原恆星系統L1551 IRS5及L1551 NE之分子雙極噴流

(1)

國立臺灣大學理學院天文物理研究所碩士論文

Graduate Institute of Astrophysics College of Science

National Taiwan University Master Thesis

原恆星系統 L1551 IRS5 及 L1551 NE 之分子雙極噴流 Bipolar Molecular Outflows in

L1551 IRS5 and L1551 NE Protostellar Systems 吳柏鋒

Wu Po-Feng 指導教授：林仁良博士高桑繁久博士賀曾樸博士 Advisor: Jeremy Lim, Ph.D.

Shigehisa Takakuwa, Ph.D.

Paul T. P. Ho, Ph.D.

中華民國 98 年 7 月

(2)

致謝

這篇碩士論文主要在中研院的林仁良博士(Dr. Jeremy Lim)與高桑繁久博士 (Dr. Takakuwa Shigehisa)指導下完成。

首先要感謝 Jermey，你是一個好老師。這兩年我沒有花太多的時間迷航，

是你帶領我往正確的大方向前進，並教導我如何用宏觀的角度思考科學問題。謝謝 Shige，我用掉了你很多時間。多少個下午，我們在你的辦公室討論各種問題，

從你那我得到了許多實用的意見。謝謝中研院天文所賀曾樸主任，從我的論文出發，你提醒了我什麼才是探索這個宇宙的正確態度。而這份研究的背後還有一個強大的顧問團支持—所有中研院天文所的研究員們。對於任何層次的科學問題或技術上的麻煩，你們總是能提供有用的支援。

謝謝 820 的住戶們，有你們在的辦公室總是充滿笑聲，踏進辦公室成為一件開心的事。因為如此，我才能持續工作一天，一天，又一天。也謝謝所有中研院天文所的研究助理和學生，我喜歡每一次與你們的互動。無論是科學上的討論，

日常閒聊，亦或是關起門來偷偷說三道四，每一件小事都令人開心。而我的大學同學們，你們也是不可或缺的一部分。每次見面的笑靨都是最有用的補給品。知道大家都過得很好，各自在人生的道路上努力，讓我能開心地接下一次又一次的全新挑戰。

最後，最重要的是我的家人。即使你們完全搞不清楚我究竟在做什麼研究，

也不同意我的選擇，你們仍然放手讓我走自己的路，並且無條件地給予我一切所需的援助。如果我有任何小小的成就，那不屬於我，一切都屬於你們。

(3)

Acknowledgment

This thesis work is finished under the supervision of Dr. Jeremy Lim and Dr. Takakuwa Shigehisa. I would like to thank Jeremy, who helps me with developing my ability to do research, teaches me how to think as a scientist and leads me go throught the process from playing with the data to publishing a paper. Thanks to Shige for taking care of my detail problems. I spent a lot of your time, sitting in your office and asking any kinds of questions and receiving practical opinions from you. I also would like to thank Dr.

Pual Ho, who gives me crucial comments on my work and reminds me of the right attitude for doing science. This work is also supported by a strong consultant group – all IAA researchers. You always kindly provide your helps at all levels, give me your precious opinions on science questions and technical problems.

Thanks to the residents of room 820. You fill this office with happiness.

Because of you, stepping into this office and working whole day becomes a pleasure. Also for all IAA studuents and RAs, I really delight in our interactions, including academic or non-academic, sharing our science, chatting or gossiping. For my undergraduate classmates in physics department, even you are not directly related to the thesis work, I still want you to know that you are also an important part of it. Your laughter always energizes me for taking the coming challenges.

A special thank is for my family. Even you have no ideal on what I am doing and disagree the road I choose, you still allow me go my way and unconditionally provide your supports. All my achievements belong to you.

(4)

摘要

L1551 分子雲為一低質量恆星形成區域，且為距太陽系最近的恆星形成區域之一。觀測上顯示，此區域中存在許多不同演化階段的原恆星系統，意味著長時間以來，此區域皆為一活躍的恆星形成區域。我們使用次毫米微波陣列(SubMillimeter Array, SMA)觀測 L1551 分子雲中兩個原恆星系統—L1551 IRS5 及 L1551 NE—周圍區域的一氧化碳分子譜線，以探測該原恆星分子雙極噴流的物理性質。

L1551 IRS5 是 L1551 分子雲中輻射熱光度最強的多重原恆星系統，且擁有巨大的分子雙極噴流。其大尺度分子雙極噴流呈東北(紅移端)—西南(藍移端)方向，各自向外延伸約 1.5 秒差距。我們藉由角分辨率約4 角秒(約 560 天文單位) 的 CO(2—1)分子譜線影像探測此噴流之中心約4000 天文單位的區域。在此中心區域，我們分辨出三個不同噴流結構：(1) X 形結構，自中心向外延伸約20 角秒，其方位角與大尺度噴流相近，且有相同之紅藍移方向。此 X 形結構為中央高速噴流所開之錐狀空腔之臨邊增亮現象。(2) S 形結構，自中心向外延伸約 10 角秒，其方位角與大尺度噴流相近，但紅藍移方向相反。此 S 形結構為新發現的分子雙極噴流。S 形結構以及視線方向速度的變化顯示此噴流正在進動。此噴流可能是被此系統中一新發現的原恆星所驅使，且其環星盤與鄰近的原恆星有強烈潮汐作用而使環星盤進動。

(3) 中央緊密結構，自中心向外延伸約 1.4 角秒，其方位角與大尺度噴流相近，且有相同之紅藍移方向。此結構為原恆星系統中心近期受中央高速噴流所驅使的分子氣體。

L1551 NE 是 L1551 分子雲中輻射熱光度次強的多重原恆星系統，位於 L1551 IRS5 東北方2.5 角分處，為 L1551 IRS5 的巨大分子雙極噴流影響。L1551 NE 所驅使的分子雙極噴流尚未被確認。我們利用角分辨率約0.8 角秒(約 110 天文單位)的 CO(3—2)分子譜線影像探測此系統周圍較溫暖之氣體，首次成功拍攝下此系統分子雙極噴流的電波波段影像。此分子雙極噴流自中心延伸僅約8 角秒，且其動力學年齡極短，顯示此為一年輕的原恆星系統，但其質量明顯小於其他位於同樣演化階段的分子雙極噴流。此現象指出，L1551 NE 鄰近區域的氣體分子已被L1551 IRS5 的強烈噴流掃除。另外，我們的一氧化碳分子譜線影像與紅外光觀測中之扇形反射星雲吻合，但卻未發現在紅外光波段同樣出現的束狀噴流。此現象暗示 L1551 NE 系統中的兩原恆星處在不同演化階段，驅使扇形反射星雲的原恆星較早演化或演化較快，其噴流將鄰近區域之分子氣體掃除。較晚演化的原恆星之束狀噴流驅使之分子噴流

