When Space Plasmas Collide: Understanding the Interactions between the Solar Wind and

the Interstellar Medium

Jeffrey L. LinskyJILA and the Department of Astrophysical and

Planetary SciencesUniversity of Colorado

Boulder CO USA7-8 June 2007

In collaboration with:Seth Redfield (University of Texas)

Brian Wood (University of Colorado)

Seminar Outlinehttp://www.astro.uu.se/forskarutb/

long_term_V06.html• Structure of the heliosphere and local interstellar

medium• Plasma and kinetic models of the outer heliosphere• What has Voyager I told us about the termination

shock?• Observations of neutral hydrogen in the heliosphere

from Lyman-αbackscattering• The ISM: theory and reality• Measuring the properties of interstellar clouds• Constituents of the Local Bubble (gas, dust, magnetic

fields)• Astrospheres and winds of dwarf stars as a function

of activity and age

The Dynamical Structure of the Local Interstellar Medium

Seth Redfield(University of Texas at Austin)

Adapted from Mewaldt & Liewer (2001)

The Dynamical Structure of the Local Interstellar Medium

Seth Redfield(University of Texas at Austin)

Adapted from Mewaldt & Liewer (2001)

Plasma and Kinetic Models of the Outer Heliosphere: Two

Important Reviews

• B. Wood (2004) “Astrosphere and Solar-like Stellar Winds” Living Reviews in Solar Physics, Vol 1, No.2.

• G. P. Zank (1999) “Interaction of the Solar Wind with the Local Interstellar Medium: A Theoretical Perspective”, Space Science Reviews Vol. 89, 413.

Solar Wind Properties

• Mass flux ~2x10^-14 solar masses/yr with little variation.

• Fast speed wind in open field regions (~800 km/s).

• Slow speed wind in magnetically complex regions (~400 km/s).

Heliosphere/ISM Interaction Terminology

• Termination shock: where the supersonic solar wind (400-800 km/s) becomes subsonic and heated (94 AU for Voyager 1)

• Heliopause: Interface around the Sun between the subsonic solar wind and ISM plasma (~150 AU)

• Bow shock: where the incoming ISM (26 km/s) becomes subsonic (~250 AU). May not shock depends on magnetic field.

• Hydrogen wall: Pileup of neutral H gas mostly in upwind direction with charge exchange (150-250 AU)

• Plasma models: include electromagnetic and gravity forces on all ionized particles (either as one or multifluid models)

• Kinetic models: treat neutral particles (e.g. H) with long path lengths by Boltzmann equation or Monte Carlo techniques.

• Pickup ions: interstellar neutrals that are ionized (photoionization or charge exchange) and captured by the solar wind magnetic field and accelerated at the termination shock.

Heliosphere/ISM interaction: Examples of Code Types

• Monte Carlo codes beginning with Baranov & Malama (1993, 1995).

• Hydrodynamic four-fluid codes beginning with Zank et al. (1996) (one fluid for protons, 3 fluids for hydrogen:primary, secondary, tertiary atoms)

• Hybrid codes beginning with Muller et al. (2000).

• MHD codes (cf. Opher 2004).

Müller (2004)

Interaction of Stars with their LISMHeliosphere is the structure caused by the momentum balance (v2) between the outward moving solar wind and the surrounding interstellar medium.

Magnetized solar wind extends out to heliopause, diverts plasma around Solar System, and modulates the cosmic ray flux into Solar System.

Most neutrals stream in unperturbed, except neutral hydrogen, which due to charge exchange reactions, is heated and decelerated forming “Hydrogen Wall” (log NH (cm-2) ~ 14.5).

Reviews of heliospheric modeling: Wood (2004), Zank (1999), and Baranov (1990).

Reviews of interaction of LISM with heliosphere: “Solar Journey” Frisch (2006), and Redfield (2006)

Zank & Frisch (1999)

Increase the density of the surrounding LISM by only a factor of 50 (nH from 0.2 to 10 cm-3) and the termination shock shrinks from 100 AU to 10 AU.

