1: \documentclass[preprint2]{aastex}
2: \shorttitle{NGC 1333 IRAS 4A Outflow Variability}
3: \shortauthors{Choi et al.}
4: \received{}
5:
6: % For preprint only
7: \usepackage{times,epsfig}
8: \frenchspacing
9: \slugcomment{To appear in the Astrophysical Journal}
10: % custom macros
11: \newcommand{\HCOp}{{\rm HCO$^+$}}
12: \newcommand{\Ht}{{\rm H$_2$}}
13: \newcommand{\JJ}[2]{\mbox{$J = #1\rightarrow#2$}}
14: \newcommand{\kms}{\mbox{km s$^{-1}$}}
15: \newcommand{\ploteps}[2]{\centerline{\ifdim#2=0mm\epsfig{file=#1}%
16: \else\epsfig{file=#1,width=#2}\fi}}
17: \newcounter{panel}[figure]
18: \renewcommand{\thepanel}
19: {{\count20=1\advance\count20by\value{figure}%
20: \stepcounter{panel}\small%
21: \sc Fig. \the\count20\it\alph{panel}}}
22: \newcommand{\plotpanel}[3]
23: {\vbox{\hsize=#2\epsfig{file=#1,width=#2}
24: \ifnum#3=1\makebox[\hsize]{\thepanel}\fi}}
25: \renewcommand{\floatpagefraction}{0.8}
26: \renewcommand{\dblfloatpagefraction}{0.8}
27:
28: \begin{document}
29: \fontsize{10}{10.6}\selectfont
30: \title{Variability of the NGC 1333 IRAS 4A Outflow:
31: Molecular Hydrogen and Silicon Monoxide Images\altaffilmark{1}}
32: \author{\sc Minho Choi\altaffilmark{2,3},
33: Klaus W. Hodapp\altaffilmark{4},
34: Masahiko Hayashi\altaffilmark{5},
35: Kentaro Motohara\altaffilmark{5,6},}
36: \author{\sc Soojong Pak\altaffilmark{2,7},
37: and Tae-Soo Pyo\altaffilmark{5}}
38: \altaffiltext{1}{Based in part on data collected at Subaru Telescope,
39: which is operated
40: by the National Astronomical Observatory of Japan.}
41: \altaffiltext{2}{Korea Astronomy and Space Science Institute,
42: Hwaam 61-1, Yuseong, Daejeon 305-348, South Korea.}
43: \altaffiltext{3}{minho@kasi.re.kr.}
44: \altaffiltext{4}{Institute for Astronomy, University of Hawaii,
45: 640 North A'ohoku Place, Hilo, HI 96720;
46: hodapp@ifa.hawaii.edu.}
47: \altaffiltext{5}{Subaru Telescope,
48: National Astronomical Observatory of Japan,
49: 650 North A'ohoku Place, Hilo, HI 96720.}
50: \altaffiltext{6}{Institute of Astronomy, University of Tokyo,
51: Mitaka, Tokyo 181-0015, Japan.}
52: \altaffiltext{7}{Department of Astronomy and Space Science,
53: Kyung Hee University, Seocheon, Giheung,
54: Yongin, Gyeonggi 446-701, South Korea.}
55: \setcounter{footnote}{7}
56:
57: \begin{abstract}
58: \fontsize{10}{10.6}\selectfont
59: The NGC 1333 region was observed in the H$_2$ 1--0 $S$(1) line.
60: The H$_2$ images cover a 5$'$ $\times$ 7$'$ region around IRAS~4.
61: Numerous H$_2$ emission features were detected.
62: The northeast-southwest bipolar outflow driven by IRAS 4A was studied
63: by combining the H$_2$ images with SiO maps published previously.
64: The SiO-H$_2$ outflows are continuous on the southwestern side
65: but show a gap on the northeastern side.
66: The southwestern outflow lobe curves smoothly,
67: and the position angle increases with the distance from the driving source.
68: The base and the outer tip of the northeastern outflow lobe
69: are located at positions
70: opposite to the corresponding parts of the southwestern lobe.
71: This point-symmetry suggests
72: that the outflow axis may be drifting or precessing clockwise
73: in the plane of the sky
74: and that the cause of the axis drift may be intrinsic to the outflow engine.
75: The axis drift model is supported
76: by the asymmetric lateral intensity profile of the SiO outflow.
77: The axis drift rate is $\sim$0\fdg011 yr$^{-1}$.
78: The middle part of the northeastern outflow
79: does not exactly follow the point symmetry
80: because of the superposition
81: of two different kinds of directional variability:
82: the axis drift of the driving source and the deflection by a dense core.
83: The axis drift model provides a good explanation
84: for the large deflection angle of the northeastern outflow.
85: Other H$_2$ emission features around the IRAS 4 region
86: are discussed briefly.
87: Some of them are newly found outflows,
88: and some are associated with outflows already known before.