(5)

Abstract

L1551 molecular cloud is located at a distance of ∼ 140 pc, in the constellation of Taurus, and is one of the nearest active low-mass star formation region. We used the Submillimeter Array (SMA) to image the CO molecular outflows from two multiple protostellar systems in L1551 region, L1551 IRS5 and L1551 NE, at an arcsecond angular resolution. We observed the CO(2–1) line emission in L1551 IRS5 system. First, we imaged the base of the outflow cavity and found signatures for multiple outflow events. We also found newly entrained gas in the cavity, indicating that the ambient gas is still falling into the cavity. Second, we discovered a precessing outflow, which has an opposite inclination to it of L1551 IRS5 large-scale outflow. This newly-found component is likely driven by the recently proposed candidate third protostar in L1551 IRS5 system. For the L1551 NE binary system, we shows a clear bipolar structure of its molecular outflow at the first time. We saw a cavity-like structure corresponding to the reflection nebula at infrered wavelength, but no jet-like structure, which is also seen at infrared. It is likely that, one of the binary launched the outflow first and cleaned ambient gas away, thus the outflow from another protostar entrained less gas and is unable to be detected. The mass of L1551 NE outflow and the dusty envelope are both by orders of magnitude smaller than those of other Class 0 protostars, suggesting the ambient gas has been swept up by the strong outflow from L1551 IRS5.

(6)

List of Figures

1.1 Star-formation processes for single low-mass stars . . . 3

1.2 Spectral energy distributions (SED) of protostars in different classes . . . 5

1.3 Wind–driven model . . . 8

1.4 Position-Velocity diagram of VLA 05487 outflow . . . 9

1.5 Bow–shock model . . . 11

1.6 Molecular outflow of HH 211 . . . 12

1.7 Evolution of outflow opening-angles . . . 15

1.8 Precessing outflow from Cep E . . . 16

2.1 L1551 star-forming region . . . 20

2.2 Spectral energy distribution of L1551 IRS5 . . . 21

2.3 CO outflow of L1551 IRS5 . . . 22

2.4 Gas envelope, circumbinary disk and circumstellar disks of L1551 IRS5 . . . 24

2.5 Duplicity of L1551 IRS5 . . . 25

2.6 Near infrared image of L1551 NE system . . . 27

2.7 Multiplicity of L1551 NE . . . 28

2.8 JCMT image of CO gas surrounding L1551 NE . . . 30 2.9 Single-dish and interferometry CS images of L1551 NE system 31

(9)

3.1 Bent jet model . . . 37

4.1 0.8 mm continuum image of L1551 NE . . . 43

4.2 Model fitting of L1551 NE 0.8 mm image, one disk plus one Gaussian . . . 45

4.3 Model fitting of L1551 NE 0.8 mm image, one disk plus two Gaussians . . . 46

4.4 CO(3–2) channel maps of L1551 NE . . . 48

4.4 Continue . . . 49

4.5 CO(3–2) integrated intensity map of L1551 NE . . . 50

4.6 CO(3–2) spectra of L1551 NE from JCMT and SMA . . . 52

4.7 Mass of dusty envelopes of Class 0 and Class I protostars . . . 56

(10)

List of Tables

4.1 Observation parameters of L1551 NE SMA CO(3–2) observation 41 4.2 Physical properties of CO outflow from L1551 NE . . . 53

(11)

Chapter 1 Star Formation

Stars are the building rocks of the visible sky. The process by which stars from is thus one of the basic questions of Astronomy. Stars are self-gravitating gaseous balls whose central temperatures are high enough for nucleus fusion reactions. Their masses can be as high as several tens of, maybe over 100 solar masses, or only a few tenths of solar mass, just at the margin that hydrogen fusion can occur.

1.1 Star Formation Environment

Molecular clouds are the principle sites of star formation. Typically clouds have sizes of several tens of parsecs, and contain masses of 10⁵ to 3× 10⁶ M⊙. These giant clouds have complex substructures named “clumps”. Clumps have relatively higher densities of nH2 ∼ 10^2.5 cm⁻³, sizes of a few parsecs, and contain 10³ ∼ 10⁴ M⊙. Subregions in clumps, called —cores —, have even higher densities of nH2 > 10⁴ cm⁻³. These sites are believed to be the birthplaces of stars, because they are tightly associated with known T Tauri stars and infrared sources (Beichman et al., 1986).

(12)

1.2 Star Formation Processes

Star formation processes have proposed a generally accepted framework for low-mass stars form Shu et al. (1987). During the early stages, a dense core in the molecular cloud contracts slowly instead of collapsing violently (see Figure 1.1a).

The magnetic force resists the self-gravity of the core. Here the electrons and ions are species which sense the magnetic field and couple with magnetic field lines, while neutrals sense gravity and drift across magentic field lines.

The neutrals move relative to charged particles and the frictional drag force prevents the neutrals directly fall into the center, slows down the contration speed. At some point, as the core gradually loses it magnetic support, the central concentration will cause a gravitational collapse. The gas in the molecular cloud with sufficiently large specific angular momentum first lands on the equatorial plan, forms a disk, then transfer to the central star. This is the main accretion phase of a protostar.

During the accretion, material is also ejected out from the star-disk system. It is now generally agreed that the outflowing material is launched centrifugally along magnetic filed lines, although the preceise mechanism is still under debate. Material is ejected out in the direction of the rotational axis, leads to bipolar outflows (see Figure 1.1c).

As time proceed, the outflow opens a wider angle and sweeps up the outside envelope, cease the mass accretion. The system in this stage remains a central star and a disk but not the envelope (see Figure 1.1d). This is observed as a T Tauri star. Finally, the disk becomes the material reservior for planet formation or disperses into space. The system goes into a pre- main-sequence stage.

(13)

Figure 1.1: Four stages of star formation for single low-mass stars. (a) Mag- netic supported cores within molecular clouds slowly contract. (b) A protostar with a surrounding disk collapses rapidly. This is the main accretion phase. (c) An outflow emerges, creates bipolar outflow, sweeps up ambient materials. (d) Acrretion stops. The system remains a newly formed star with a circumstellar disk. (Shu et al., 1987).

(14)

1.3 Classification of Protostars

Young stellar objects are obseved to exihibit excess radiation at infrared wavelengths, that is, more flux than blackbody spectrum. This infrared excess radiation is due to circumstellar material, which absorb stellar photons and reemit at infrared wavelengths. The spectral index αIR of the radiation is defined as

αIR ≡ d log(λFλ)

d log λ (1.1)

It is conventional to evaluate the αIR by differencing the flux between 2.2µm and 10 µm. By varying the properties of stars and those of the circumstellar material, theoretical models can successfully reproduce the infrared excess in many observed sources (Adams et al., 1987). Therefore, the spectral index reflects the basic properties of the protostellar system and is used to cate- gorized protostars into different classes, which are associated with different evolutionary stages.