How large of a density increase is needed to significantly alter the structure of the

heliosphere?NH(LISM) = 10 cm-3

NH(LISM) = 0.2 cm-3

Müller (2004)

Consequences of a compressed heliosphere:

(1) Cosmic rays flux modulated by magnetized solar wind:

Cloud nucleation, increase in planetary albedo (Marsh & Svensmark 2000, Carslaw et al. 2002)

Lightning production increase (Gurevich & Zybin 2005)

CR alters ozone layer chemistry (Randall et al. 2005)

CR source DNA mutation (Reedy et al. 1983)

(2) Direct deposition of ISM material onto planetary atmosphere:

Dust deposition could trigger “snowball” Earth episode (Pavlov et al. 2005)

Create mesospheric ice clouds, increase in planetary albedo (McKay & Thomas 1978)

What has Voyager 1 told us about the Termination Shock?

• Voyager 1 was launched 5 Sept 1977, passed Jupiter and Saturn and crossed the termination shock (TS) on 16 Dec 2004 at 94 AU from the Sun.

• Onboard detectors measured the magnetic field and energies and directions of energetic electrons, protons, and He nuclei.

• Weak shock: velocity jump (r=2.6).• TS not spherical due to magnetic field.• TS moves in and out with the solar cycle.• Solar wind in heliosheath slower, hotter, and denser than inside

the TS.• The TS was expected to be the location where anomalous

cosmic rays (ACRs) are accelerated, but peak in ACR flux not at the TS.

Stone et al. (2005) Science 309, 2017

• A: Energetic proton intensity streaming from the Sun.

• B: Termination Shock Particle (TSP) streaming anisotropy decreases when cross the TS due to scattering by magnetic turbulence.

• C: Intensities of high-energy protons and electrons.

• D: Intensities of Galactic cosmic rays and He ions (anomalous cosmic rays).

Change in Magnetic Field Strength and Direction (heliographic) at the TS: Burlaga

(2005) Science 309, 2027

Change in Magnetic Field at the TS: Burlaga et al. (2005)

• Magnetic compression ratio across TS: 3+/-1.

• Complete change in average field strength perhaps due to thermalization of the plasma by the TS.

• Inward moving TS (>90 km/s) may have crossed Voyager 1: Jokipii (2005) ApJ 631, L163.

Effect of the Interstellar Magnetic Field on the Heliosphere (Opher et al. 2006, 2007)• 3-D MHD adaptive grid models with

interstellar magnetic field (B~1.8μG) α=30-60 degrees from inflow direction. (60-90 degrees from Galactic plane).

• vSW=450 km/s, vism=25.5 km/s Parker spiral B=2μG at equator

• Produces a N-S asymmetry in the TS and heliosheath and a deflection in the current sheet.

• TS moved in 2.0 AU for Voyager 1 with TSP streaming outward from Sun along a spiral field line that does not first go through the TS.

• Moves TS inward more in S (Voyager 2) than N (Voyager 1).

• Where Voyager 2 crosses the TS will be the critical test of value of α.

Observations of Neutral Hydrogen in the Heliosphere from Lyman-α Backscattering

Measurements: References

• Bertaux et al. (1995) Solar Phys. 162, 403 describes the SWAN (Solar Wind Anisotropies) experiment on SOHO (Solar and Heliospheric Observatory) satellite.

• Lallement et al. (2005) Science 307, 1447.

• Quémerais et al. (2006) A+A 455, 1135

SWAN maps Backscattered Lyman-αRadiation from Neutral Hydrogen Flowing into the Heliosphere using a

Hydrogen Gas Absorption Cell Spectrometer

• Maximum in the Lyman-αsky glow when inflowing H has maximum Doppler shift relative to SOHO.

• Maps show Lyman-αglow at different times of year when SOHO has different velocities relative to the H inflow vector.

Deflection of Neutral H vs. He due to charge-exchanged Solar Wind Protons becoming H Secondaries

• Incoming He atoms retain the ISM inflow direction.

• H secondaries will be deviated if the interstellar magnetic field is inclined relative to the gas flow direction.

• Measured deviation is 4+/-1 degrees.

Inferring the Direction of the Local Interstellar Magnetic Field

• MHD models consistent with the deviation of H relative to He predict the magnetic field direction in Galactic coordinates (30-60 degrees from Galactic plane).

• Magnetic field is parallel to the edges of the LIC and G clouds and likely compressed by the relative motion of the two clouds.

