FAST [Fe ii] WIND WITH A WIDE OPENING ANGLE FROM L1551 IRS 5 1

Transcription

1 The Astrophysical Journal, 618: , 2005 January 10 # The American Astronomical Society. All rights reserved. Printed in U.S.A. FAST [Fe ii] WIND WITH A WIDE OPENING ANGLE FROM L1551 IRS 5 1 Tae-Soo Pyo, 2 Masahiko Hayashi, 2, 3 Naoto Kobayashi, 2, 4 Alan T. Tokunaga, 5 Hiroshi Terada, 2 Masahiro Tsujimoto, 6 Saeko S. Hayashi, 2, 3 Tomonori Usuda, 2 Takuya Yamashita, 2 Hideki Takami, 2 Naruhisa Takato, 2 and Ko Nedachi 4 Receivved 2004 May 20; accepted 2004 September 23 ABSTRACT We present new velocity-resolved spectra of the [Fe ii] k1.644 m line emission toward the L1551 IRS 5 outflow. The spectra were taken toward the bright [Fe ii] knots PHK 1 and PHK 2 with the slit positions perpendicular to the northern jet. We have two major conclusions: (1) At PHK 1, located 1B2 away from the L1551 IRS 5 VLA sources, the spatial profile of the low radial velocity component at V LSR 110 km s 1 shows two spatial subcomponents with their FWHMs of 0B83 and 2B84. The wide subcomponent has an wide opening angle of 100,whichis consistent with the opening angle suggested by the broad velocity width of the narrow one. It favors the interpretation that both subcomponents of the low-velocity component arise from the same outflow. The gas corresponding to the wide subcomponent fills the space between the optical jet and the shell of the CO molecular outflow and may be sweeping up envelope material in the vicinity of the protostars. (2) At PHK 2, located 4B2 away from the VLA sources, we confirmed that the northern jet has two radial velocity components: V LSR 270 and 140 km s 1. The former velocity component is highly collimated because it has the same spatial width of 0B78 at both of the two [Fe ii] knots. Subject headinggs: infrared: ISM ISM: Herbig-Haro objects ISM: individual (HH 154, L1551) ISM: jets and outflows stars: formation stars: pre main-sequence 1. INTRODUCTION L1551 IRS 5 is the first object toward which a bipolar molecular outflow was discovered (Snell et al. 1980). The outflow was slow (jvj P 20 km s 1 ) and showed apparent linear acceleration (Uchida et al. 1987; Moriarty-Schieven & Snell 1988), which suggested that the molecular outflow is swept-up ambient material set into motion by an invisible wind or jet driven by the central protostar. While optical jets can provide significant energy and momentum to molecular outflows if the jets are mostly neutral (Raga et al. 1990; Hartigan et al. 1994; Bacciotti & Eislöffel 1999), it is unlikely that well-collimated, spatially narrow jets account for the wide opening angle and large spatial width of molecular outflows. This led some researchers (e.g., Shu et al. 1991) to claim the existence of a widely opened high-velocity wind surrounding a well-collimated optical jet. In fact, Lizano et al. (1988) and Giovanardi et al. (1992, 2000) detected a high-velocity atomic wind with a radial velocity of jv max j100 km s 1 inside the molecular outflow lobes of L1551IRS5intheHi21 cm emission. A number of prominent optical and H 2 emission-line features (Herbig-Haro objects) located in the blueshifted CO outflow lobe (Graham & Heyer 1990; Davis et al. 1995) show large proper motions and radial velocities of km s 1 (Stocke et al. 1988), which also suggest the existence of a high-velocity wind inside the CO 1 Based on data collected at the Subaru Telescope, which is operated by the National Astronomical Observatory of Japan. 2 Subaru Telescope, National Astronomical Observatory of Japan, 650 North A ohoku Place, Hilo, HI School of Mathematical and Physical Science, The Graduate University for Advanced Studies (SOKENDAI ), Hayama, Kanagawa , Japan. 4 Institute of Astronomy, University of Tokyo, Mitaka, Tokyo , Japan. 