The typical spectral enery distribution at different evolution stages are shown in Figure 1.2. Class I sources have characterized by αIR > 0. They have only remnant infalling envelopes, and have accumulated a large fraction of their final stellar mass (Andr´e & Montemerle, 1994; Andr´e, 1995). Class II protostars, or classical T Tauri stars, have infrared spectral index −1.5 <

αIR < 0. They are at a more evolved stage, contain circumstellar disks, but no infalling envelopes. Those whose spectral index α < −1.5 are Class III protostars, or weak–lined T Tauri stars, which have very weak or no disks.

Later a new class, Class 0, is purposed. Class 0 sources are the youngest protostars, still highly self-embedded in surrounding molecular clouds, and in their main accretion phase. Observationally, Class 0 protostars do not have emission at wavelength longer than 10µm, but high submillimeter luminosity

(15)

Figure 1.2: Spectral energy distribution (SED) and schematic figures for young stellar objects in different classes. Different SEDs are resulted from the structures of systems.

(Andr´e et al., 1993).

1.4 Molecular Outflows in Protostellar Sys- tems

Bipolar outflows is a common and essential ingredient in the star formation processes. The outflowing material is observable over a wide range of wavelengths. The first signature of outflow activities are Herbig–Haro

(16)

(HH) objects, which are small knots with optical line emission. HH objects were later recognized as hot, ionized gas, heated up by the collision between high speed outflowing material and ambient material (Schwartz, 1975). Subsequent proper motion measurement and detection of collimated wind confirmed that the wind emanants from the vicinity of young stellar objects (Cudworth & Herbig, 1979; Mundt & Fried, 1983). Signatures at other wavelengths including centimeter free-free continuum of ionized gas, infrared continuum of warm gas and emission lines. When the collimated outflow interacts with ambient molecular cloud, it sets the gas into motion, gives rise to the molecular outflow.

1.4.1 Outflow Launching Mechanisms

Outflow is regarded as one of the key processes druing the early evolutionary stage which carries away excess angular momentum of the inner region of the system thus allowing accretion. However, the detail launching mechanisms are still poorly understood. So far, two most promising models are X-wind model and disk-wind model.

The X-wind model suggests that material is accelerated outwards from the zone where the stellar magnetosphere co-rotates with the disk. This region lies typically a few stellar radii from the sources, and is called the X-point (Shu et al., 1988, 1994). The disk-wind model proposed that winds are launched from the rotating, magnetized disks. It was first suggested as the origin of jets from accretion disks around black holes and soon proposed as the mechanism for protostellar jets (Pudritz & Norman, 1983, 1986).

(17)

1.4.2 Models for Molecular outflows

Molecular outflow is believe to be entrained by the ejected matter to to velocities higher than those of the cloud. Several models are proposed to explain how the ambient is entrained. Two most succesful models in explaining observed molecular characteristics are: (1) wide-angle magnetized winds, (2) jet-driven bow shocks.

Wide-angle Magnetized winds

The first hypothesis asserts that the bipolar molecular outflows are swept up by the wide-angle magnetized wind, as illustrated in Figure 1.3. The ambient gas is set into motion by acquiring momentum from the wind. If the wind has an axial density gradient, the wind can appear as a collimated jet (Shu et al., 1991, 1995). The wide-angle wind model can produce a molecular outflow with wide-opening angle and performs well in explaining the kinematics in several outflow systems (Lee et al., 2000). For example, Figure 1.4 shows the comparison between observation and model calculation of the kinematics and spatial structure of the protostellar system VLA 05487. Both the PV cut along the jet axis and the total integrated intensity map in gray scale show the V-shaped structure. Contours from model caculations can reproduce the overall structure.

Jet Driven Bow Shocks

The other proposed mechanism of molecular outflows is that they are the shocked gas when high–velocity jets propagate into the ambient gas. When a supersonic jet emanantes toward the surrounding envelope, large amount of gas is trapped in the working surface, then flows sideways. This gas

(18)

Figure 1.3: Schematic diagram of wind–driven model. The inclination i is the angle between observer. In the scenario, the magnitized wind transfer its momentum to ambient gas, which is observed as bipolar molecular outflows.

(19)

Figure 1.4: Comparison between wind-driven model and observation for VLA 05487. The gray-scale images are the observations and the contours are from the calculations for wind-driven model. The left panel shows the PV diagram cut along the jet axis. The right panel shows the integrated intensity map (Lee et al., 2000).

(20)

flow backwards in the reference frame of the jet, as illustrated in Figure 1.5.

The bow-shocked layer reexpands after the working surface passes by, which appears as the molecular outflow we see (Raga & Cabrit, 1993). The bow- shock model is appealing because it explains several observed properties of some molecular outflows. For example, the CO outflow of a Class 0 protostar, HH 211, exhibits a collimated, higher velocity component and a wider cavity- like, lower velocity component. The overall structure fits into the picture of a jet-driven model. The shaped of lower velocity cavity, shown by the contours in Figure 1.6, can be reproduced well by the solid curve, which is generated from a simple model of jet-driven flow.

(21)

Figure 1.5: Sketch of the structure of the bow–shock. The velocities are in the reference from of shock . A densed, collimated jet is ejected into the ambient medium, creates a bow–shock (Gueth & Guilloteau, 1999).

(22)

Figure 1.6: Comparison between bow–shock model and observation of HH 211. The contours are 230 GHz continuum emission (centered at (0,0)) and low velocity CO(2–1) emission. Thick line is the model calculation of the cavity (Gueth & Guilloteau, 1999).

(23)

1.4.3 Physical Properties of CO Outflow

CO is the second abundent molecule amoung interstellar gas species, only next to molecular hydrogen. It has simple line structure, thus simple rotational energy levels, that can be easily observed in millimeter and sub- millimete wavelengths. Outflow properties can be obtained by converting observed CO lines into physical gas parameters. Measuring the line ratio between two rotational lines is often used to derive the gas kinetic temperature.

As for the optical depth, measuments of lines from different isotopes can give a good estimate. Having both excitation temperature and optical depth, it is possible to derive the CO gas column density by using equations of radiative transfer. The mass of outflowing gas can be calculate by assuming a CO abundence. Moreover, the mometum and energy of outflowing gas can also be estimated from obtaining the mass distribution and line-of-sight velocity.

