4.2. Zone of Influence of Winds
The impact of galactic winds on the host galaxies and the environment depends sensitively on the size of the "zone of influence" of these winds, i.e. the region affected either directly (e.g., heating, metals) or indirectly (e.g., ionizing radiation) by these winds. This section summarizes the methods used to estimate this quantity.
Indirect Measurements based on Estimates of the Escape
Velocity. The true extent of galactic winds is difficult to
determine in practice due to the steeply declining density profile of
both the wind material and the host ISM. The zone of influence of
galactic winds is therefore often estimated using indirect means which
rely on a number of assumptions. A popular method is to use the
measured velocity of the outflow and compare it with the local escape
velocity derived from some model for the gravitational potential of
the host galaxy. If the measured outflow velocity exceeds the
predicted escape velocity and if the halo drag is negligible, then
the outflowing material is presumed to escape the host galaxy and be
deposited in the IGM on scales
50 - 100
kpc. This method was used for instance by
Rupke et al. (2002)
to estimate the average escape fraction <fesc>
esc /
tot and
"ejection efficiency"
<
>
esc /
SFRi for 12
ULIRGs, which were found to be ~ 0.4 - 0.5 and ~ 0.1, respectively. These
calculations assumed that the host galaxy could be modeled as a singular
isothermal sphere truncated at some radius
rmax. Neither the escape fraction nor the ejection
efficiency were found to be sensitive to the exact value of
rmax. Other strong cases for escaping
material include the "H
cap" of M 82 and the ~ 1500 km
s-1 line-emitting material in the superbubble of NGC 3079; both
objects were discussed in Section 3.2.
Note that the outflow velocities measured by Rupke et al. (2002) refer to the neutral component of the outflow, not the hot enriched wind fluid. Unfortunately, direct measurements of the wind velocity are not yet technically possible so one generally relies on the expected terminal velocity of an adiabatic wind at the measured X-ray temperature TX [vX ~ (5KTX / µ)0.5, where µ is the mean mass per particle] to provide a lower limit to the velocity of the wind fluid (this is a lower limit because it only takes into account the thermal energy of this gas and neglects any bulk motion; e.g., Chevalier & Clegg 1985; Martin 1999; Heckman et al. 2000).
Arguably the single most important assumption made to determine the fate of the outflowing gas is that halo drag is negligible. Silich & Tenorio-Tagle (2001) have argued that halo drag may severely limit the extent of the wind and the escape fraction. Drag by a dense halo or a complex of tidal debris may be particularly important in ULIRGs if they are created by galaxy interactions (e.g., Veilleux, Kim, & Sanders 2002b).
Deep X-ray and Optical Maps of Local Starbursts. The fundamental limitation in directly measuring the zone of influence of winds is the sensitivity of the instruments. Fortunately, CXO and XMM-Newton now provide powerful tools to better constrain the extent of the hot medium (e.g., M 82, Stevens et al. 2003; NGC 3079, Cecil et al. 2002; NGC 6240, Komossa et al. 2003; Veilleux et al. 2003; NGC 1511, Dahlem et 2003). The reader should refer to the contribution of M. Ehle at this conference for a summary of recent X-ray results (see also Strickland et al. 2003 and references therein).
The present discussion focusses on optical constraints derived from
the detection of warm ionized gas on the outskirts of wind
hosts. Progress in this area of research has been possible thanks to
advances in the fabrication of low-order Fabry-Perot etalons which are
used as tunable filters to provide monochromatic images over a large
fraction of the field of view of the imager. The central wavelength
(3500 Å - 1.0 µm) is tuned to the emission-line feature of
interest and the bandwidth (10 - 100 Å) is chosen to minimize the
sky background. Continuum and emission-line images are produced nearly
simultaneously thanks to a "charge shuffling/frequency switching"
mode, where the charges are moved up and down within the detector at
the same time as switching between two discrete frequencies with the
tunable filter, therefore averaging out temporal variations associated
with atmospheric lines and transparency, seeing, instrument and
detector instabilities. The narrow-band images are obtained in a
straddle mode, where the off-band image is made up of a pair of images
that "straddle" the on-band image in wavelength (e.g.,
1 =
6500 Å and
2 = 6625
Å for rest-frame
H); this
greatly improves the accuracy of the continuum removal since it
corrects for slopes in the continuum and underlying absorption features.