5 Institute for Astronomy, University of Hawaii, 2680 Woodlawn Drive, Honolulu, HI Department of Astronomy and Astrophysics, Pennsylvania State University, 525 Davey Laboratory, University Park, PA lobe. For classical T Tauri stars, the P Cygni profiles of optical permitted lines demonstrate the existence of similar fast winds (Edwards et al. 1993; Mundt et al. 1985). The winds are, however, compact and have never been spatially resolved, which make it difficult to understand their location and morphology. In a previous paper we presented a position-velocity diagram of the [Fe ii] k1.644 m emission along the northern jet of L1551 IRS 5 (Pyo et al. 2002, hereafter Paper I). We detected in the northern jet two distinct components that are well separated from each other in radial velocity and space. The high-velocity component (HVC; 300 km s 1 ) shows a narrow line width (40 km s 1 ) and is more extended and located farther away from the origin than is the low-velocity component (LVC; 100 km s 1 ), which has a broad, spatially varying line width and is located closer to the origin. We interpreted the broad line width of the LVC in terms of diverging stream lines over a large opening angle of ; the LVC may be a widely opened wind. Similar results were obtained for the young T Tauri star DGTau(Pyoetal.2003). In this article we present results of our continued [Fe ii] line spectroscopy of L1551 IRS 5 at the two bright knots PHK 1and PHK 2 (defined in Paper I) carried out with the slit positions perpendicular to the jets. We confirmed that the LVC has a spatial subcomponent that shows a large spatial width across the jets. 2. OBSERVATIONS AND DATA REDUCTION We observed the [ Fe ii] k1.644 m emission from the L1551 IRS 5 outflow on 2002 November 27 using the Infrared Camera and Spectrograph (IRCS; Tokunaga et al. 1998; Kobayashi et al. 2000) mounted on the Cassegrain focus of the Subaru Telescope atop Mauna Kea, Hawaii. Near-infrared H-band spectra were obtained with the IRCS echelle spectrograph equipped with a Raytheon 1024 ; 1024 InSb array with an Aladdin II multiplexer. The pixel scale was 0B060 0B006 pixel 1. The slit width was 0B6, corresponding to the velocity resolution of

2 818 PYO ET AL. Vol km s 1 (k=k 5000). The spectra were taken at PHK 1 and PHK 2 with a slit of 5B79 long placed perpendicular to the northern jet at the position angle (P.A.) of 166.Theseeing was 0B52 at the H band. The total on-source exposure time was 1200 s for each slit position. Sky spectra were observed at northeast (SKY) of the L1551 IRS 5 VLA sources. We observed the standard star HD (A0 V, T ea ¼ 9480 K, V ¼ 6:403 mag) for calibrating flux and removing telluric absorptions. The spectroscopic data were reduced with the IRAF packages as described in Paper I. 7 Narrowband imaging observations were made simultaneously with the H-band echelle spectroscopy with [Fe ii](k c ¼ 1:644 m, k ¼ 0:026 m) and H-cont (k c ¼ 1:573 m, k ¼ 0:020 m) filters by using the imaging section of IRCS as a near-infrared slit viewer. The slit viewer was equipped with a Raytheon 1024 ; 1024 InSb array with an Aladdin III multiplexer and had a pixel scale of 0B05825 pixel 1. The total on-source exposure times were 840 and 480 s for the [Fe ii] andh-cont filters, respectively. Each single exposure was 120 s. Image dithering was automatically achieved as we took the spectra of PHK 1, PHK 2, and SKY successively. We constructed a sky flat frame by combining the dithered frames with median filtering. With the IRAF packages, we did dark-current subtraction, flat-fielding with the normalized sky flat frame, sky subtraction, and bad pixels and cosmic-ray events elimination for each frame. The final object frames were then combined with median filtering, and the fluxes were scaled to the values for the unit exposure time. 3. RESULTS Figure 1a is a continuum-subtracted image of the [Fe ii] k1.644 m outflow from L1551 IRS 5 showing two wellseparated jetlike features (Fridlund & Liseau 1998; Itoh et al. 2000). The peak of the [Fe ii]emissionatphk1istakentobe Y ¼ 0 00, while the continuum peak is chosen to be X ¼ 0 00, with the direction of positive Y along P:A: ¼ 256. The northern jet is nearly on the Y axis, and the southern jet is seen at the left of it at Y k The slit apertures across PHK 1 and PHK 2 are marked by the rectangles. Also marked for reference is the slit aperture along the northern jet for which the [Fe ii] spectrum showninfigure1bwaspreviously obtained (Paper I). The extended low-brightness emission at both sides of PHK 1 may be attributed to the residual of the H-band continuum because its subtraction is not complete for