Measuring the Properties of Interstellar Clouds: References

• Linsky et al. (2000) ApJ 528, 756

• Redfield & Linsky (2000) ApJ 534, 825

• Redfield & Linsky (2002) ApJS 139, 439

• Redfield & Linsky (2004) ApJ 602, 776

• Redfield & Linsky (2004) ApJ 613, 1004

• Redfield & Linsky (2007) ApJ, in press

Observational Diagnostics

1 ion, 1 sightline Velocity, Column Density

Redfield & Linsky (2004a)

Observational Diagnostics

1 ion, 1 sightline Velocity, Column Density

multiple ions, 1 sightline

Temperature, Turbulence, Volume Density, Abundances, Depletion, Ionization Fraction

Redfield & Linsky (2004b)

Observational Diagnostics

1 ion, 1 sightline Velocity, Column Density

multiple ions, 1 sightline

multiple ions, multiple sightlines

Temperature, Turbulence, Volume Density, Abundances, Depletion, Ionization Fraction

Global Morphology, Global Kinematics, Intercloud Variation

Origin and Evolution, Interaction of LISM Phases

Global Kinematics

Our “observables”:

Centroid velocity (vR) of LISM absorption, i.e., the radial component of the projected velocity in direction (l, b)

Assume the simplest dynamical structure: a single vector bulk flow.

vR = v0 (cos b cos b0 cos (l0 - l) + sin b0 sin b)

Questions:

Can the observed LISM velocities be characterized by a rigid bulk flow, or are they chaotic?

How significant are any departures from a rigid flow vector?

Is there a correlation between direction and departure from the bulk flow?

Can multiple velocity vectors successfully characterize the majority of LISM observations?

160 targets within 100 pc observed at high and moderate resolution that contain 270 LISM absorption components (~60% of UV observations taken for other purposes).

LISM Sample

LISM Sample

Dynamical Modeling Procedure

Iterative process:Fit datasetRemove outliers (in velocity and space)Repeat until satisfactory fit (use F-test to determine stopping point)Remove successful dynamical cloud sightlines from datasetStart over.

Assumptions:All motions are rigid velocity vectors (i.e., a tensor dynamical characterization may result

in two separate dynamical clouds, close in space, and with similar velocity vectors).The projected morphology of all clouds are contiguous (i.e., no “swiss cheese” clouds)

Results: Fits largest surface area structures first (e.g., LIC and G vectors derived immediately).15 velocity vectors satisfy 80% of LISM database.All velocity vectors are similar, and approximately opposite to the motion of the Sun.About 1/3rd of the dynamical clouds are filamentary.Search for unique cloud properties.

Heliocentric Vector Solutions

Vector Distribution:

All LISM vectors are coming from a narrow range of directionsA relatively wide range of velocity magnitudes, with mean around 20 km/s relative to LSRThe solar trajectory (upstream: l ~ 208, b ~ -32, v ~ -13 km/s) is in almost the exact

opposite direction (Dehnen & Binney 1998)Spatial correlation of velocity vectors are clear (e.g., Aql Cloud)Many dynamical clouds have filamentary structures possibly due to collisional interfaces

Upwind Velocity Distribution Relative to Sun:

Upwind Velocity Distribution Relative to LSR:

Do individual dynamical clouds have unique physical properties?

Is the Heliosphere located inside of the LIC or the G Cloud?

Location V0 (km/s) T(K)

LIC 23.88±0.91 7500±1300

He inside the heliosphere

26.24±0.45 6300±390

G Cloud 29.6±1.1 5500±400

Macroscopic Velocity Differences

All V magnitudes for clouds that share the same line of sight.

Without distance information, it is unclear how many velocity differences are realized.

Indicate that macroscopic motions can induce compression and shear flows.

Transonic turbulent compression by warm interstellar clouds can successfully create cold neutral clouds (Vazquez-Semadeni et al. 2006).

LIC/G

1% of all radio quasars show short timescale variability (hours).

Diffraction of intervening screen can cause time delays in signal - use to get transverse motion of scintillating screen.

As Earth moves through projection of diffraction pattern, the characteristic timescale of scintillation varies - which is repeatable over several years.