Molecular outflows driven by low–mass protostars (< 1M⊙) have been extensively studied and show some basic common features. The most char- acteristic property of outflows is their bipolarity. Two separated lobes of gas, one redshifted and the other blueshifted, lie next to each other, with a young stellar object in between. The length of outflows is typically 0.1 to a few pc, with a wide range of collimation. The collimation factor is defined by major radius divided by minor radius. Typical values obtained from single-dish observations are 3–20, while interferometric maps tend to give higher numbers, that is more collimated, due to its better angular resolution. The mass of molecular outflows also ranges over a few orders of magnitudes, from ∼10⁻² M⊙ to as high as > 10 M⊙. The momentum supply rate is about 10⁻⁵–10⁻⁴ M⊙ km s⁻¹ yr⁻¹ (Bontemps et al., 1996; Arce et al., 2005a).

(24)

Outflow Opening Angle

The outflow opening angle is often considered as a signature of the age of the driving source. Arce & Sargent (2006) showed a clear trend that the outflow opening angle becomes larger as the central star evolves, based on their survey of molecular outflow within . 10⁴ AU of nine protostellar systems in different evolutionary stages, together with previous observations. Figure 1.7 plots the outflow opening angles versus source ages. Data are from Arce &

Sargent (2006) and literatures therein (see Table 9 of Arce & Sargent, 2006, for the reference list). Data points populate in the upper-right and lower- left parts of the diagram, mean that older protostars posses outflows with larger opening angle. Class 0 protostars exhibit relatively collimated jet-like outflow or cone-shaped lobes with opening angles < 55^◦, where as Class I protostars have outflow lobes with wider opening angles of > 75^◦(with ∼125^◦ the largest opening angle seen).

Outflow Precession

Observations with higher angular resolution revealed bending or winding structures or misalignment between two lobes of the outflow, for example, Cep E in Figure 1.8 (Eisl¨offel et al., 1996), and L1557 (Gueth et al., 1996).

The bending structure indicates the ejection direction changes over time.

However, the reason for the changing ejection direction has not been well established yet.

Multiple Outflow

Several outflow systems have quadropolar morphology, that is, four lobes are observed and seem to be driven by the same condensation. The quadropolar outflows are an indication for the multiplicity of the central protostellar sys-

(25)

Figure 1.7: Outflow opening angle versus source age (Arce & Sargent, 2006).

The ages are estimated from bolometric temperatures. Circles are from observations in Arce & Sargent (2006), triangles are from literatures therein.

In average, the outflow opening angles of older sources are larger than those of younger sources.

(26)

Figure 1.8: Precession model (black curve) superposed on molecular hydrogen 1–0S(1) line emission of Cep E. (Eisl¨offel et al., 1996). In this case, the outflow precesses at 4^◦.

(27)

tem, without directly resolving the driving sources (see Gueth et al., 2001;

Lee et al., 2002, for examples).

(28)

Chapter 2 INTRODUCTION TO L1551 STAR–FORMING REGION

2.1 L1551 Dark Cloud

The L1551 dark cloud (Lynds, 1962) is one of the nearest and therefore best studied regions of low-mass star formation. It lies in the constellation Taurus at a distance has been measured to be 140±10 pc (Kenyon et al., 1994), and has diameter of ∼20^′ (∼1 pc). Surveys at different wavelengths have found numerous signatures of star-formation: infrared sources (Storm et al., 1976;

G˚alfalk et al., 2004), Herbig–Haro (HH) objects (Garnavich et al., 1992;

Devine et al., 1999), young T Tauri stars (Brice˜no et al., 1998), infrared reflection nebula (Hayashi & Pyo, 2008, Figure 2.1), and bipolar molecular outflows (Pound & Bally, 1991; Moriarty-Schieven & Wannier, 1991).

Observations in different CO lines have revealed the overall morphology and kinematics of outflows in the L1551 region on parsec scales and nearly all outflows have been identified with their driving sources. Several star–formation sites have been identified, such as L1551 IRS5, L1551 NE, HH 30, HL Tau,

(29)

XZ Tau that concentrate in the densest parts of the cloud, and more than 30 classical and weak T Tauri stars are distributed in the cloud (G˚alfalk et al., 2004). Both younger (Class 0 and Class I) objects, such as L1551 NE and L1551 IRS5, and more evolved classical and weak–line T Tauri stars (Class II and Class III) are present, indicating that star formation activities has lasted at least for a few million years (Moriarty-Schieven et al., 2006).

2.2 L1551 IRS5 Protostellar System

L1551 IRS5 was first detected as an infrared survey in the L1551 cloud (Storm et al., 1976), it exhibits a luminosity of ∼ 32L⊙, obtained by summing up luminosities measured from 1.2 µm to 160 µm (Cohen et al., 1984), and is the brightest source in L1551 star-forming region.

Althought L1551 IRS5 is prominent at infrared wavelength, it is invisible in optical. Modeling of its infrared spectral energy distribution (SED), shown in Figure 2.2 suggests it belongs to Class I evolutionary stage (Adams et al., 1987).

L1551 IRS5 is the first protostar from which a molecular bipolar outflow, detected in¹²CO(2–1), was first recognized (Snell et al., 1980). A more recent

12CO(1–0) image is shown in Figure 2.3, taken by Stojimirovi´c et al. (2006) with James Clerk Maxwell Telescope (JCMT). Its outflow extends ∼ 1.5 pc towards the NE (redshifted lobe) and SW (blueshifted lobe) directions at a position angle of ∼ 50^◦. Each outflow lobe has a U-shaped structure, together with knotty components having distinct velocities, suggestive of multiple ejection events (interpretation of Bachillar et al., 1994).

Apart from the large-scale bipolar molecular outflow, L1551 IRS5 is surrounded by a compact molecular condensation. This condensation is elon-

(30)

Figure 2.1: Composite color image of L1551 star–forming region taken by Subaru. The blue, green and red colors are assigned to the broad-band filter H, narrow-band filter [Fe II] λ 1.644 µm and narrow-band filter H₂, respectively. (Hayashi & Pyo, 2008)

(31)

Figure 2.2: Spectral energy distribution (SED) of L1551 IRS5, Figure 1g in Adams et al. (1987). This plot shows log(νLν) vs. log(ν) in cgs units. Solid and dashed curves are theoretical models used in Adams et al. (1987).

(32)

Figure 2.3: CO(1–0) integrated intensity map overlaid on Hα data, (Sto- jimirovi´c et al., 2006). The outflow associated with L1551 IRS5 is in NE(red)–

SW(blue) direction. Another redshifted red lobe extends to the west, which its driving source is still unclear.