image frames; the residual brightness is 5% of the continuum at (X ; Y ) ¼ ( 3 00 ; 0 00 ). Figure 1c shows a continuum-subtracted position-velocity diagram (PVD) of the [Fe ii] emission at PHK 1. There are two velocity components in the figure; their existence and relative strength are consistent with the previously obtained PVD along the northern jet (Fig. 1b). The low-velocity component (LVC) is stronger and has a peak velocity of V LSR ¼ 111 km s 1, while the high-velocity component (HVC) has a weaker peak at V LSR 279 km s 1. To examine their spatial extents along the slit, i.e., perpendicular to the outflow, we plotted in Figures 2a 2c the spatial profiles of the LVC, HVC, and continuum emissions averaged over the velocity ranges of 240 V LSR 20, 400 V LSR 240, and 200 V LSR 1000 km s 1, respectively. Also shown in the figures are the results of Gaussian fitting to the spatial profiles. 7 Details of the data reduction method for IRCS echelle spectra are available at IRCS_ and_reduction/ IRCS_ reduction_ html.html. Fig. 1. (a) Continuum-subtracted [ Fe ii] k1.644 m image of the L1551 IRS 5 after being smoothed by a Gaussian with G ¼ 2 pixels. Two horizontal slit apertures are shown across PHK 1 and PHK 2. Contours are shown from 3 to 183 with equal intervals in a logarithmic scale, where is the standard deviation of the background noise. The plus sign marks the midpoint of the two radio sources ( Rodríguez et al. 2003; Campbell et al. 1988). The filled circle at the lower right shows the seeing size of 0B52. (b) PVD previously obtained in Paper I along the vertical slit aperture shown in (a). The slit aperture in (a)was expanded so as to scale the proper motion of PHK 2 with respect to PHK 1 (0B25 0B15 yr 1 ), while the Y 0 axis in (b) measures the original angular distance when the data was taken. The dash-dotted horizontal line indicates the position of the L1551 IRS 5 VLA sources. The dash-dotted vertical line indicates the systemic velocity of L1551 IRS 5 (V LSR ¼þ6:2 kms 1 ;Momose et al. 1998). The ellipse at the lower right shows the velocity resolution and the seeing size of 59 km s 1 and 0B3, respectively. (c) Continuum-subtracted PVD perpendicular to the outflow at PHK 1. Contours are shown from 5 to 140 with equal intervals in a logarithmic scale, where ¼ 3 ; Wm 2 m 1. The dash-dotted horizontal line indicates the systemic velocity. The dotted line at V LSR ¼ 240 km s 1 divides the LVC and HVC. The ellipse at the lower left shows the velocity resolution and seeing size of 60 km s 1 and 0B52, respectively. The spatial widths of the emissions are broader than that of the standard star, which is plotted in Figure 2d. While the spatial profile of the HVC is well fitted by a single Gaussian with a FWHM of 0B78 0B28 (Fig. 2b), that of the LVC cannot be fitted by a single Gaussian because a spatially wide subcomponent (LVC wide )ispresent.figure2a shows that two Gaussian components give a satisfactory fit with the FWHMs of 0B83 0B14 and 2B84 0B47. Although a single Gaussian gives a moderately good fit to the spatial profile of the continuum emission at PHK 1, we show a better fit with the two Gaussian components CONT narrow and CONT wide in Figure 2c in order to clarify that even the wider spatial component CONT wide has the spatial width smaller than that of LVC wide. Figure 3a is the PVD for the slit position at PHK 2. It shows that the northern and southern jets are well separated from each

3 No. 2, 2005 [Fe ii] WIND FROM L1551 IRS Fig. 2. Spatial profiles of different radial velocity components at PHK 1 averaged over the following velocity ranges: (a)lvc( 240 V LSR 20 km s 1 ), (b) HVC( 400 V LSR 240 km s 1 ), and (c) continuum (200 V LSR 1000 km s 1 ). (d ) Normalized spatial profile of the standard star continuum averaged over the velocity range of 400 V LSR 1000 km s 1 with respect to the [ Fe ii] rest wavelength. The dotted lines show the fitted Gaussian components. The dash-dotted horizontal lines indicate the zero level of the intensity. other. No continuum emission was detected here. The northern jet around X ¼ 0 00 has two velocity components, while the southern jet at X ¼ 1B6 has only one velocity component with 100 km s 1. For