Set characteristic scale length (~104 km) to Fresnel scale to put limits on screen distance: L < 10 pc!

Radio Scintillation: IntraDay Variables (IDVs)

Dennett-Thorpe & de Bruyn (2002)Bignall et al. (2006)

Annual Scintillation Signature

Annual variation of scintillation timescale is fit with 5 free parameters: diffraction scale length, axial ratio, orientation, and transverse velocity.

For B1257-326, our derived transverse velocity of the Aur Cloud, which lies along the line of sight toward the quasar B1257-326, is consistent with the radio scintillation observations.!

Bignall et al. (2006)

Future

Past

Radio Scintillation Sources

Leo Cold Cloud (Meyer et al. 2006)

Astrospheres: References

• Wood et al. (2001) ApJ 547, L49 (αCen and Prox Cen).

• Wood et al. (2002) ApJ 574, 412 (4 new stars

• Wood et al. (2005) ApJ 628, L143 (14 stars).

• Frisch (1993) ApJ 407, 198 (first estimate of astropause radii for stars).

Wood et al. (2002)

Journey of a Lyman-α Photon

Detailed central profile of late-type star has little influence since at the core of absorption.

log NH(LISM) < 18.7, otherwise obliterates any helio- or astrospheric signature.

Nearby DI line critical to constraining fit of LISM absorption (and constancy of D/H in Local Bubble).

Since HI is decelerated at heliosphere, the heliospheric absorption is always redshifted and the astrospheric absorption is always blueshifted.

Comparison of Lyman-α Profiles with hydrodynamic models assuming wind

speed of 400 km/s and LISM parameters

Maps of Neutral Hydrogen Density with Sightline to Sun

Mass Loss rates vs. X-ray Surface Flux and Age of the Sun

Consequences of much higher mass loss rates and UV/X-ray fluxes from young solar-like stars

• Ionization of the outer parts of young disks → chemistry, large molecule formation.

• Erosion of planetary atmospheres when no planetary magnetic field (e.g., Mars beginning 3.9 Gyr ago.

• Evaporation of atmospheres of close in giant planets.• Changes in asteroid minerology (melting by giant

flares• Little change to stellar mass but large change in

stellar angular momentum.

Classical Theoretical Models of the ISMReferences Assumptions

• Field, Goldsmith, Habing (1969) ApJ 155, L149 (CNM,WNM/WIM)

• McKee, Ostriker (1977) ApJ 218, 148 (CNM, WNM/WIM,HIM)

• Wolfire et al.(2003) ApJ 587,278 (3-phase)

• Ferrière (2001) Rev.Mod.Phys. 73(4)1031 (ISM data)

• Cox(2005) ARAA 43, 337 (important review)

• Hydrostatic equilibrium• Thermal equalibrium• Heating by UV photoelectric

(on gas and grains)(depends on SN rate)

• Cooling (radiation by forbidden lines, PAHs)

• Thermally unstable P(n) curve allows coexistence of 3 phases at n=0.3 to 30/cm³, P/k=1700-4400, T=270-5500K)

• Nonthermal pressure terms not included

Constituents of the Local Bubble: Pressure Components (K/cm³)

Weight of overlying gas: Pmid/k ≈ 22,000Cosmic rays: Pcr/k ≈ 9300Magnetic pressure: Pmag/k = B²/8πk = 7200 (B/5μG)² Ram pressure: Pram/k = ρv²/k = 2400(ntot/0.3)

(v/8 km/s)²Thermal pressure: Pth/k = ntotT=2300[Pmag=Pth when Beq=2.8μG][Pmag+Pcr estimated from observed

synchrotron emission]

Thermal Pressure Disequilibrium

• Theoretical two-phase equilibrium ISM model (Wolfire et al. 2003; Cox 2005).

• Blue dotted curve is 10X higher heating rate.

• Black dashed line is total midplane pressure due to overlying matter.

• Orange dashed line is mean magnetic pressure.

• Nonthermal pressure (dynamic, magnetic, cosmic ray) dominates the thermal pressure. Therefore, a wide range of pressures (1700-20,000 can be compensated by different nonthermal terms.

• Is this the whole story? Dynamic pressures can be measured by high-resolution UV spectroscopy.