(33)

gated in the NW–SE direction, perpendicular to the major axis of the large–

scale bipolar molecular outflow. Figure 2.4a shows the C¹⁸O image of the condensation taken by Momose et al. (1998). It has a linear size of ∼2400 AU in the direction of major axis. The kinematics of the condensation indicates it is a flattened envelope which is contracting and rotating (Ohashi et al., 1996; Saito et al., 1996; Momose et al., 1998). In the inner, warmer region of the envelope, traced by high rotational transition lines of CS, rotation dominates over infalling motion (Takakuwa et al., 2004, see Figure 2.4b).

The first indication of its duplicity came from Bieging & Cohen (1985), who discovered two sources in L1551 IRS5, separated by ∼0.3^′′(∼42 AU) at 2 cm in north-south direction (Figure 2.5a). Looney et al. (1997) showed observations in the 2.7 mm continuum that L1551 IRS5 contains two circumstellar dust disks surrouded by a possible circumbinary dust disk and/or extended envelope. Rodr´ıguez et al. (1998) spatially resolved the two circumstellar dust disks for the first time at 7 mm. These two protostellar components were shown to have the same proper motion through space except for a much smaller component of motion attributed to their orbital motion about each other (Rodr´ıguez et al., 2003a), and each drives a protosteller jet (Rodr´ıguez et al., 2003b). Lim & Takakuwa (2006) spatially resolved the two circumstellar disk along both major and minor axes, and showed that these disks are aligned with each other as well their surrounding rotating and flattened pseudodisk. The major axis of each disk is perpendicular to the ionize jet.

They also showed that the two disks orbit each other in the same direction as the rotation of the pseudodisk. This feature may indicated the two disks form throught a fragementation of parent rotating cloud. Finally, a candidate third protostellar component is detected, located near the northern protostellar component (Figure 2.4c). The circumstellar disk of this third

(34)

Figure 2.4: (a) Intensity-weighted mean velocity map of L1551 IRS5 obtained from the C¹⁸O emission, adapted from Momose et al. (1998). The C¹⁸O emission is elongated in the direction perpendicular to the outflow axis, indicating it comprises a flattend circumstellar envelope around L1551 IRS5. The velocity pattern suggests both infall and rotation appears in the central part. (b) CS(7–6) emission (blue and red contours) superposed on the 343 GHz continuum emission (gray scale) and VLA 3.5 cm radio jet image (white contours). The CS(7–6) line traces the warmer and denser gas in the inner part of the envelope. The NW-SE velocity gradient is perpendicular to the jet axis, suggests rotation dominates in the inner envelope. (c) 7 mm image of L1551 IRS5, adapted from Lim

& Takakuwa (2006). The two main protostars orbit with each other clockwise, same direction as that of the inner gas envelope. These features may indicate that the two main protostars forms from a disk fragmentation process. In addition, a smaller third condensation is resolve in this map.

component is misaligned relative to the circumstellar disks of the two main components as well as their surrounding pseudodisk.

2.3 L1551 NE Protostellar system

L1551 NE was discovered by Emerson et al. (1984) from IRAS data, and it is the second brightest embedded source in L1551 cloud (Lbol ∼ 6L⊙) and sits only ∼2.5^′ northeast to L1551 IRS5 (Figure 2.1). It is classified as a

(35)

Figure 2.5: (a) L1551 IRS5 2 cm continuum image. The two sources are located in north-south direction, with a separation of 42 AU (Bieging &

Cohen, 1985). (b) The binary jet of L1551 IRS5 imaged at 3.5 cm. Each jet is approximately centered on the 7 mm protoplanetary disks. This confirmed that L1551 IRS5 is a binary system where each star is surrounded by a disk and powers a collimated jet (Rodr´ıguez et al., 2003b).

(36)

Class 0 source, mainly because of its shallower density distribution of its circumstellar envelope, which is also observed in other Class 0 protostars (Barsony & Chandler, 1993).

Optical and infrared observation suggests that L1551 NE drives a high- speed protostellar jet. A recent infrared image is shown in Figure 2.6. HH objects and a collimated [Fe II] jet have been attributed to L1551 NE (Hodapp

& Ladd, 1995; Devine et al., 1999; Reipurth et al., 2000; Hayashi & Pyo, 2008). A reflection nebula, whose apex is at the position of L1551 NE, is also seen in optical and infrared (Draper et al., 1985; Reipurth et al., 2000;

Hayashi & Pyo, 2008). Closely examing of the infrared emission indicates that, the collimated [Fe II] jet does not pass through the brightest point of the reflection nebular. The brightest point of the reflection nebular offsets 0.^′′2 northwest from the jet axis.

The binarity of L1551 NE is first shown by Rodr´ıguez et al. (1995), who discovered two sources at 3.5 cm, separate by 0.5^′′ in east-west direction.

Later observation with better angular resolution and high signal-to-noise ratio at the same wavelength confirmed that it is indeed a binary (Reipurth et al., 2002), as shown in Figure 2.7a. At this wavelength, the eastern component NE A is 1.4 times brighter than the western component NE B. The NW-SE orientation of the binary system gives a good explanation to the offest of the jet axis and the brightest point of the reflection nebular. The eastern source drives the [Fe II] jet, while the western source is associated with the reflection nebular. Finally, as shown in Figure 2.7b, Moriarty-Schieven et al.

(2000) detected a continuum source at 1.3 mm located at 1.4^′′ southeast to the binary and interpreted it as the third protostar of L1551 NE system. But this peak was undetected in 3.5 cm.

The molecular outflow, one of the common features of protostellar sys-

(37)

Figure 2.6: Composite color image of L1551 NE, adapted from Hayashi &

Pyo (2008). The blue, green and red colors are assigned to the broad-band filter H, narrow-band filter [Fe II] λ 1.644 µm and narrow-band filter H2, respectively. The reflection nebula extends southwestward, with its brightest point at L1551 NE. A collimated [Fe II] jet with ∼30^′ length and a position angle of 244^◦ is emananted.

(38)

Figure 2.7: (a)3.6 cm continuum of L1551 NE. Two sources are located in east-west direction with 0.5^′′ separation. The western source (labeled A) is brighter than the eastern source (labeled B). (b)1.3 mm continuum.

The stronger central peak is the position of two 3.6 cm continuum source.

Another peak is located 1.4^′′ southeast of the central peak, proposed as a candidate protostar. The diffuse north-south elongation is interpreted as a circumbinary disk.

(39)

tems, has not been confirmed in the L1551 NE system. Moriarty-Schieven et al. (1995) detected blueshifted emission extending westward from L1551 NE with Jams Clerk Maxwell Telescope (JCMT), coincided with the optical/infrared reflection nebula (Figure 2.8) and infered that the blueshifted emission is the molecular outflow. But the redshifted emission distributes in the surrounding of L1551 NE, not only in the eastern region, thus its bipolarity is still unclear. One reason makes it difficult to detect the molecular outflow is that, L1551 NE is located in the prominent outflow powered by L1551 IRS5, as shown in Figure 2.3. The redshifted lobe extends ∼1 pc to the northeast of L1551 IRS5 and passes through the L1551 NE region.

