
highly reactive
F, Cl have subshells “missing” an electron in
outer shell.
So Na, Cl have strong affinity for each other
– NaCl a common molecule
Pauli’s principle and the
Periodic Table
1 outer e
Filled
subshells
Atomic transitions
In a stellar atmosphere, an atom or ion is
normally in its ground state, but can be
excited (by e- collisions) to an upper level.
A radiative transition back to the ground
state or some lower state may follow,
resulting in line emission, i.e. emission of
a quantum hν
E.g. H-like ions may undergo excitation from
1s to 2p, followed by a 2p →1s transition,
resulting in a “Lyman-α” line photon
(1216Å for H, 304Å for He II etc.)
Ionization conditions
Stellar atmospheres are low-density, hot
plasmas.
Generally, for ionization, only collisional
processes important (e- = a free electron):
X+m + e-1 -> X+m+1 + e-1 + e-2
Recombination processes are either
radiative (hν = photon) :
X+m+1 + e- -> X+m +hν
or dielectronic:
X+m+1 + e- -> (X+m)** (doubly excited)
Ionization equilibrium
In the quiet solar corona, and to a 1st
approximation in active regions (or maybe even
flares), there is ionization equilibrium:
Number of ionizations/unit vol. and time = No. of
recombinations/unit vol. and time:
Ne N(X+m) Q(T) = Ne N(X+m+1) α(T)
where Q(T) = rate coefft. of ionization, α(T) = rate
coefft. of recombination.
From this one can calculate all the ion fractions for
a particular element as a function of T.
[Note: ionization equilibrium is NOT the same as
LTE! LTE holds in the solar photosphere.]
Fe ion fractions as a
function of temperature Te
9 = fractional abundance of Fe+9 ions: N(Fe+9)/N(Fe)
Line excitation from stellar
atmospheres
UV and X-ray lines excited by collisions of
e’s with ions, followed by spontaneous
radiative de-excitation:
X+m + e- -> (X+m)* + e-
(X+m)*
-> X+m + hνline
[hνline = line photon; asterix * means ion is
excited]
Line radiant fluxes
Line emission F (photons s-1) from an emitting volume V,
electron density Ne, ion density N(X+m) is:
F = Ne N (X+m) Cij(T ) V
= Ne2 V×[(N(X+m)/N(X)]×[N(X)/N(H)]×[N(H)/Ne]×Cij (T)
where Cij = collisional rate coefficient (cm3 s-1)
F = Ne2 V × f (T ) × Ab (X) × 0.8 × Cij (T )
Ne2 V is the (volume) emission measure (cm-3)
f (T ) = ionization fraction = N(X+m)/N(X)
Ab (X) = abundance of element (X) relative to H.
Line irradiance (flux) at
Earth or SOHO etc.
Line radiant flux (or power) F is number of
photons (or ergs or J) from an emitting volume
V.
At Earth, the line irradiance (or flux) is
F / (1 A.U.)2 where 1 A.U. = 1 astronomical unit
= 150 × 106 km.
Units are photons cm-2 s-1.
Note SOHO is at inner Lagrangian (L1) point which
is 148.5 × 106 km from Sun ~ 1 A.U. to ~1%
(therefore fluxes to within ~2%).
Lagrangian points: an aside!
There are 5 Lagrangian points for the
Earth’s motion round the Sun.
3 of these points are on the Sun—Earth line,
L1 (inner), L2 and L3 are beyond the
Earth and Sun.
L1 is NOT the centre-of-mass point!!
It is defined by the balance of the
centripetal acceleration (v2/r) and the net
gravitational acceln. of the Sun and Earth.
Hinode/EIS spectrum from quiet Sun
Fe XII
Fe XI
He II
CHIANTI: spectral synthesis
Some strong lines in EIS spectra
Fe IX 171.0 Å Transition 3p6 – 3p5 3d. A
“resonance” line. Near edge of EIS
channel 1, so always weak.
Fe X 174.8 Å Transition 3p5 – 3p4 3d.
Fe XI 180.4 Å Transition 3p4 – 3p3 3d.