89: \end{abstract}
90:
91: \keywords{ISM: individual (NGC 1333 IRAS 4) --- ISM: jets and outflows
92: --- ISM: structure --- stars: formation}
93:
94: \section{INTRODUCTION}
95:
96: The NGC 1333 region contains
97: numerous young stellar objects and outflows
98: (Aspin et al. 1994; Hodapp \& Ladd 1995; Rodr\'{\i}guez et al. 1999).
99: It is one of the most active sites of star formation
100: in the solar neighborhood,
101: and the large number of active outflow sources
102: was described as a ``microburst'' of star formation (Bally et al. 1996).
103: Several protostars are located in the NGC 1333 molecular cloud,
104: and IRAS 4A is one of the brightest submillimeter sources among them
105: (Sandell \& Knee 2001).
106: IRAS 4A is a Class 0 protobinary system
107: (Sandell et al. 1991; Lay et al. 1995; Looney et al. 2000).
108:
109: Single-dish observations showed
110: an interesting directional variability of the IRAS 4A molecular outflow
111: (Blake et al. 1995).
112: Detailed structure of the IRAS 4A molecular outflows
113: within $\sim$1$'$ of the driving source
114: was revealed by interferometric observations.
115: The interferometric maps showed
116: that the apparently large change of outflow direction
117: is partly owing to the existence of two outflows
118: driven by IRAS 4A (Choi 2001b)
119: and partly owing to the sharp bend (or deflection)
120: in the northeastern lobe of the main bipolar outflow
121: (Girart et al. 1999; Choi 2005a, hereafter Paper I).
122: Choi (2005a) suggested
123: that the sharp bend was caused by a collision
124: between the northeastern outflow
125: and a dense core in the ambient molecular cloud.
126: The jet-core collision model was supported by several lines of evidence
127: including the asymmetric morphology of the main outflow,
128: the HCN emission from shocked high-density gas near the impact point,
129: and the high-density core located in the course of the northeastern outflow
130: (Paper I; Choi 2005b).
131: However, the deflection cannot completely explain
132: the difference of position angles
133: between the SiO outflow near the driving source
134: and the large-scale ($>$ 1$'$) CO/\Ht\ outflow.
135:
136: To investigate the overall structure
137: of the NGC 1333 IRAS 4A outflow system,
138: we observed the outflows in several molecular tracers.
139: In Paper I, we presented high angular resolution observations
140: in the SiO $v=0$ \JJ10\ line using the Very Large Array (VLA).
141: In this paper, we present our observations in the \Ht\ 1--0 $S$(1) line
142: using the University of Hawaii (UH) 2.2 m telescope
143: and the Subaru telescope.
144: In \S~2 we describe our \Ht\ observations.
145: In \S~3 we report the results of the \Ht\ imaging.
146: In \S~4 we compare the \Ht\ and the SiO images
147: and discuss the variability of the IRAS 4A main outflow.
148:
149: \section{OBSERVATIONS}
150:
151: \subsection{UH 2.2 {\rm m} Telescope Observations}
152:
153: The molecular hydrogen emission features associated with IRAS 4
154: were included in a wide field image of the NGC 1333 region
155: in the \Ht\ 1--0 $S$(1) emission line
156: obtained in the nights of 1996 July 30 to August 3
157: using the QUIRC camera (Hodapp et al. 1996) at the UH 2.2 m telescope.
158: The individual integration time was 300 s.
159: The image was produced by stepping the telescope
160: in steps of 30$''$ in the north-south direction.
161: Therefore, most positions in the image were
162: covered by six individual frames for a total integration time of 30 min.
163: In the overlap region between adjacent vertical stripes,
164: the coverage is 12-fold for a total integration time of 60 min.
165:
166: The UH \Ht\ image shows a set of artifacts
167: that are the result of the residual excess dark current
168: of the HgCdTe array used for the observations.
169: This particular array, one of the earliest HAWAII-I devices,
170: showed a strong excess of dark current
171: in pixels previously exposed to high signal levels,
172: with a decay time of the order of many minutes.
173: As a result of this effect,
174: combined with our method of collecting the individual images of this mosaic
175: in a north-to-south stepping pattern,
176: bright sources, most prominently SVS 13,
177: are associated with a long string of residual images south of them,
178: in 30$''$ spacing, and with slowly decaying intensity.
179: The faint extended nebulosity discussed in this paper
180: is not affected by this artifact.
181:
182: \subsection{Subaru Telescope Observations}
183:
184: Imaging was carried out on 2000 September 13
185: using the Cooled Infrared Spectrograph and Camera
186: for OH-Airglow Suppressor (CISCO),
187: which was equipped with a 1024 $\times$ 1024 HgCdTe detector
188: covering a field of view of 110$''$ $\times$ 110$''$
189: with a plate scale of 0\farcs11 pixel$^{-1}$ (Motohara et al. 2002).
190: We obtained narrow band images
191: with and without \Ht\ 1--0 $S$(1) line emission.