These techniques have been used with the Taurus Tunable Filter (TTF; Bland-Hawthorn & Jones 1998; Bland-Hawthorn & Kedziora-Chudczer 2003) on the AAT and WHT to produce emission-line images of several "quiescent" disk galaxies (Miller & Veilleux 2003a) and a few starburst galaxies (Veilleux et al. 2003) down to unprecedented emission-line fluxes. Gaseous complexes or filaments larger than ~ 20 kpc have been discovered or confirmed in a number of wind hosts (e.g., NGC 1482 and NGC 6240; the presence of warm ionized gas at ~ 12 kpc from the center of M 82 was discussed in Section 3.2). Multi-line imaging and long-slit spectroscopy of the gas found on large scale reveal line ratios which are generally not H II region-like. Shocks often contribute significantly to the ionization of the outflowing gas on the outskirts of starburst galaxies. As expected from shock models (e.g., Dopita & Sutherland 1995), the importance of shocks over photoionization by OB stars appears to scale with the velocity of the outflowing gas (e.g., NGC 1482, NGC 6240, or ESO484-G036 versus NGC 1705; NGC 3079 is an extreme example of a shock-excited wind nebula; Veilleux et al. 1994), although other factors like the starburst age, star formation rate, and the dynamical state of the outflowing structure (e.g., pre- or post-blowout) must also be important in determining the excitation properties of the gas at these large radii (e.g., Shopbell & Bland-Hawthorn 1998 and Veilleux & Rupke 2002).
Influence of the Wind on Companion Galaxies. Companion galaxies located within the zone of influence of the wind will be affected by the wind ram pressure. Irwin et al. (1987) noticed that the dwarf S0 galaxy NGC 3073 exhibits an elongated H I tail that is remarkably aligned with the nucleus of NGC 3079. Irwin et al. have argued that ram pressure due to the outflowing gas of NGC 3079 is responsible for this tail. If that is the case, the wind of NGC 3079 must extent to at least ~ 50 kpc. This is the only system known so far where this phenomenon is suspected to take place.
Absorption-Line Studies. Absorption-line spectroscopy of bright background galaxies (e.g., high-z quasars, Lyman break galaxies) can provide direct constraints on the zone of influence of galactic winds. Norman et al. (1996) have used this method to estimate the extent of the wind in NGC 520. A strong and possibly complex Mg II, Mg I, and Fe II absorption-line system was found near the systemic velocity of NGC 520 at a distance from the galactic nucleus of 24 h-1 kpc. A weaker system at a distance of 52 h-1 kpc is also possibly present. Unfortunately, NGC 520 is undergoing a tidal interaction so the absorption may arise from tidally disrupted gas rather than material in the purported wind. Norman et al. also looked for absorption-line systems associated with the wind of NGC 253, but the proximity of this system to our own Galaxy and to the line of sight to the Magellanic Stream makes the identification of the absorption-line systems ambiguous. No other local wind galaxy has been studied using this technique.
Large absorption-line data sets collected on high-z galaxies provide new constraints on the zone of influence of winds in the early universe. Adelberger et al. (2003) have recently presented tantalizing evidence for a deficit of neutral hydrogen clouds within a comoving radius of ~ 0.5 h-1 Mpc from z ~ 3 LBGs. The uncertainties are large and the results are significant at less than the ~ 2 level. Adelberger et al. (2003) argue that this deficit, if real, is unlikely to be due solely to the ionizing radiation from LBGs (e.g., Steidel et al. 2001; Giallongo et al. 2002). They favor a scenario in which the winds in LBGs directly influence the surrounding IGM. They also argue that the excess of absorption-line systems with large CIV column densities near LBGs is evidence for chemical enrichment of the IGM by the LBG winds.