In Figure 2.9 Yokogawa et al. (2003) showed an arc-shaped structure of CS emission westward of L1551 NE, with a velocity gradient along the outflow axis of L1551 IRS5. They were interpreted as the swept-up gas from L1551 IRS5 has collided with L1551 NE envelope, forming the bow-shock-like structure and the deceleration. The L1551 NE region is severely disturbed by the L1551 IRS5 outflow.

(40)

Figure 2.8: CO(3-2) emission around L1551 NE, obtained by Moriarty-Schieven et al.

(1995) using James Clerk Maxwell Telescope (JCMT) with ∼14^′′beam size. L1551 NE is located at (0,0). Blueshifted emission (above) places westwards, coincident with infrared reflection nebula, but redshifted peaks (below) are not located eastward to L1551 NE.

Channels with velocity between 4.5 km s⁻¹ and 10 km s⁻¹ were omitted because of con- fusion due to self-absorption and probable contamination from L1551 IRS5 outflow.

(41)

Figure 2.9: CS(2–1) maps taken by Nobeyama 45m telescope (the black contours) and Nobeyama Millimeter array (the white contours). The large–

scale CS distribution mapped by Nobeyama 45m telescope shows an arc–

shaped structure opens toward L1551 IRS5 (indicated by the black dash-dot line). This is resulted from the collision between the L1551 IRS5 outflow and L1551 NE envelope. A small-scale CS arc–shaped structure opens toward L1551 NE (the white dash) and has a velocity gradien to the direction of L1551 IRS5. This feature is due to the jet from L1551 NE is decelerated by the outside envelope (Yokogawa et al., 2003).

(42)

Chapter 3 Observation of L1551 IRS5 Protostellar System

3.1 Abstract

We studied the outflow of L1551 IRS5 within ∼4000 AU of its driving source.

Our SMA CO(2–1) observation provided the first high-angular resolution toward the vicinity of L1551 IRS5. Under an angular resolution of ∼ 4^′′, we indentified three distinct components. The most prominent component is an X-shaped structure spanning ∼ 20^′′from the center. Its redshifted emission is located in the northeast side and blueshifted emission in the southwest side, which is similar to the large-scale outflow. The second prominent structure is an S-shaped structure spanning ∼ 10^′′ from center. It has the similar symmetry axis to the large-scale outflow, but an opposite velocity pattern.

The outer portion of this feature appears to twist north on the NE side and twist south on the SW side, creating an S-shape. The thrid component is a compact central component spanning 1.^′′4 from center, with the same velocity pattern and similar symmetry axis to the large-scale outflow. This

(43)

component has a small linear extension, but the highest velocity among three components.

The X-shaped component likely comprises the limb-brightened walls of a cone-shaped cavity evacuated by the outflow from two main protostellar components. One of the cavity arms shows multiple kinematic components, which is also seen in single-dish telescope. The multiple kinematic components may be resulted from multiple outflow events. The compact central component likely comprises material within this cavity newly entrained by one or both outflow from the two main protostellar components. The S- shaped likely comprises a precession outflow, with its symmetry axis inclined to the opposite side of plane of the sky than the other two components. The morphology and kinematics of the S-shaped component can be described by a jet with a constant velocity and precesses at a constant angular velocity.

This outflow may be driven by a recently reported candidated protostellar component in the L1551 IRS5 system, whose circumstellar disk is misaligned to the two main protostellar components. The tidal interaction between this circumstellar disk and the more massive northern component may cause the circumstellar disk to precess, thus the outflow of this component also precesses.

This work has been publish in The Astrophysical Journal (Wu et al., 2009) and is presented in section 3.2. Section 3.3 is a supplement to Wu et al. (2009).

(44)

3.2 Multiple Bipolar Molecular Outflow from

the L1551 IRS5 Protostellar System

(45)

The Astrophysical Journal, 698:184–197, 2009 June 10 doi:10.1088/0004-637X/698/1/184

MULTIPLE BIPOLAR MOLECULAR OUTFLOWS FROM THE L1551 IRS5 PROTOSTELLAR SYSTEM Po-Feng Wu^1,2, Shigehisa Takakuwa², and Jeremy Lim²

1Institute of Astrophysics, National Taiwan University, No. 1, Section 4, Roosevelt Road, Taipei 106, Taiwan

2Academia Sinica Institute of Astronomy and Astrophysics, P.O. Box 23-141, Taipei, 106, Taiwan Received 2008 December 4; accepted 2009 March 30; published 2009 May 19

ABSTRACT

The multiple protostellar system L1551 IRS5 exhibits a large-scale bipolar molecular outflow that spans∼1.5 pc on both the northeast (redshifted) and southwest (blueshifted) sides of the system. We have studied this outflow within∼4000 AU of its driving source(s) with the Submillimeter Array. Our CO(2–1) image at ∼4(∼560 AU) resolution reveals three distinct components: (1) an X-shaped structure spanning∼20from center with a similar symmetry axis and velocity pattern as the large-scale outflow; (2) an S-shaped structure spanning∼10from center also with a similar symmetry axis but opposite velocity pattern to the large-scale outflow; and (3) a compact central component spanning∼ 1.4 from center again with a similar symmetry axis and velocity pattern as the large-scale outflow. The X-shaped component likely comprises the limb-brightened walls of a cone-shaped cavity excavated by the outflows from the two main protostellar components. The compact central component likely comprises material within this cavity newly entrained by one or both outflows from the two main protostellar components. The S-shaped component mostly likely comprises a precessing outflow with its symmetry axis inclined in the opposite sense to the plane of the sky than the other two components, taking the S-shaped component out of the cone-shaped cavity along most, if not all, of its entire length. This outflow may be driven by a recently reported candidate third protostellar component in L1551 IRS5, whose circumstellar disk is misaligned relative to the two main protostellar components.

Gravitational interactions between this protostellar component and its more massive northern (and perhaps also southern) neighbor(s) may be causing the circumstellar disk and hence outflow of this component to precess.

Key words: ISM: jets and outflows – stars: formation – stars: individual (L1551 IRS5) – stars: low-mass, brown dwarfs – stars: pre-main sequence

Online-only material: color figure

1. INTRODUCTION

Most near-solar-mass stars are born as members of binary or multiple systems (Duquennoy & Mayor1991; Mathieu1994).