192: The \Ht\ 1--0 filter was centered on $\lambda$ = 2.120 $\mu$m
193: with a width of 0.020 $\mu$m (FWHM),
194: while the N215 filter for continuum emission
195: was centered on 2.147 $\mu$m with a width of 0.021 $\mu$m (FWHM).
196: Sixteen exposures with 60 s each were made for each of the filters
197: by stepping the telescope in steps of 10$''$--20$''$
198: along the east-west and/or north-south directions,
199: resulting the total on-source exposure time of 960 s
200: for the central 90$''$ $\times$ 90$''$ region.
201: Each frame of \Ht\ 1--0 and N215 filter images was flat-fielded
202: using sky-flat data and was subtracted by a corresponding sky frame,
203: which was made for each of the two filters with median filtering.
204: It was found that all the extended features in the \Ht\ 1--0 filter image
205: arise from the \Ht\ 1--0 $S$(1) line
206: by comparing it with the N215 filter image.
207: We thus use only the \Ht\ 1--0 filter image in this paper.
208:
209: \subsection{Astrometric Calibration}
210:
211: Accurate astrometry is essential
212: to comparing the \Ht\ image with the SiO image.
213: Since the adjacent continuum emission was not subtracted,
214: the \Ht\ images include both the line and the continuum emission.
215: In the UH \Ht\ image,
216: 20 field stars listed in the 2MASS Point Source Catalogue%
217: \footnote{This publication makes use of data products
218: from the Two Micron All Sky Survey (2MASS),
219: which is a joint project of the University of Massachusetts
220: and the Infrared Processing
221: and Analysis Center/California Institute of Technology,
222: funded by the National Aeronautics and Space Administration
223: and the National Science Foundation.}
224: were identified.
225: In the Subaru \Ht\ image, only three 2MASS stars were found.
226:
227: With the 2MASS coordinate data,
228: the {\it geomap} package of {\it IRAF} was used
229: to find the spatial transformation functions.
230: Then the {\it geotran} package was used to transform the observed images.
231: Comparing the stellar coordinates
232: of the transformed UH \Ht\ image and of the 2MASS catalogue,
233: the rms of position difference is 0\farcs19.
234: This value is about the same order of the position uncertainties
235: in the 2MASS catalogue,
236: which confirms that the transformation is acceptable.
237:
238: Angular resolutions of the resulting images were measured
239: by fitting Gaussian intensity profiles to stellar objects.
240: The angular resolutions (FWHM) are
241: 0\farcs82 for the UH \Ht\ image and 0\farcs46 for the Subaru \Ht\ image.
242:
243: \subsection{VLA Data}
244:
245: Details of the VLA observations and the results were presented in Paper I.
246: The NGC 1333 IRAS 4 region was observed using VLA
247: in the SiO $v=0$ \JJ10\ line.
248: The resulting images have a restoring beam of FWHM = 1\farcs96.
249: The SiO image covers a region up to $\sim$47$''$ from IRAS 4A.
250:
251: \section{RESULTS}
252:
253: Figure 1 shows the UH \Ht\ image of the NGC 1333 cloud
254: covering a 5\farcm2 $\times$ 7\farcm5 region.
255: The Subaru \Ht\ image (Fig. 2) has a better sensitivity
256: and a higher angular resolution than the UH image,
257: but the field of view is smaller.
258: In Figures 3$b$ and 4$a$, the VLA SiO map of the IRAS 4A outflows
259: is superposed on the \Ht\ images for comparison.
260:
261: \input{fig1.tex} \input{fig2.tex}
262: \input{fig3.tex} \input{fig4.tex}
263:
264: The structure of the \Ht\ emission features in the NGC 1333 region
265: was previously discussed by Hodapp \& Ladd (1995).
266: Recent observations in the submillimeter and centimeter continuum
267: (Rodr\'{\i}guez et al. 1999; Sandell \& Knee 2001)
268: provide helpful information for understanding the relation
269: between the \Ht\ emission features
270: and young stellar objects in this region.
271: The NGC 1333 region is crowded with outflows,
272: and the \Ht\ emission features should be inspected carefully
273: to avoid confusion.
274:
275: Since the main interest of this paper
276: is the structure of the IRAS 4A molecular outflows,
277: we describe the \Ht\ emission features related to these flows first.
278: While the \Ht\ images show the large-scale structure of the main outflow,
279: little \Ht\ emission can be seen near the driving source,
280: probably owing to a high extinction through the dense core.
281: In contrast, the SiO line nicely shows
282: the outflow structure near the driving source,
283: probably owing to a high density.
284: Therefore, the \Ht\ images and the SiO map
285: provide complementary information.
286:
287: On the northeastern side,
288: there is a relatively wide ($\sim$35$''$) gap
289: between ASR 57 and the northeastern end of the SiO outflow (Fig. 3$b$).
