Next Contents Previous

4.2. Structure of the PC images of 53W002

Here we discuss the components of this young galaxy revealed in the PC images:

4.2.1. The maximum AGN contribution

To constrain the maximum possible AGN contribution to 53W002's continuum, a scaled PSF was subtracted from its image core, using the faint red star "S", noted by W92. The maximum possible point source that can be subtracted from 53W002's core without making its central flux negative is 25 ± 2% of the total light in B, 21 ± 3% in V, and 20 ± 2% in I. The upper error boundaries of these fractions are firm, so that the AGN contribution to 53W002's total continuum is definitely leq 30% (W91, W92), but the lower boundaries are soft. The AGN contribution could be ltapprox 15% of its total continuum, which would in Fig. 6b produce light-profiles that are straighter in r1/4-space. This light is contained within 0".06 FWHM or ~ 500 pc and has blue colors (B - I appeq 0.06 mag), as expected for an AGN at z appeq 2.4. The moderately weak radio source 53W002 (S1.4 = 50 mJy, logP1.4 = 27.5 W Hz-1) thus has a similarly modest optical-UV AGN contribution: VAGN appeq 24.3, MAGNV appeq -21.8 (power-law K-correction with alpha appeq 1.0; W91).

Figure 8

Figure 8. Grey scale images of 66 x 66 pixel sections (appeq 3".0 x 3".0) of the PC-exposures on 53W002 (see W98): (a) 15 x 2700s orbits in Lyalpha410; (b) 24 x 2400s orbits in B450; (c) 12 x 1700s orbits in V606; (d) 12 x 1700s orbits in I814; (e) equal to (b) after subtraction of a central point source (with 25% of the total flux) and of the best fit r1/4-profile (Fig. 6b); (f) the (B - I) color image [= (b)/(d)]. The VLA 8.4 GHz contours of W91 are superimposed and the CO peaks of S97 are indicated in (a) & (d), as explained in W98. The AGN is located at each panel's central pixel. Note: the low SB and "arclike" feature of the Lyalpha cloud in (a); the dominant r1/4-like profile and brighter one-sided blue cloud in (b)-(d); the brighter and fainter blue clouds in (e); and the blue (dark) central AGN and the reddish (white) linear feature (a "dust lane"?) separating both blue clouds in (f). The radio source covers most of the blue clouds plus the Lyalpha "arc", suggesting jet-induced starformation; the remainder of the Western continuum and Lyalpha cloud may be AGN light seen in reflection.

4.2.2. The inner resolved continuum "core"

To address the symmetric extended component, only the "clean" quadrant between the blue clouds discussed in Section 4.2.4 are fit (see Fig. 8a-8f). The central parts of the galaxy in F450W cannot be simultaneously fit by a single r1/4 or exponential law. Comparing these profile-fits to the PSF shows that a small additional central light distribution (with rhl ~ 0".05) is required, containing about half the flux of the nuclear point source, but with redder colors ((B - I) appeq 0.76 ± 0.1). The distinctions among this small extended component, the nuclear point-source (4.2.1), and the more extended r1/4-like distribution (Section 4.2.3) are artificial to some extent, reflecting a radial structure more complex than our simple library of pieces (see Section 4.2.4).

4.2.3. The remaining r1/4-like profile

After subtraction of the central unresolved AGN component (Section 4.2.1), the underlying galaxy can be measured only in the quadrant between the two blue clouds (Section 4.2.4). W98 concentrate on a region south of the nucleus, where nearly a full 90° quadrant is clear of these contaminating sources. Details of the elliptical profile fitting technique and its errors are given by W92, W94b, Mu97, & Sc97. The BVI light-profiles are shown in Fig. 6b. Most of the deviation from an r1/4-law at r1/4 gtapprox 0.75 (r gtapprox 0".32) occurs in B & V, and is due to the faint blue cloud leaking into the uncontaminated quadrant, and not only due to sky-subtraction errors, which affect the profile for SBBVI gtapprox 25.5-26.0 mag arcsec-2 (see Fig. 6b). The BVI light profiles (Fig. 6b) follow an r1/4-like profile closer than an exponential disk, although an early-type galaxy with a bulge-to-disk ratio gtapprox 3-5 cannot be ruled out from the PC data. The best r1/4 fit has (a/b) = 1.25 ± 0.1 and rhl appeq 0".20 ± 07 in B & V and rhl appeq 0.27 ± 0.05 in I (or 1.8-2.5 kpc). The total flux of this symmetric component is well constrained at 27 ± 3% of the total. Its position angle (PA ~ 110°) is uncertain, but consistent with the orientation of the aligned clouds (PA ~ 95°), which appear to be separated by a redder feature (Section 4.2.4 & Fig. 8f). 53W002's average color is (B - I) appeq 1.3. Its (V - I) color (appeq 0.70 mag) is less contaminated by the blue cloud. Within the errors, both colors are relatively constant with radius, so that any color gradient must be small for r ltapprox 1".0 (ltapprox 0.3 mag across the PC image). W91 & W94b present 12-band (Lyalpha UBVRIgriJHK) photometry for 53W002 and surrounding objects, and W94b, Mu97, & Sc97 discuss spectral evolution model fits to these and similar color-redshift data to constrain stellar population ages (defined as the onset of the major, currently visible starburst). Fig. 8f suggests that, with the exception of the region possibly affected by a "dust lane", 53W002 is not enormously reddened by dust (Section 4.2.4). Following these spectral evolution models, the colors of the symmetric component of 53W002 - if interpreted as coming from stars only - would suggest a stellar population with an average age of ~ 0.4 Gyr (W91, W94b), which is of the same order as its dynamical time scale (Section 4.3). The lack of discernable color gradient does not allow us to distinguish whether 53W002 formed through a sudden, global halo collapse (cf. ELS62) or through rapid merging of many sub-galactic sized units (e.g., SZ78, P96b). Any color gradient between the current PC and NICMOS images would help decide between these scenarios. Below we show that there may be other processes triggering the (star)formation of 53W002.