The formation of such systems, although the preferred mode of star formation at least for near-solar-mass stars, is poorly understood. The main competing theoretical models are fragmentation of dense molecular condensations to form systems comprising multiple protostars, or capture of originally single protostars still embedded in their individual condensations to form multiple systems (e.g., review by Tohline2002).

The most direct way of distinguishing between these two possibilities is to compare for a given system the properties of its individual protostars (i.e., orientation of their circumstellar disks and their orbital motion) with the properties of its surrounding molecular condensation (i.e., orientation and rotation). Such a study of the low-mass protostellar system L1551 IRS5 provides the most direct evidence yet that its two main protostellar components formed via the fragmentation of their surrounding parent molecular condensation (Lim & Takakuwa 2006). For many systems (especially those with closely separated components), however, the properties of their individual protostellar components are difficult to measure until the Extended Very Large Array (EVLA) or Atacama Large Millimeter and Submillimeter Array (ALMA) is completed. Indeed, at the present time, even the number of protostellar components present in a given system is not always clear.

Bipolar molecular outflows, found ubiquitously around protostars, are more easily accessible, and can be used to indi-

undetectable or spatially unresolved protostellar components through their individual outflows. Outflows that wind in space with a helical pattern may have driving sources whose circumstellar disks and therefore jets are precessing due to gravitational interactions with closely separated protostellar companions. The degree of collimation and/or dynamical age of the outflow, as well as the opening angle of the outflow cavity at its base, can provide a guide to the evolutionary stage of the driving source.

Here, we study the bipolar molecular outflow associated with L1551 IRS5 at a high angular resolution with an interferometric array. This system was first found as an infrared source by Strom et al. (1976) in the nearby (distance ∼140 pc) L1551 molecular cloud located in the constellation Taurus. L1551 IRS5 was the first protostellar source from which a bipolar molecular outflow, detected in single-dish observations of the CO(1–0) and CO(2–1) lines, was recognized (Snell et al.1980).

Subsequent single-dish observations in CO(1–0), CO(2–1), and CO(3–2) at higher angular resolutions have revealed the overall morphology and kinematics of this and a number of other outflows in the L1551 star-forming region on parsec scales (Uchida et al. 1987; Bachiller et al. 1994; Moriarty-Schieven et al.1987; Moriarty-Schieven & Snell1988; Moriarty-Schieven et al. 2006; Stojimirovi´c et al. 2006). The most prominent of these outflows is centered on L1551 IRS5 and extends

∼1.5 pc toward the northeast (NE) (redshifted lobe) and southwest (SW) (blueshifted lobe) directions at a position angle of

∼ 50^◦. Each outflow lobe has a U-shaped structure, together with knotty components having distinct velocities. The other outflows detected in the L1551 cloud have been associated with other driving sources except for a collimated redshifted outflow

(46)

No. 1, 2009 MULTIPLE BIPOLAR MOLECULAR OUTFLOWS FROM L1551 IRS55 185 Apart from the large-scale bipolar molecular outflow, L1551

IRS5 is surrounded by a relatively compact molecular condensation (radius of∼1200 AU in C¹⁸O) that has been best studied by Momose et al. (1998). This condensation is elongated in the northwest (NW)–southeast (SE) direction, perpendicular to the major axis of the large-scale bipolar molecular outflow. The measured kinematics indicate a flattened envelope that is rotating and contracting, also suggested in earlier observations by Ohashi et al. (1996) and Saito et al. (1996). This condensation has been observed in CS (7–6), which traces warmer and denser gas closer to the central protostellar system, by Takakuwa et al.

(2004). In the inner region of the envelope, the kinematics is dominated by rotation.

The first indication that L1551 IRS5 may be a binary system was reported by Bieging & Cohen (1985), who showed that this system comprised two sources separated by ∼0.3 (∼42 AU) in the north-south direction at a wavelength of 2 cm. Subsequent studies resolved these two sources into twin ionized jets closely aligned with the large-scale bipolar molecular outflow (Rodr´ıguez et al. 2003b). Looney, Mundy

& Welch (1997) showed from high angular-resolution (∼ 0.5) observations in the 2.7 mm continuum that L1551 IRS5 contains two circumstellar dust disks surrounded by a circumbinary dust disk and/or extended envelope. Rodr´ıguez et al. (1998) spatially resolved the two circumstellar dust disks (along their major axes) for the first time at 7 mm, and showed that they are centered on the abovementioned twin ionized jets. Rodr´ıguez et al. (2003a) showed that these two protostellar components were (likely) orbiting each other in a clockwise direction. Lim

& Takakuwa (2006) spatially resolved the two circumstellar disks along both their major and minor axes, and showed that these disks are aligned with each other as well their surrounding rotating and flattened molecular condensation (pseudodisk).

They also showed that the two components were orbiting each other in the same direction as the rotation of the pseudodisk.

All these properties are consistent with the notion that the two protostellar components formed as a result of fragmentation in the central region of their surrounding pseudodisk. Finally, Lim & Takakuwa (2006) found a candidate third protostellar component located (in projection) near the northern protostellar component. The circumstellar disk of this third component is misaligned relative to the circumstellar disks of the two main components as well as their surrounding pseudodisk.

We report in this paper high angular resolution (∼4) observations in CO(2–1) centered on the base of the large-scale bipolar molecular outflow from L1551 IRS5. There is, to the best of our knowledge, only one published interferometric image of this molecular outflow in CO that presented by Barsony et al. (1993) in CO(1–0), who studied relatively extended ( 1) features in the blue outflow lobe. In our study, we investigate the structure and kinematics of the outflow components detected within∼30 of L1551 IRS5, their connection with the large-scale bipolar molecular outflow imaged with single-dish telescopes, and the likely nature of their driving sources.

Readers, interested in how the observations were conducted and data reduced, should now proceed to Section 2. Those interested only in the results of the CO(2–1) and simultaneous 1.3 mm continuum measurements can skip ahead to Section3.

In Section4, we show that our CO(2–1) maps likely trace three separate outflow components: the limb-brightened walls of an outflow cavity, material in the cavity entrained by one or both

Table 1 Observational Parameters

Parameter Value

Right ascension (J2000) 04^h31^m34.^s14

Declination (J2000) 18^◦0805.1

Primary beam (FWHM) ∼57

Synthesized beam (FWHM) 4.66× 2.30 (P.A.= −66.^◦7)

Central frequencies 231.3 GHz, 241.3 GHz

Frequency (velocity) resolution 203.125 kHz (∼0.264 km s⁻¹) Gain calibrators 0423-013 (9.4 Jy), 3C 120 (3.3 Jy) Absolute flux and passband calibrator Uranus

Rms noise level (continuum) 0.014 Jy beam⁻¹

Rms noise level (line) 0.35 Jy beam⁻¹at 203.125 kHz resolution

concise summary of our results and interpretation can proceed directly to Section5.