the northern jet, we label the more blueshifted component as the HVC and the less blueshifted component as the pedestal and wing component (PWC) following the definition in Paper I, which reported that the relation between the less blueshifted component at PHK 2 and the LVC at PHK 1 is not certain (see Fig. 1b). We plotted in Figures 3b and c the spatial profiles of the PWC and southern jet, and HVC averaged over the velocity ranges of 220 V LSR 0and 350 V LSR 220 km s 1, respectively. The spatial profile of the southern jet is averaged over the same velocity range as the PWC. The results of Gaussian fitting to the spatial profiles of the PWC and southern jet, and HVC across PHK 2 are also shown. The spatial profiles of the HVC and southern jet are well fitted by single Gaussians with their FWHMs of 0B78 0B05 and 1B22 0B38, respectively, while asinglegaussianfittothespatialprofileofthepwcshows residuals near its peak and at X k 0B8. Note that the spatial widths of the HVC are the same at PHK 1 and PHK 2, suggesting that it is a highly collimated jet. Table 1 summarizes the line profile parameters obtained from the Gaussian fitting to the spatial profiles at PHK 1 and PHK 2. Fig. 3. (a) PVD for the slit position perpendicular to the northern jet at PHK 2, as marked in Fig. 1a. Contours are shown from 5 to 97 with equal intervals in a logarithmic scale, where ¼ 3 ; Wm 2 m 1. The dash-dotted horizontal line indicates the systemic velocity. The dotted boxes divide the three velocity components: the HVC, PWC, and southern jet. The ellipse at the lower left shows the velocity resolution and seeing size. (b) Spatial profiles of the PWC and southern jet averaged over 220 V LSR 0kms 1.(c) Spatial profile of the HVC at PHK 2 averaged over 350 V LSR 220 km s 1. For (b)and(c) the dotted lines show fitted Gaussian components and the dashdotted horizontal lines indicate the zero level of the intensity. 4. DISCUSSION The spatial width of the LVC wide is remarkably larger than those of the other emission components. Because the nearinfrared continuum at PHK 1 is known to be scattering emission from the deeply embedded VLA sources, the lack of a corresponding wide spatial component seen in the continuum profile in Figure 2c indicates that the LVC wide is not caused by the scattered [Fe ii] emission. There is no corresponding wide subcomponent seen in the spatial profile of the HVC in Figure 2b either. A marginal feature at 1 00 < X < 2 00 is seen in its spatial profile in Figure 2b. However, the absence of similar feature on the opposite side at 2 00 < X < 1 00 means that the marginal feature, even if it is real, does not correspond to the entire spatial extent of the LVC wide. Hence the LVC wide emission is not attributed to the scattering of the strong [Fe ii] emission around (X ; Y ; V LSR ) ¼ (0 00 ; 0 00 ; 100 km s 1 ). It should also be noted that the southern part ( 2 00 < X < 1 00 ) of the LVC wide is not part of the southern jet, as is clear from Figure 1a. The southern jet merges with the northern jet at PHK 1, probably with a velocity similar to that of the LVC as expected from its velocity at PHK 2 (Fig. 3a), and cannot be discerned as an independent spatial or velocity component at all. We must note, however, that the two widths of the LVC wide and CONT wide overlap at the 2 uncertainties of their widths. This gives the probability of 10% for the two widths being actually the same. We thus have a confidence level of 90% to conclude that the LVC wide emission is not caused by the scattering that is responsible for the CONT wide.wealsonotethat