290: This gap is probably real
291: because CO outflow maps also show a similar gap
292: (see Fig. 3 of Blake et al. 1995).
293: A chain of \Ht\ features (ASR 57, HL 10, and HL 11 = HH 347A)
294: seems to belong to the IRAS 4A main outflow.
295: Each of these features is elongated in the general direction of the flow.
296: HH 347B is located away from the chain of \Ht\ features,
297: and it probably does not belong to the IRAS 4A outflow.
298:
299: On the southwestern side,
300: the blueshifted SiO outflow is connected with the \Ht\ outflow,
301: and there is a considerable overlap (Fig. 4$a$).
302: Near the driving source,
303: the faint \Ht\ feature CHH 15 coincides with SiO outflow peak (SiOOP) 7.
304: The \Ht\ feature CHH 17 coincides with SiOOP 10 and 11
305: within a few arcseconds,
306: and the northeastern end of CHH 18 coincides with SiOOP 12 (Fig. 4$a$).
307: Further away from the driving source,
308: a chain of \Ht\ features (HL 5/3)
309: seems to belong to the IRAS 4A main outflow.
310:
311: Along the southern SiO outflow,
312: the \Ht\ feature HL 6 is located in the middle of SiOOP 14, 15, and 16.
313: Unlike the main outflow,
314: the southern outflow shows no \Ht\ emission feature
315: beyond the SiO outflow lobe.
316: Therefore, the VLA SiO map probably traces
317: the whole extent of the southern outflow.
318:
319: \subsection{Proper Motion}
320:
321: Proper motions of \Ht\ features were measured
322: by comparing the Subaru and the UH images.
323: The coordinate systems of the two images were aligned
324: by comparing the positions of three stars near the \Ht\ features
325: (ASR 53, ASR 103, and CHH 18 IRS).
326: Since these stars are visible only in the near-IR or longer wavelengths,
327: they are either stars embedded in the NGC 1333 molecular cloud
328: or background stars,
329: and their proper motions may be negligible.
330: For bright \Ht\ features with well-defined peaks,
331: position offsets were measured by comparing pixels around the peaks,
332: typically pixels within $\sim$2$''$ from the peak positions.
333: Additional measurements were made by comparing the Subaru image
334: with the \Ht\ image of Hodapp \& Ladd (1995) in a similar manner,
335: which provides a longer time baseline but a larger uncertainty
336: owing to the worse angular resolution.
337: The two sets of measurements were consistent to each other,
338: and the average proper-motion vectors are shown in Figure 5.
339: The uncertainty of position offset was measured
340: from the residual position differences of the five stars
341: common to both images (the three stars above and ASR 54/55).
342:
343: \input{fig5.tex}
344:
345: The proper-motion vectors of HL 4 peaks
346: show a systematic motion toward southeast,
347: and those of HL 5 peaks show a motion toward southwest.
348: Assuming a distance of 320 pc (de Zeeuw et al. 1999),
349: the proper motion of HL 5 peaks ranges from 30 to 120 \kms\
350: with an uncertainty of 50 \kms.
351: In the discussions below (\S~4),
352: a proper motion of 100 \kms\ or 0\farcs07~yr$^{-1}$ is assumed
353: for calculations of the timescale of the IRAS 4A main outflow.
354:
355: \section{DISCUSSION}
356:
357: \subsection{Large-Scale Variability of the IRAS 4A Main Outflow}
358:
359: As discussed in Paper I,
360: the northeastern SiO outflow shows a directional variability
361: that is probably caused by a collision with a dense core.
362: On top of this variability,
363: yet another directional variability is revealed
364: by combining the SiO map with the \Ht\ images.
365: This variability appears to affect
366: both the southwestern and the northeastern outflows
367: in a large ($\sim$5$'$) scale.
368: Since the northeastern outflow is more complicated
369: owing to the superposition of the two different variabilities,
370: we describe the southwestern outflow first.
371:
372: The southwestern lobe of the SiO outflow appears very straight:
373: all the SiO outflow peaks are located
374: within a beam size from a straight line (see Fig. 2$a$ of Paper I).
375: When the SiO and the \Ht\ images are superposed, however,
376: the southwestern outflow appears to curve smoothly.
377: That is, the position angle (PA) of the outflow peaks
378: with respect to the driving source (IRAS 4A2)
379: increases with the angular distance (Fig. 3$a$ {\it top panel}).
380: Polynomial fits were drawn in Figure 3 to outline the flow.
381: A change of flow direction can be caused
382: either by intrinsic variability,
383: such as geometric change of the outflow engine,
384: or by external perturbation,
385: such as density gradients in the external medium.
386: The nearly constant gradient of the PA-distance relation suggests
387: that the large-scale variability is
388: probably owing to an intrinsic variability.
389: Another important clue to the variability mechanism
390: may come from the symmetry of the outflow morphology.
391:
392: Near the driving source,
393: the direction of the northeastern outflow
394: is exactly opposite to that of the southwestern outflow.