2. OBSERVATIONS AND DATA REDUCTION Observations of L1551 IRS5 in CO(2–1) and 1.3 mm continuum were carried out using the SMA³ in its compact configu- ration on 2003 December 7. A description of the SMA can be found in Ho et al. (2004). The parameters of the observations are summarized in Table1. Seven of the eight antennas of the SMA were available for the observations. We used the stronger quasar 0423-013, lying 19.^◦6 from L1551 IRS5, for amplitude calibra- tion, and the weaker quasar 3C 120, lying 12.^◦8 from L1551 IRS5, for phase calibration. The flux density of 0423-013 was 9.4 Jy and that of 3C 120 was 3.3 Jy as measured with respect to Uranus, which served as the absolute flux calibrator.

The SMA has a double-sideband receiver in the 230 GHz (1.3 mm) band with a bandwidth of 2 GHz in each sideband.

The lower sideband was tuned to a central frequency of 231.3 GHz, and the upper sideband 241.3 GHz. The signal in each sideband is fed into a correlator that distributes this signal to 24 spectral windows (“chunks” in the SMA nomenclature).

Six of the chunks were divided into 512 channels, which in the CO(2–1) line results in a velocity resolution of 0.265 km s⁻¹ over a velocity range of 135 km s⁻¹. The remaining chunks were divided into 128 channels, and following calibration vector- averaged to make a single continuum channel at 1.3 mm with a total bandwidth of 4.0 GHz. The minimum projected baseline in our observation was∼8 kλ, allowing us to recover ∼30% of the flux from the largest structure visible in our CO(2–1) maps of∼15(e.g., see Wilner & Welch1994).

We calibrated the raw visibility data using MIR, which is an IDL-based data reduction package adopted for the SMA from the MMA software package originally developed for Owens Vally Radio Observatory (OVRO) (Scoville et al 1993). We adopted antenna-based calibration, which provides the best sensitivity as there were no significant baseline-based errors in our data. The calibrated visibility data were Fourier-transformed to produce DIRTY images, and the point-spread function of the telescope deconvolved (CLEANed) from these images using MIRIAD (Sault et al. 1995) to produce the final maps. A ROBUST parameter of 0.5 was adopted, which provided the best compromise between sensitivity and angular resolution.

To make the CO(2–1) channel maps, we first subtracted the continuum emission from the visibility data as derived from

3 The Submillimeter Array (SMA) is a joint project between the Smithsonian

(47)

186 WU, TAKAKUWA, & LIM Vol. 698

Figure 1.Continuum map of L1551 IRS5 at 1.3 mm. Contour levels are plotted at

−2, 2, 10, 20, 30, 40, 50, 60σ , where the rms uncertainty σ = 0.014 Jy beam⁻¹. Crosses indicate the position of the two main protostellar components at 7 mm as measured by Lim & Takakuwa (2006). The synthesized beam is shown as a filled ellipse at the bottom left corner.

a fit to the line-free channels. Unlike in the continuum where the emission is centrally concentrated, in CO(2–1) the emission in many channels is distributed in a complex manner over an extended region resulting in complicated sidelobe patterns. To CLEAN the channel maps, we first selected (i.e., place boxes enclosing) features that we believed to be real. Through trial and error, we converged on features that changed smoothly in structure between neighboring channels, as would be expected for real features. An examination of the residual maps (i.e., after the candidate real features and their sidelobes had been removed) revealed no systematic pattern (i.e., only random noise), indicating that all the real feature had indeed been correctly recovered.

Thus, we believe that any artificial components incorrectly selected for CLEANing are much weaker than the real components in our maps.

3. RESULTS 3.1. 1.3 mm Continuum

Figure1shows our 1.3 mm continuum map of L1551 IRS5 at an angular resolution of 4.66× 2.33. The continuum emission peaks at the location of the protostellar system, as seen in images at comparable angular resolution at 3 mm (Looney, Mundy &

Welch1997) and 0.8 mm (Takakuwa et al.2004). We fitted a two-dimensional Gaussian to the continuum source to derive a total flux density of 1.20± 0.01 Jy and a deconvolved size at full-width half-maximum (FWHM) of (1.5±0.1)×(0.8± 0.2) along a position angle of 146^◦± 8^◦. The size of the continuum source is much larger than the separation of the two main protostellar components (∼0.3), implying that at least a part of the emission arises from the surrounding envelope. Looney, Mundy & Welch (1997) found that, at 3 mm, about half of the continuum emission arises from the circumstellar dust disks of the two main protostellar components (their observation did not have sufficient angular resolution to spatially separate the

Figure 2.SED of the continuum emission from L1551 IRS5. The data points and their corresponding±1σ measurement uncertainties (where large enough to be seen) are tabulated in Table2. The solid curve shows our best chi-squared fit assuming two power-law components with each having Fν ∝ ν^α. One component has α≈ −0.1 as indicated by the short-dashed line and dominates at low frequencies, and the other α≈ 3.0 as indicated by the long-dashed line and dominates at high frequencies (see text).

Table 2

Continuum Flux Density of L1551 IRS5 Wavelength Flux Density (mJy) References^a

18 cm 1.3± 0.4 1

3.6 cm 1.48± 0.04 1

2.0 cm 2.1± 0.2 1

1.3 cm 3.5± 0.4 1

7 mm 12.2± 1.0 1

7 mm 10.1± 0.7 2

2.7 mm 171± 19 3

2.7 mm 162± 6 4

1.3 mm 1200 This work

850 μm 7230 4

450 μm 45700 4

Note.^a1. Rodr´ıguez et al. (1998); 2. Lim & Takakuwa (2006); 3. Momose et al. (1998); 4. Moriarty-Schieven et al. (2006).

that the extended part of the continuum emission traces the inner regions of the pseudodisk.

Both free-free emission from ionized jets and thermal emission from dust can contribute to the continuum emission. In L1551 IRS5, as in many protostellar systems, the ionized jets dominate the emission at centimeter wavelength, whereas dust contributes an increasingly larger fraction of the emission toward shorter wavelengths. To assess the relative contribution from each component at a given wavelength, we have collected all relevant measurements of the continuum emission for L1551 IRS5 at centimeter to submillimeter wavelength as listed in Table2. Figure2shows the spectral energy distribution (SED) of the continuum emission based on the tabulated data. We mod- eled the SED as the sum of two power-law components with each component having Fν ∝ ν^α, where Fνis the intensity at a given frequency ν and α the power-law index. A chi-squared fit to the data gives one component with α= −0.1 ± 0.1 that dominates at low frequencies (the short-dashed line in Figure2), consistent with optically thin free-free emission. The other component has