Measurements of thermal gas pressures

• In LIC, P/k = 2300 from spectral line widths and gas densities.

• Jenkins & Tripp (2001) study of C I fine structure excitation (STIS 1.5 km/s spectra).

• Mean thermal P/k=2240.• 15% of gas at P/k>5,000• A very small amount of

gas at P/k>100,000. Turbulent compression? Very small sizes (0.01 pc)?

C I fine structure line spectra obtained by Jenkins & Tripp (2001) with STIS at 1.5 km/s resolution

Interstellar Radiation Environment near the Sun: Slavin & Frisch (2002) ApJ 565,

364Stars: ε CMa (B2Iab, 132 pc)

and hot white dwarfs dominate (EUVE data).

Hot gas: Local Bubble (if T=1.25 MK).

Cloud Boundary: evaporative interface (O VI etc.)

FUV background: Beyond the Local Bubble.

Absorption: log N(HI)=4x10^17 cm^-2 (outer part of LIC).

Radiation hardens to center of warm clouds.

Higher ionization of He than H.

Lallement, Welsh, et al. (2003)

Local Bubble

R ~ 100 pc; ne ~ 10-3 cm-3; T ~ 106 K

NaI spectroscopy (Lallement, Welsh, et al. 2003)Soft X-rays ROSAT (Snowden et al. 1998)OVI absorption lines FUSE (Oegerle et al. 2005)OVII, OVIII emission Chandra (Smith et al. 2005)EUV emission lines CHIPS (Hurwitz et al. 2005)

LISM

R ~ 1-10 pc; n ~ 0.2 cm-3; T ~ 7000 K

Lyman-α backscatter SOHO (Quémerais et al. 2000)He I emission Ulysses (in situ) (Witte et al. 1996)UV/Optical spectroscopy (Redfield & Linsky 2002, 2004, Welty, Hobbs, Morton 2003, Crawford 2001)

Cold Dense Gas?

R~1.4 pc; nH ~ 30 cm-3; T ~ 20 K

NaI spectroscopy (Lallement, Welsh, et al. 2003)NaI + HI (Meyer et al. 2006; Heiles & Troland 2003)CO (Magnani et al. 1996)Cores (Chol Minh et al. 2003)

How much hot gas in the Local Bubble? (Hurwitz et al. ApJ 623, 911 (2005))

• Observation with the Cosmic Hot Interstellar Plasma Spectrometer (CHIPS) satellite.

• Diffuse emission (revised to an upper limit measured in the Fe IX 171Å line (formed at log T = 5.4-6.4).

• Emission measure (ne^2 length) upper limit depends on T and Fe depletion and assumption of collisional ionization equilibrium.

• Significant foreground component (large in daylight) not removed!

Possible hot gas models

• If ne = 0.1 cm^{-3} like in warm clouds, then EMLB= 1 cm^{-6}pc (>1000 times too large).

• If thermal gas pressure equilibrium P/k=2280 Kcm^{-3}, and log T=6.0, then EMLB=0.001 (too large).

• If thermal gas pressure equilibrium and logT as low as 5.6, then EMLB=0.001 (but does not account for significant foreground emission).

• Some semi-hot gas could fill most of volume of the Local Bubble consistent with Fe IX flux upper limit..

• But, there are other pressure terms and the LB could be far from thermal gas pressure equilibrium.

Could the soft X-ray background be foreground? (Koutroumpa et al. A+A 460, 289 (2006))

• Calculation of EUV/soft X-ray emission from charge transfer between solar wind heavy ions and interstellar neutral atoms.

• Solid lines are: N ecliptic pole, South ecliptic pole, equitorial antisolar. Solid: solar min. Dashed: solar max.

• Important emission ions: C VI, O VII, O VIII, Ne IX, and Mg XI.

• A large fraction of the EUV and soft X-ray background is heliospheric foreground.

• Earlier studies by Cravens and Lallement with similar conclusion.

• No O VI (T=≈300,000K) in LB.• Is there any hot gas in the Local

Bubble? Is most of the Local Bubble volume empty?

Comparison of simulated soft X-ray flux calculated from charge exchange of the solar wind with interplanetary

neutral H and XMM observations of diffuse soft X-rays