Fe XII 195.1 Å Transition 3p3 – 3p2 3d. Main
contributor to TRACE/EIT/STEREO “195”
channel.
He II 256.3 Å Transition 1s – 3p (Ly-β),
emitted in chromosphere.
Typical X-ray stellar spectra
Sun-like stars with
hot coronae
Degree of
activity
Age
Solar flare X-ray spectrum
What do UV and X-ray spectra
tell us?
Temperatures (Te) or distribution of
material with Te differential emission
measure
Densities (Ne)
Plasma flows and turbulence
Element abundances
Temperatures from spectral
line ratios
Several temperature “diagnostics”, including
lines from different ions of same element
(e.g. Fe XXVI/Fe XXV lines in solar flare Xray spectra)
Lines of same ion with different excitation
energies (e.g. Ly-α/Ly-β of H-like ions)
For solar flares, “dielectronic satellite”
lines/nearby resonance X-ray lines often
used.
P78-1 Fe XXV solar flare spectra
Flare
peak
Flare
decay
w = 1s2 – 1s2p
j = 1s2 2p – 1s2p2
Doschek et al. (1980)
Satellite j / res.
line w = f(Te)
Electron densities from line
emission: O VII lines
OVII Spectrum line ratio
R = I(22.1Å)/I(21.8Å)
Level diagram
R = line m-1 / line 3-1
O VII ratios in a solar flare
SOLEX spectra from P76-1 spacecraft: Doschek et al. (1981)
O VII lines in AB Doradus
XMM-Newton RGS spectra of rapidly rotating K main sequence star
AB Dor (Güdel et al. 2001)
Case of Capella
Capella: spectroscopic binary - two G-type giants,
separation 157 solar radii, period ~100 days.
O VII triplet in Chandra spectra – Ne = 2 x 1010
cm-3.
EM (for both stars) is Ne2 V ~ 1052 cm-3.
Volume ~ 2.5 x 1031 cm3, x100 less than solar
corona, 105 less than the vol. of each star.
If uniform, coronae of two stars have depth ~ 30
km
Maybe coronae are made up of lots of small loops.
Capella’s “Quiescent” X-ray
Spectrum (8 – 20 Å)
Quiescent Capella spectrum
Solar flare in 1980
Capella spectrum is like a low-temperature (4-6MK) solar flare
Density-dependent line ratios
in the UV, EUV & X-ray regions
Fe+13 level diagram
4d level
59.6 Å
59.0 Å
3p levels
5303 Å
Ratio of Fe XIV
X-ray lines,
I (59.6)/I (59.0)
= function of Ne
Turbulence and flows

Turbulent events (TE) and jets observed
by HRTS in the UV from solar transition
region lines. Brueckner & Bartoe (1983)
Note: SUMER results suggest they are
the same. Velocities up to 400 km/s.
X-ray lines at the start of solar
flares: upflows and turbulence
Ca XIX lines formed at ~10 MK,
seen at the “impulsive” stage of
a solar flare in 1980 with the
BCS on SMM.
Upflow vels. ~ 200 km/s
Spectrum from
upflowing plasma
Spectrum from
stationary
plasma
Element abundances: the
“FIP” effect
In the solar atmosphere, the abundances of
Mg, Al, Si, K, Ca, Fe are enhanced relative
to the photosphere – up to x 4.
These elements all have low first ionization
potentials, FIPs (< 10 eV)
Possibly a magnetic/electric field mechanism
which takes the partially ionized material
of photosphere into the corona as it rises.
For some or even most active stars, there is
an “inverse” FIP effect (O is v. abundant).
FIP bias = coronal/photospheric
abundances vs. FIP
Element Abundances
With atomic parameters
describing the line, the
element abundance can be
deduced from line flux
measurements
K abundance from He-like K
(K XVII) line in RESIK X-ray
solar flare spectra
(Sylwester et al. 2004)
Summary
The Sun and flare (“active”) stars have hot
coronae and produce flares
Their energetic atmospheres are the source
of X-ray and UV emission
UV and X-ray spectroscopy tells us some
parameters describing their atmospheres:
temperatures, densities, flows, and
element abundances