In some cases these winds are indeed driven by supernovae and massive stellar winds from a central starburst (e.g., M82, NGC 1569; [Lehnert & Heckman 1995]; [Lehnert & Heckman 1996]), while other winds appear to be powered by more exotic forces associated with the central engines of AGN (e.g., NGC 3079, NGC 4051; [Colbert et al. 1996a]; [Colbert et al. 1996b]), and some galaxies appear to exhibit both starburst and AGN characteristics (e.g., NGC 1808 [[Forbes, Boisson, & Ward 1992]], Mk 231 [[Lípari, Colina, & Macchetto 1994]]). Because of the low densities of the wind material, emission from optical lines is often very difficult to detect, except in nearby galaxies. The x-ray emission from these winds is comparable in luminosity to the optical emission lines, but is visible on larger spatial scales and therefore detectable to greater distances.
|Designations||M82, NGC 3034, UGC 5322, ARP 337, 3C 231|
|Position (2000.0)||09.^h 55.^m 54.^s0; 69o 40' 57''|
|Galaxy type||I0, Irr II|
|Dimensions||11.'2 4.'3 (10.6 4.1 kpc)|
|Disk inclination||81.o5 a|
|Radial velocity||km s-1|
|Total magnitude||B =|
|Galactic extinction||A = 0.13|
|Total colors||B - V =|
|U - B =|
|Radio luminosity||L = 1039 ergs s-1 c|
|Infrared luminosity||L = 1044 ergs s-1 c|
|H luminosity||L = ergs s-1 d|
|X-ray luminosity||L = ergs s-1 c|
Due to its proximity and favorable inclination angle (see Tab. 1), the irregular disk galaxy M82 (NGC 3034) has been studied extensively as a prototype galactic wind system. Following the original discovery ([Lynds & Sandage 1963]), the first detailed spectroscopic study of the optical emission line filaments ([Burbidge, Burbidge, & Rubin 1964]) revealed kinematics indicative of a bipolar, roughly conical, outflow of gas along the minor axis of the galaxy.
Thirty years and over 500 published papers later, the initial interpretation of the M82 filaments in terms of an outflow still stands. Support for this picture extends from radio to gamma ray wavelengths. The starburst nature of the nucleus of M82 has been verified through its strong infrared emission (e.g., [Soifer, Houck, & Neugebauer 1987]; [Rice et al. 1988]; [Telesco et al. 1991]) and numerous compact radio supernovae (e.g., [Kronberg, Biermann, & Schwab 1985]; [Huang et al. 1994]; [Muxlow et al. 1994]). Models of the starburst evolution produce appropriate quantities of energy and mass on plausible timescales to create and sustain the observed nuclear and galactic wind behavior (e.g., [Rieke et al. 1993]; [Doane & Mathews 1993]). The wind itself has been observed optically in both spectral (e.g., [McCarthy, Heckman, & van Breugel 1987]; [McKeith et al. 1995]) and imaging studies (e.g., [Ichikawa et al. 1994]), in emission lines (e.g., [Armus, Heckman, & Miley 1990]) and broadband radiation (e.g., [O'Connell & Mangano 1978]). In particular, kinematic evidence for the existence of a galactic wind in M82 has been presented by several optical emission line studies (e.g., [Heckathorn 1972]; [Axon & Taylor 1978]; [Bland & Tully 1988]; [Heckman, Armus, & Miley 1990]; [McKeith et al. 1995]), and even by molecular observations ([Nakai et al. 1987]).
An extensive x-ray halo has been observed oriented along the minor axis, extending 5-6 kpc from the disk, far beyond the visible extent of the optical filaments (e.g., [Watson, Stanger, & Griffiths 1984]; [Schaaf et al. 1989]; [Bregman, Schulman, & Tomisaka 1995]). A few percent of the supernova energy from the starburst has been deposited in this hot ( K) ``x-ray wind.'' A large ( kpc) spherical halo has been reported at radio continuum wavelengths ([Seaquist & Odegard 1991]), and has been interpreted as synchrotron emission from relativistic electrons in the outflowing wind.
With this wealth of data, it is surprising that little attempt has been made to undertake detailed comparisons of models and observations. Large-scale galactic winds were first proposed to account for the lack of gas in elliptical galaxies ([Mathews & Baker 1971]). The theory of these winds has since evolved in tandem with research on starburst-driven galactic winds (e.g., [White & Chevalier 1983]; [Smith, Kennel, & Coroniti 1993a]; [Smith, Kennel, & Coroniti 1993b]). Since the original starburst wind model ([Chevalier & Clegg 1985]), advances have been made in both analytical studies (e.g., [Koo & McKee 1992a]; [Koo & McKee 1992b]) and hydrodynamic simulations (e.g., [Tomisaka & Ikeuchi 1988]; [Tomisaka & Bregman 1993]; [Suchkov et al. 1994]; [Suchkov et al. 1996]). Models of winds in other astrophysical situations, such as stellar winds (e.g., [Shu et al. 1991]) and winds from AGN (e.g., [Smith 1993]; [Arav & Li 1994]; [Arav, Li, & Begelman 1994]), have contributed to our understanding as well. On a broader scale, starburst-driven winds have important ramifications for a variety of astrophysical and cosmological situations, including enrichment of the intergalactic medium, contributions to the diffuse X-ray background, the evolution of dwarf and interacting galaxies, and the formation of elliptical galaxies through mergers (see [Heckman, Armus, & Miley 1990] and references therein).
Toward the goal of investigating this model, we have obtained high-resolution imaging Fabry-Perot observations of M82 in several optical emission lines. In Section 2 of this paper we present our Fabry-Perot observations and describe the methods of data reduction. In Section 3, we present the derived two-dimensional emission-line maps of M82 illustrating the distribution of line flux, ionized gas velocity, and ionization state across the outflow. In Section 4, we describe our new data for the disk, halo, and outflow and provide comparisons with other published observations and a number of kinematic models. We present our conclusions in Section 5. A subsequent paper describes refinements to our kinematic and ionization models to accommodate new observations from the Keck and Hubble Space Telescopes.