395: The northeastern outflow changes direction abruptly
396: at a point $\sim$23$''$ away from the driving source,
397: and the position angle increases at a very high rate
398: (Fig. 3 {\it bottom panel}).
399: At a large distance, however, the PA-distance relation converges
400: back to the reverse of the southwestern flow.
401: Therefore, it appears that the southwestern and the northeastern outflows
402: would have shown a change of flow direction in a point-symmetric way,
403: if it were not for the jet-core collision of the northeastern outflow.
404: The symmetric morphology suggests
405: that the change of flow direction in the large scale
406: is caused by a geometric reconfiguration of the outflow engine.
407: That is, the axis of the IRAS 4A2 accretion disk may be changing direction,
408: and the outflow axis may be drifting%
409: \footnote{Outflows with changing directions
410: are often described as ``precessing''.
411: In this paper, we use the term ``drifting'' instead of precessing
412: because there is no evidence that the change is periodic.}
413: together.
414: The axis appears to drift clockwise in the plane of the sky.
415:
416: The IRAS 4A outflow is not the only example
417: of point-symmetric, or {\sf S}-shaped, outflow
418: driven by young stellar objects.
419: Other examples include
420: the HH 340/343 outflow driven by NGC 1333 J032845.3+310542 (K)
421: and the \Ht\ jet driven by V380 Ori NE
422: (Hodapp et al. 2005; Davis et al. 2000).
423: Some outflows driven by high-mass (proto)stars
424: appear to have similar shapes,
425: e.g., IRAS 20126+4104 (Shepherd et al. 2000).
426:
427: \subsection{Lateral Structure of the {\rm SiO} Outflow}
428:
429: The lateral structure of the SiO outflow
430: was briefly described in \S~3.3 of Paper I.
431: Here we describe the structure in more detail
432: and discuss it in the context of the drifting axis.
433: Figure 4$b$ shows the lateral intensity profiles.
434: The SiO intensity profiles
435: of the undeflected part of the main outflow (peaks 4--12)
436: usually show a steep slope on the clockwise side
437: and a relatively slower slope (or shoulder) on the counterclockwise side.
438:
439: The asymmetric slopes of the intensity profiles
440: can be explained by the drift of the flow axis.
441: Since the axis of the primary jet drifts clockwise,
442: the ambient molecular gas on the leading (clockwise) side
443: would be relatively fresh (and probably denser),
444: while the molecular gas on the opposite side
445: would have been already disturbed and accelerated in the past.
446: The shear zone between the primary jet and the ambient medium
447: may be thinner on the leading side than on the trailing side.
448: Then the SiO molecules in the shear zone of the leading side
449: would be more strongly shocked, excited more highly,
450: and make stronger emission,
451: which could produce the asymmetric intensity profile
452: across the outflow lobe.
453: On the trailing side,
454: the SiO emitting gas may be gradually cooling down
455: as the jet drifts away.
456: Therefore, the asymmetric lateral intensity profile provides
457: additional support for the drifting axis model.
458:
459: \subsection{Drifting Axis of the IRAS 4A Main Outflow}
460:
461: The PA-distance relation of the main outflow (Fig. 3$a$) suggests
462: that the amount of axis drift
463: within the observed extent of the main outflow
464: (i.e., from the vicinity of the driving source to HL 3)
465: is $\sim$20\arcdeg\
466: and that the rate of drift is $\sim$10\arcdeg\ arcminute$^{-1}$.
467: Assuming a proper motion of 0\farcs07 yr$^{-1}$ (see \S~3.1),
468: the temporal rate of axis drift is $\sim$0\fdg011 yr$^{-1}$.
469:
470: Several authors suggested possible mechanisms
471: of intrinsic outflow variability (e.g., Eisl{\"o}ffel \& Mundt 1997;
472: Fendt \& Zinnecker 1998; Shepherd et al. 2000),
473: but most of them cannot generate a large amount of drift
474: as seen in the IRAS 4A outflow.
475: The tidal interaction between IRAS 4A1 and A2 cannot explain the drift
476: because the separation between them ($>$ 540 AU) is
477: too large to cause a strong effect.
478: In principle, A2 itself could be a close binary system
479: that has not been observationally resolved yet.
480: Assuming a proper motion of 0\farcs07 yr$^{-1}$,
481: the 130$''$ separation between IRAS 4A and HL 3b
482: gives an outflow timescale of $\sim$2000 yr.
483: Since the flow is not showing multiple turns,
484: the precession period would be longer than 4000 yr,
485: but it should be shorter than $\sim$16000 yr
486: to keep the precession angle reasonably small, say less than 90\arcdeg.
487: Since the precession period would be longer
488: than the binary orbital period by a factor of $\sim$20 (Bate et al. 2000),
489: the expected orbital period of the unresolved binary system is 200--800 yr.