4 820 PYO ET AL. Vol. 618 TABLE 1 [Fe ii] Line Parameters at PHK 1 and PHK 2 Parameter Peak Position X (arcsec) Spatial Width FWHM (arcsec) Peak Velocity V LSR (km s 1 ) At PHK 1 HVC LVC narrow LVC wide CONT narrow CONT wide At PHK 2 HVC PWC a Southern jet a The single-gaussian fit shows residuals at X k 0B8 andx the baselines to define the spatial and spectral profiles are welldefined, flat and stable, so that even the small excess seen in the spatial profile of the LVC at X P 2B5 in Figure 2a is real. This can be understood when we examine the continuum profile in Figure 2c at the positions jx j k 2 00, where no sign of emission exits on a well-defined flat baseline. The difficulties in attributing the large spatial width of the LVC wide to scattering suggest that there should be [Fe ii] emitting gas actually flowing over the wide spatial width. The spatially wide outflow has a width of 2B84 0B47 (FWHM) at a distance of 1B2 from the driving sources, suggesting that the outflow has an opening angle of 100.InPaperIweobtained a smaller, although still large, opening angle of for the spatially narrow subcomponent of the LVC (LVC narrow ) 8 by interpreting its large velocity width in terms of diverging stream lines. The existence of a large opening angle outflow suggests that it fills the space between the highly collimated outflow such as the HVC and the shell of the CO outflow. Such an outflow with a wide opening angle will be effective in sweeping up envelope material from the close vicinity of its driving source. It is important to note that no systematic difference was found in the radial velocities between the LVC wide and LVC narrow ;the overall feature of the LVC emission in Figure 1c is symmetrical with respect to the common centroid velocity at V LSR 110 km s 1. This result may rule out the possibility that the LVC wide is the emission arising from the bow shock wing regions trailing behind an internal working surface in the outflow responsible for LVC narrow. Because in such a case we would expect a curved feature on the PVD such that the radial velocity is largest for LVC narrow and systematically decreases toward both edges of the LVC wide (Mundt et al. 1987). In other words, the transverse spatial extent is larger at lower radial velocities (see, e.g., Fig. 5 of Raga et al. 2001). We may also reject the possibility that the LVC wide originates in the gas entrained and accelerated by the LVC narrow, because in this case we would again expect that the LVC wide would have a lower radial velocity than the LVC narrow. These arguments, together with the similarity in their large opening angles, support the interpretation that the 8 The LVC referred to in Paper I corresponds to LVC narrow in this paper. LVC wide and LVC narrow are different parts of a single lowvelocity outflow: the LVC wide represents its outermost part while the LVC narrow arises mainly from its inner part, which has a smaller opening angle. If this is the case, then both LVC wide and LVC narrow are associated with the northern jet because the LVC narrow is most probably associated with the northern jet as was discussed in Paper I. The issue of the LVC wide and LVC narrow having a similar radial velocity requires further study. This suggests that the outflow gases with wide and narrow opening angles both have a similar Alfvén radius if they are magnetocentrifugally launched, because the Alfvén radius is the dominant factor in determining the outflow terminal velocity (see e.g., Matt et al. 2003; Bacciotti et al. 2002; Königl & Pudritz 2000). All the LVC gas filling the gap between the highly collimated HVC and the poorly collimated shell of the CO outflow may be accelerated at a similar radius of an accretion disk. 5. SUMMARY We presented [ Fe ii] k1.644 m emission spectra along the slits across PHK 1 and PHK 2 perpendicular to the outflow from L1551 IRS 5. We have two main conclusions: 1. At PHK 1 the low radial velocity component at V LSR 110 km s 1 shows two spatial subcomponents with widths of 0B83 and 2B84 ( FWHM). The wide subcomponent has an opening angle of 100. The gas responsible for this emission probably fills the space between the northern jet and the shell of the CO molecular outflow, and it may be sweeping up the envelope material around the embedded protostars. 2. At PHK 2 the northern jet shows two radial velocity components at V LSR 270 and 140 km s 1. The high-velocity component has the same spatial width of 0B78 at both PHK 1 and PHK 2, indicating that it is a highly collimated outflow. We thank Michihiro Takami for his critical comments and inspiring discussion. We are grateful to the entire staff of the Subaru Telescope for their dedicated support of the telescope and observatory operations. M. Tsujimoto is supported by the Japan Society for the Promotion of Science.