4.2.4. The nature of the blue clouds

The residuals after removing the symmetric pieces described in Section 4.2.1 - 4.2.3 trace the aligned blue "clouds" nicely. The larger continuum cloud to the west - in the direction of the extended Lyalpha distribution (Fig. 8a & W91) - is quite extended and vaguely triangular (see Fig. 8e and the high contrast insert in the color Plate of W98). It peaks 0".45 west of the nucleus and extends gtapprox 1" from the core with an opening angle of about 45° (Fig. 8b), and has a brighter "arc" which is dominated by Lyalpha emission (Fig. 8a) about 0".6 from the core. On the opposite side is a very small blue object - possibly a "counter-cloud" - elongated perpendicular to the nucleus-cloud direction and confined within 0".2 from the core (Fig. 8e) - at the very limit of the HST/PC resolution. With the exception of the "arc" at the edge of the larger cloud, the aligned components and the core are essentially free of Lyalpha line-contamination, as seen in Fig. 8a-8f. The nucleus is a weak Lyalpha source, contributing only about 20% of the total Lyalpha flux (Fig. 8a). Smoothing of the medium-band Lyalpha image shows emission associated with the arc in the western aligned cloud, and otherwise very diffuse Lyalpha emission only (Fig. 8a). The arc at the outer edge of the western cloud contributes as much as 93% of the B-band light. This is the only feature seen in the Lyalpha image with significant contrast against the rest of the galaxy in terms of equivalent width. These two blue clouds could represent:

(1) Reflection of the AGN-light shining through a cone, including Lyalpha and C-IV emission lines from gas lit up by the cone. The asymmetry in size and flux between eastern and western clouds (Fig. 8b & 8e) may represent obscuration or genuine physical differences. The fact that we can see a much larger and somewhat symmetric ground-based Lyalpha cloud (compare our Fig. 8a to Fig. 3 of W91) argues against obscuration, although this extended Lyalpha gas could be mostly in front of - or away from - any dust, and in part unrelated to the AGN. Two other z appeq 2.40 objects - Nos. 18 & 19 of P96b - also are AGN with continuum reflection cones (see the color Plate of W98), but with a relatively stronger AGN component compared to the surrounding material (at the 50-80% level of the total flux; P96b). The presence of these reflection cones implies the existence of a substantial amount of gas and/or dust well beyond the optical extent of these galaxies.

(2) A star-bursting region induced by the weak radio jet. Compared to the spectral evolution models described by W94b, the much bluer colors of the cloud - if caused by stars - would suggest a star-bursting region ltapprox 108 years old. This is similar to the typical radio source lifetime, but younger than the galaxy's dynamical time scale. A color map - produced by matching the registration, sampling and resolution of the B450 & I814 images - shows the color contrast between the inner and outer regions of 53W002 (Fig. 8f). A red, almost linear feature appears to separate the smaller cloud from the nucleus. In nearby galaxies, this would suggest an organized dust lane. If this feature is indeed a dust lane, it would have a differential optical depth between 1300 and 2400 Å ranging from tau appeq 0.75-1.5 averaged over the resolution limit. This is rather mild by standards of present-day galaxies. The visual or blue extinction expected for this amount of far-UV extinction would be easy to miss in nearby radio galaxies, so the total amount of dust required is not excessive for objects like 53W002, which might be chemically younger and correspondingly more metal-poor (KW91, K97).

Next Contents Previous