490: The corresponding binary separation would be 30--80 AU,
491: assuming that the mass of IRAS 4A2 is $\sim$0.9 $M_\odot$,
492: which is about half the mass of IRAS 4A
493: (mass from Choi 2001b, scaled to 320 pc).
494: Then the angular separation would be 0\farcs09--0\farcs25,
495: which is too small for existing millimeter interferometers to resolve.
496: The accretion disk in such a system, however,
497: would probably be too small to drive an energetic outflow.
498:
499: Another possible mechanism is the anisotropic accretion
500: that can occur
501: when the angular momentum of currently accreting mass is not parallel
502: to the rotation axis of previously accreted matter.
503: Considering that IRAS 4A is a binary system,
504: we propose the following scenario.
505: The initial IRAS 4A core had fragmented into two subcores (A1 and A2),
506: and each of them collapsed to form a protostar
507: and started an outflow activity.
508: While the total angular momentum of the A1/A2 system
509: may be parallel to the rotation axis of the initial core,
510: the spin axis of each individual protostar
511: (and the initial direction of each outflow)
512: could have been oriented differently.
513: (This scenario is consistent with the observations
514: because there is a considerable difference
515: between the directions of the main outflow driven by A2
516: and the southern outflow driven by A1.)
517: As the protostar accumulates mass from the large outer envelope,
518: the axis of the accretion disk would shift gradually
519: and converge to the direction of the average angular momentum
520: of the whole protostellar core.
521: This scenario can be tested in principle
522: by observationally determining the rotation axis of the IRAS 4A core,
523: but the velocity structure of this region
524: may be too complicated (Choi et al. 2004) for such a test.
525: In the anisotropic accretion model,
526: the change of direction may happen only once,
527: i.e., may not be periodic or oscillating,
528: which can explain
529: why most of the direction-changing outflows observed so far
530: do not show multiple turns.
531:
532: Yet another possibility is the misalignment
533: between the rotation and magnetic fields.
534: Numerical simulations show
535: that the rotation axis, magnetic fields, and disk orientation
536: can precess significantly (Matsumoto \& Tomisaka 2004).
537: In such systems, changes of outflow direction may be observable.
538: The evolution of the field direction with respect to the rotation axis
539: may depend on the relative strength
540: between the rotation and the magnetic fields (Machida et al. 2006).
541:
542: \subsection{Deflection of the Northeastern Outflow}
543:
544: Theoretical models of jet-cloud collision predict
545: that, in the simple cases of steady jet,
546: the deflection process finishes in a few hundred years
547: either because the jet penetrates the dense clump
548: or because the jet is pinched off
549: (Raga \& Cant{\'o} 1995; de Gouveia Dal Pino 1999).
550: When the whole extent of the \Ht\ outflow is considered,
551: the timescale of the IRAS 4A main outflow is $\sim$2000 yr.
552: Then the deflection process of a steady jet
553: would have been almost finished in this timescale,
554: and the deflection angle would have been small.
555: The large deflection angle observed in the northeastern outflow
556: suggests that the real situation is more complicated than the simple models.
557: It was suggested that the deflection timescale can be longer
558: if the jet beam is not steady
559: so that the impact point does not stay at one position (Raga et al. 2002).
560: Therefore, the drifting axis model provides a good explanation
561: for the large deflection angle currently observed in the IRAS 4A outflow.
562:
563: Assuming that the width of the (undeflected) primary jet at the impact point
564: is smaller than 4\farcs3 (deconvolved width of SiOOP 4),
565: and considering the axis drift rate in the previous section,
566: it takes less than $\sim$1000 yr
567: for the impact point at the ``surface'' of the dense core
568: to move across a jet width.
569: Since this crossing time is
570: comparable to or shorter than the cloud penetration timescale,
571: the large deflection angle can be maintained.
572:
573: The deflection history of the northeastern outflow
574: can be summarized in the following scenario.
575: Choi (2005b) suggested
576: that the dense core obstructing the northeastern outflow is
577: located $\sim$50$''$ from IRAS 4A2
578: with a position angle of $\sim$16\arcdeg\ (Figs. 3$b$ and 6).
579: The detection of \HCOp\ and HCN emission suggests
580: that the obstructing core is very dense ($\sim$10$^5$ cm$^{-3}$)
581: (Paper I; Choi 2005b).
582: In the distant past,
583: the direction of the outflow had a large position angle,
584: and the flow would have passed by the eastern side of the obstructing core
585: with a large impact parameter
586: and would have been deflected only slightly.
587: For example, HL 11 would have had an (projected) impact parameter
588: of $\sim$19$''$ (or 6000 AU at 320 pc).
589: As the outflow axis drifts clockwise,
590: the impact point would move
591: on the ``surface'' of the dense core to the west,
592: and the impact parameter decreases.
593: Consequently, the outflow would be deflected more and more severely,
594: and the deflection angle may increase with time.