5 No. 2, 2005 [Fe ii] WIND FROM L1551 IRS Bacciotti, F., & Eislöffel, J. 1999, A&A, 342, 717 Bacciotti, F., Ray, T. P., Mundt, R., Eislöffel, J., & Solf, J. 2002, ApJ, 576, 222 Campbell, B., Persson, S. E., Strom, S. E., & Grasdalen, G. L. 1988, AJ, 95, 1173 Davis, C. J., Mundt, R., Eislöffel, J., & Ray, T. P. 1995, AJ, 110, 766 Edwards, S., Ray, T., & Mundt, R. 1993, in Protostars and Planets III, ed. E. H. Levy & J. I. Lunine ( Tucson: Univ. Arizona Press), 567 Fridlund, C. V., & Liseau, R. 1998, ApJ, 499, L75 Giovanardi, C., Lizano, S., Natta, A., Evans, N. J., II, & Heiles, C. 1992, ApJ, 397, 214 Giovanardi, C., Rodríguez, L. F., Lizano, S., & Cantó, J. 2000, ApJ, 538, 728 Graham, J. A., & Heyer, M. H. 1990, PASP, 102, 972 Hartigan, P., Morse, J., & Raymond, J. 1994, ApJ, 436, 125 Itoh, Y., et al. 2000, PASJ, 52, 81 Kobayashi, N., et al. 2000, Proc. SPIE, 4008, 1056 Königl, A., & Pudritz, R. 2000, in Protostars and Planets IV, ed. V. Mannings, A. P. Boss, & S. S. Russel ( Tucson: Univ. Arizona Press), 759 Lizano, S., Heiles, C., Rodríguez, L. F., Koo, B.-C., Shu, F. H., Hasegawa, T., Hayashi, S., & Mirabel, I. F. 1988, ApJ, 328, 763 Matt, S., Winglee, R., & Böhm, K.-H. 2003, MNRAS, 345, 660 REFERENCES Momose, M., Ohashi, N., Kawabe, R., Nakano, T., & Hayashi, M. 1998, ApJ, 504, 314 Moriarty-Schieven, G. H., & Snell, R. L. 1988, ApJ, 332, 364 Mundt, R., Brugel, E. W., & Bührke, T. 1987, ApJ, 319, 275 Mundt, R., Stocke, J., Strom, S. E., Strom, K. M., & Anderson, E. R. 1985, ApJ, 297, L41 Pyo, T.-S., Hayashi, M., Kobayashi, N., Terada, H., Goto, M., Yamashita, T., Tokunaga, A. T., & Itoh, Y. 2002, ApJ, 570, 724 ( Paper I ) Pyo, T.-S., et al. 2003, ApJ, 590, 340 Raga, A. C., Binette, L., & Canto, J. 1990, ApJ, 360, 612 Raga, A., Cabrit, S., Dougados, C., & Lavalley, C. 2001, A&A, 367, 959 Rodríguez, L. F., Curiel, S., Cantó, J., Loinard, L., Raga, A. C., & Torrelles, J. M. 2003, ApJ, 583, 330 Shu, F. H., Ruden, S. P., Lada, C. J., & Lizano, S. 1991, ApJ, 370, L31 Snell, R. L., Loren, R. B., & Plambeck, R. L. 1980, ApJ, 239, L17 Stocke, J. T., Hartigan, P. M., Strom, S. E., Strom, K. M., Anderson, E. R., Hartmann, L. W., & Kenyon, S. J. 1988, ApJS, 68, 229 Tokunaga, A. T., et al. 1998, Proc. SPIE, 3354, 512 Uchida, Y., Kaifu, N., Shibata, K., Hayashi, S. S., Hasegawa, T., & Hamatake, H. 1987, PASJ, 39, 907