595: The position angle of the outflow axis
596: in the vicinity of the driving source is $\sim$19\arcdeg,
597: and the jet currently colliding against the dense core
598: may have an impact parameter of $\sim$3$''$ or 900 AU.
599: That is, the northeastern outflow is
600: currently making a nearly head-on collision against the obstructing core,
601: and consequently the deflection angle is large.
602:
603: \input{fig6.tex}
604:
605: Reipurth et al. (1996) suggested
606: that the HH 270/110 system may be an example of deflected outflow.
607: This interpretation, however, was questioned
608: because high-density molecular gas was not detected
609: near the proposed impact point (Choi 2001a).
610: Raga et al. (2002) suggested
611: that the apparent conflict can be reconciled
612: if the impact region is located
613: far away from the crossing point of the HH 270 and the HH 110 axes.
614: If the HH 110 flow is indeed a deflected segment of the HH 270 flow,
615: its deflection history must be more complicated than the case of IRAS 4A.
616: More detailed and sensitive observations are needed
617: to understand the HH 270/110 system better.
618:
619: \subsection{Other \Ht\ Emission Features}
620:
621: The \Ht\ images (Figs. 1 and 2) show several emission features
622: that were previously undetected or unlabeled.
623: Coordinates of their peak positions are listed in Table 1.
624: These \Ht\ features are labeled with a prefix CHH and briefly described here.
625: Some of the previously known \Ht\ features are also discussed below.
626:
627: \input{tab1.tex}
628:
629: {\it CHH 1.}---%
630: This is a short jet-like \Ht\ feature.
631: A probable driving source is VLA 14.
632:
633: {\it CHH 2/4.}---%
634: These \Ht\ features and HH 347B probably belong to a single outflow.
635:
636: {\it CHH 3.}---%
637: This is a short jet-like \Ht\ feature.
638: A probable driving source is ASR 74, the star just north of CHH 3.
639:
640: {\it CHH 5.}---%
641: This is a linear chain of \Ht\ emission features.
642: Hodapp \& Ladd (1995) suggested
643: that CHH 5 and HL 4 are related with the SVS 13 complex.
644: Another possibility is
645: that the CHH~5--HL 4 flow may be driven by VLA 19 (SK 14).
646: Yet another possibility is that CHH 5b,
647: which is elongated roughly in the northeast-southwest direction,
648: may be a short outflow driven by VLA 19,
649: while CHH 5a/c (and HL 4) may belong to a separate outflow.
650:
651: {\it CHH 6.}---%
652: This is an unusually straight jet-like \Ht\ feature.
653: It appears to be related with ASR 20/21 and probably with HH 344.
654:
655: {\it CHH 7/10/14.}---%
656: These are isolated \Ht\ features with fuzzy appearances.
657: Their nature is unclear.
658: CHH 10 is associated with HH 759 (Walawender et al. 2005).
659:
660: {\it CHH 8/9.}---%
661: These \Ht\ features seem to belong to the HH 5 system.
662:
663: {\it CHH 11.}---%
664: This is a system of weak \Ht\ features.
665: It appears to be related with HL 8.
666:
667: {\it CHH 12/13.}---%
668: These \Ht\ features probably belong to a single outflow.
669: A probable driving source is SK 1.
670:
671: {\it CHH 16.}---%
672: This is a faint \Ht\ feature associated with SiOOP 17.
673: Since SiOOP 17 is kinematically distinct from the IRAS 4A main outflow,
674: it is not clear if it is related with the IRAS 4A system.
675: See \S~3.1 of Paper I for details.
676:
677: {\it CHH 19.}---%
678: This \Ht\ feature probably belongs to the IRAS 4A outflow.
679:
680: {\it HL 8.}---%
681: Hodapp \& Ladd (1995) suggested that HL 8 may be an extension of HL 4.
682: Considering the locations of SK1 and CHH 12/13, however,
683: HL 8 could be the counterpart of the CHH 12--13 outflow.
684:
685: {\it HL 9.}---%
686: At $\sim$8$''$ south of IRAS 4BI,
687: the two peaks of HL 9 line up in the north-south direction,
688: which agrees with the direction of the IRAS 4BI bipolar outflow
689: seen in the HCN line (Choi 2001b).
690:
691: {\it ASR 57.}---%
692: This is one of the brightest \Ht\ emission features in the region imaged.
693: The location of ASR 57 at the intersection
694: of the IRAS 4A outflow and the HH 7--11--HL 7 outflow
695: raised the possibility
696: that the two flows might be interacting (Hodapp \& Ladd 1995).
697: Since the dense core obstructing the IRAS 4A outflow
698: is a part of the large cloud
699: associated with SVS 13 and the HH 7--11 complex (Paper I; Choi 2005b),
700: the two outflows may be close to each other along the line of sight.
701: Further studies are needed to test
702: whether or not they are physically interacting indeed.
703: Proper motion studies of the \Ht\ peaks within ASR 57 may be useful.
704:
705: {\it VLA 13/30/32.}---%
706: These centimeter continuum sources were detected
707: in the UH image as point-like sources.
708: The agreement between the VLA position and the near-IR position
709: is good within 0\farcs2.
710:
711: \acknowledgements
712:
713: We thank Mike Nassir for help with the reduction
714: of the near-infrared image from the UH 2.2 m telescope.
715: This work was partially supported by the LRG Program of KASI,
716: by the Korea-Japan Basic Scientific Cooperation Program of KOSEF,
717: and by the Japan-Korea Basic Scientific Cooperation Program of JSPS.
718:
719: \begin{references}
720: \reference{} Aspin, C., Sandell, G., \& Russell, A. P. G. 1994,
721: A\&AS, 106, 165
722: \reference{} Bally, J., Devine, D., \& Reipurth, B. 1996, ApJ, 473, L49
723: \reference{} Bate, M. R., Bonnell, I. A., Clarke, C. J., Lubow, S. H.,
724: Ogilvie, G. I., Pringle J. E., \& Tout, C. A. 2000,
725: MNRAS, 317, 773
726: \reference{} Blake, G. A., Sandell, G., van Dishoeck, E. F.,
727: Groesbeck, T. D., Mundy, L. G., \& Aspin, C. 1995,
728: \reference{} Choi, M. 2001a, ApJ, 550, 817
729: \reference{} Choi, M. 2001b, ApJ, 553, 219
730: \reference{} Choi, M. 2005a, ApJ, 630, 976 (Paper I)
731: \reference{} Choi, M. 2005b, Journal of the Korean Physical Society, 47, 533
732: \reference{} Choi, M, Kamazaki, T., Tatematsu, K., \& Panis, J.-F. 2004,
733: ApJ, 617, 1157
734: \reference{} Davis, C. J., Dent, W. R. F., Matthews, H. E., Coulson, I. M.,
735: \& McCaughrean, M. J. 2000, MNRAS, 318, 952
736: \reference{} de Gouveia Dal Pino, E. M. 1999, ApJ, 526, 862
737: \reference{} de Zeeuw, P. T., Hoogerwerf, R., de Bruijne, J. H. J.,
738: Brown, A. G. A., \& Blaauw, A. 1999, AJ, 117, 354
739: \reference{} Eisl{\"o}ffel, J., \& Mundt, R. 1997, AJ, 114, 280
740: \reference{} Fendt, C., \& Zinnecker, H. 1998, A\&A, 334, 750
741: \reference{} Girart, J. M., Crutcher, R. M., \& Rao, R. 1999, ApJ, 525, L109
742: \reference{} Hodapp, K. W., Bally, J., Eisl{\"o}ffel, J.,
743: \& Davis, C. J. 2005, AJ, 129, 1580
744: \reference{} Hodapp, K.-W., et al. 1996, New Astronomy, 1, 177
745: \reference{} Hodapp, K.-W., \& Ladd, E. F. 1995, ApJ, 453, 715
746: \enlargethispage{-14\baselineskip}
747: \reference{} Lay, O. P., Carlstrom, J. E., \& Hills, R. E. 1995,
748: ApJ, 452, L73
749: \reference{} Looney, L. W., Mundy, L. G., \& Welch, W. J. 2000, ApJ, 529, 477
750: \reference{} Machida, M. N., Matsumoto, T., Hanawa, T., \& Tomisaka, K. 2006,
751: ApJ, in press
752: \reference{} Matsumoto, T., \& Tomisaka, K. 2004, ApJ, 616, 266
753: \reference{} Motohara, K., et al. 2002, PASJ, 54, 315
754: \reference{} Raga, A. C., \& Cant{\'o}, J. 1995, Rev. Mex. AA, 31, 51
755: \reference{} Raga, A. C., de Gouveia Dal Pino, E. M., Noriega-Crespo, A.,
756: Mininni, P. D., \& Vel{\'a}zquez, P. F. 2002, A\&A, 392, 267
757: \reference{} Reipurth, B., Raga, A. C., \& Heathcote, S. 1996,
758: A\&A, 311, 989
759: \reference{} Rodr\'{\i}guez, L. F., Anglada, G., \& Curiel, S. 1999,
760: ApJS, 125, 427
761: \reference{} Sandell, G., Aspin, C., Duncan, W. D., Russell, A. P. G.,
762: \& Robson, E. I. 1991, ApJ, 376, L17
763: \reference{} Sandell, G., \& Knee, L. B. G. 2001, ApJ, 546, L49
764: \reference{} Shepherd, D. S., Yu, K. C., Bally, J., \& Testi, L. 2000,
765: ApJ, 535, 833
766: \reference{} Walawender, J., Bally, J., \& Reipurth, B. 2005, AJ, 129, 2308
767: \end{references}
768:
769: \vfill\centerline{\small\tt arXiv version}\vspace*{-\baselineskip}
770:
771: \end{document}
772:
773: