Water: from clouds to planets
水：从云到行星

Water can exist as a gas (vapor or ‘steam’), as a solid (ice), or as a liquid. At the low pressures of interstellar space, only water vapor and ice occur, with the temperature at which the transition occurs depending on density. At typical cloud densities of $10^{4}$ particles cm^-3, water sublimates around 100 K (Fraser et al., 2001), but at densities of $10^{13}$ cm^-3, corresponding to the midplanes of protoplanetary disks, the sublimation temperature increases to $\sim$160 K. According to the phase diagram of water, liquid water can exist above the triple point at 273 K and 6.12 mbar ($\sim 10^{17}$ cm^-3). Such pressures and temperatures are typically achieved at the surfaces of bodies of the size of Mars or larger and at distances between 0.7 and 1.7 AU for a solar-type star.
水可以存在为气体（蒸汽或“蒸汽”）、固体（冰）或液体。在星际空间的低压下，只有水蒸气和冰存在，转变发生的温度取决于密度。在典型云密度为 $10^{4}$ 粒子 cm ^-3 时，水升华温度约为 100K（Fraser 等人，2001 年），但在密度为 $10^{13}$ cm ^-3 ，对应于原行星盘的中平面时，升华温度增加到 $\sim$ 160K。根据水的相图，液态水可以存在于三相点以上的 273K 和 6.12mbar（ $\sim 10^{17}$ cm ^-3 ）。这样的压力和温度通常在火星大小或更大的天体表面以及太阳型恒星 0.7 到 1.7 AU 之间的距离上实现。

Water ice can take many different crystalline and amorphous forms depending on temperature and pressure. At interstellar densities, crystallization of an initially amorphous ice to the cubic configuration, $I_{c}$, occurs around 90 K. This phase change is irreversible: even when the ice is cooled down again, the crystal structure remains and it therefore provides a record of the temperature history of the ice. Below 90 K, interstellar ice is mostly in a compact high-density amorphous (HDA) phase, which does not naturally occur on planetary surfaces (Jenniskens and Blake, 1994). The densities of water ice in the HDA, LDA and $I_{c}$ phases are 1.17, 0.94 and 0.92 gr cm^-3, respectively, much lower than those of rocks (3.2–4.4 gr cm^-3 for magnesium-iron silicates).
水冰可以根据温度和压力采取许多不同的结晶和非晶形式。在星际密度下，初始非晶冰结晶为立方体结构， $I_{c}$ ，大约在 90K 左右发生。这种相变是不可逆的：即使冰再次冷却，晶体结构仍然保持，因此它提供了冰的温度历史记录。在 90K 以下，星际冰主要处于紧凑的高密度非晶（HDA）相，这在行星表面上并不自然发生（Jenniskens 和 Blake，1994）。HDA、LDA 和 $I_{c}$ 相中水冰的密度分别为 1.17、0.94 和 0.92 克/立方厘米，远低于岩石的密度（镁铁硅酸盐的密度为 3.2-4.4 克/立方厘米）。

Clathrate hydrates are crystalline water-based solids in which small non-polar molecules can be trapped inside ‘cages’ of the hydrogen-bonded water molecules. They can be formed when a gas of water mixed with other species condenses¹¹1Strictly speaking, the term condensation refers to the gas to liquid transition; we adopt here the astronomical parlance where it is also used to denote the gas-to-solid transition. out at high pressure and has enough entropy to form a stable clathrate structure (Lunine and Stevenson, 1985; Mousis et al., 2010). Clathrate hydrates are found in large quantities on Earth, with methane clathrates on the deep ocean floor and in permafrost as the best known examples. They have been postulated to occur in large quantities on other planets and icy solar system bodies.
克拉沙水合物是结晶的基于水的固体，其中小的非极性分子可以被困在氢键结合的水分子的“笼子”内。当水气与其他物种混合在高压下凝结并具有足够的熵形成稳定的克拉沙结构时，它们可以形成。克拉沙水合物在地球上大量存在，其中深海底部的甲烷水合物和冻土层中的甲烷水合物是最为人熟知的例子。据推测，它们在其他行星和冰冷的太阳系天体上也以大量存在。

Except for in-situ mass spectroscopy in planetary and cometary atmospheres, all information about interstellar and solar system water comes from spectroscopic data obtained with telescopes. Because of the high abundance of water in the Earth’s atmosphere, the bulk of the data comes from space observatories. Like any molecule, water has electronic, vibrational and rotational energy levels. Dipole-allowed transitions between electronic states occur at ultraviolet (UV) wavelengths, between vibrational states at near- to mid-infrared (IR) wavelengths, and between rotational states from mid- to far-IR and submillimeter wavelengths.
除了行星和彗星大气中的原位质谱外，有关星际和太阳系水的所有信息都来自使用望远镜获得的光谱数据。由于地球大气中水的丰富，大部分数据来自空间天文台。像任何分子一样，水具有电子、振动和转动能级。电偶极允许的电子态之间的跃迁发生在紫外（UV）波长处，振动态之间的跃迁发生在近红外到中红外（IR）波长处，转动态之间的跃迁发生在中红外到远红外和亚毫米波长处。

Interstellar water vapor observations target mostly the pure rotational transitions. H₂O is an asymmetric rotor with a highly irregular set of energy levels, characterized by quantum numbers $J_{K_{A}K_{C}}$. Because water is a light molecule, the spacing of its rotational energy levels is much larger than that of heavy rotors, such as CO or CS, and the corresponding wavelengths much shorter (0.5 mm vs 3–7 mm for the lowest transitions). The nuclear spins of the two hydrogen atoms can be either parallel or anti-parallel, and this results in a grouping of the H₂O energy levels into ortho ($K_{A}+K_{C}=$odd) and para ($K_{A}+K_{C}=$even) ladders, with a statistical weight ratio of 3:1, respectively. Radiative transitions between these two ladders are forbidden to high order, and only chemical reactions in which an H atom of water is exchanged with an H-atom of a reactant can transform ortho- to para-H₂O and vice versa.
星际水蒸气观测主要针对纯旋转跃迁。H2O 是一个不对称转子，具有高度不规则的能级集，其特征是量子数。因为水是一个轻分子，其旋转能级的间距比重转子（如 CO 或 CS）大得多，相应的波长也要短得多（最低跃迁为 0.5 毫米，而 CO 或 CS 为 3-7 毫米）。两个氢原子的核自旋可以是平行的，也可以是反平行的，这导致 H2O 能级分为正交（奇数）和顺磁（偶数）阶梯，其统计权重比分别为 3:1。这两个阶梯之间的辐射跃迁在高阶上是被禁止的，只有在水的一个氢原子与反应物的一个氢原子交换的化学反应中，才能将正交转变为顺磁，反之亦然。

Infrared spectroscopy can reveal both water vapor and ice. Water has three active vibrational modes: the fundamental $v$=1–0 bands of the $\nu_{1}$ and $\nu_{3}$ symmetric and asymmetric stretches centered at 2.7 $\mu$m and 2.65 $\mu$m, respectively, and the $\nu_{2}$ bending mode at 6.2 $\mu$m. Overtone ($\Delta v=2$ or larger) and combination (e.g., $\nu_{2}$+$\nu_{3}$) transitions occur in hot gas at shorter wavelengths (see Fig. 1 for example). Gas-phase water therefore has a rich vibration-rotation spectrum with many individual lines depending on the temperature of the gas. In contrast, the vibrational bands of water ice have no rotational substructure and consist of very broad profiles, with the much stronger $\nu_{3}$ band overwhelming the weak $\nu_{1}$ band. The ice profile shapes depend on the morphology, temperature and environment of the water molecules (Hudgins et al., 1993). Crystalline water ice is readily distinguished by a sharp feature around 3.1 $\mu$m that is lacking in amorphous water ice. Libration modes of crystalline water ice are found at 45 and 63 $\mu$m (Moore and Hudson, 1994).
红外光谱学可以揭示水蒸气和冰。水有三种活跃振动模式：基本 $v$ =1-0 带的 $\nu_{1}$ 和 $\nu_{3}$ 对称和非对称伸展位于 2.7 $\mu$ m 和 2.65 $\mu$ m，以及 6.2 $\mu$ m 处的 $\nu_{2}$ 弯曲模式。在较短波长处，热气体中会发生倍频（ $\Delta v=2$ 或更大）和组合（例如， $\nu_{2}$ + $\nu_{3}$ ）跃迁（例如，见图 1）。因此，气相水具有丰富的振动-转动光谱，具有许多个体线，取决于气体的温度。相比之下，水冰的振动带没有旋转亚结构，由非常宽的轮廓组成，较强的 $\nu_{3}$ 带压倒了较弱的 $\nu_{1}$ 带。冰的轮廓形状取决于水分子的形态、温度和环境（Hudgins 等，1993 年）。结晶水冰可通过缺乏于无定形水冰的 3.1 $\mu$ m 附近的尖锐特征轻松区分。结晶水冰的摆动模式位于 45 和 63 $\mu$ m 处（Moore 和 Hudson，1994 年）。

Spectra of hydrous silicates (also known as phyllosilicates, layer-lattice silicates or ‘clays’) show sharp features at 2.70–2.75 $\mu$m due to isolated OH groups and a broader absorption from 2.75–3.2 $\mu$m caused by interlayered (‘bound’) water molecules. At longer wavelengths, various peaks can occur depending on the composition; for example, the hydrous silicate montmorillonite has bands at 49 and 100 $\mu$m (Koike et al., 1982).
含水硅酸盐（也称为云母状硅酸盐、层状晶格硅酸盐或“粘土”）的光谱在 2.70-2.75μm 处显示出由孤立的 OH 基团引起的尖锐特征，以及由层间（“结合”）水分子引起的 2.75-3.2μm 处的较宽吸收。在较长波长处，根据组成可以出现各种峰值；例如，含水硅酸盐蒙脱石在 49 和 100μm 处有吸收带（Koike 等，1982 年）。

Bound-bound electronic transitions of water occur at far-UV wavelengths around 1240 Å, but have not yet been detected in space.
水的束缚-束缚电子跃迁发生在远紫外波长约 1240 埃处，但尚未在太空中检测到。

The strength of an emission or absorption line of water depends on the number of molecules in the telescope beam and, for gaseous water, on the populations of the individual energy levels. These populations, in turn, are determined by the balance between the collisional and radiative excitation and de-excitation of the levels. The radiative processes involve both spontaneous emission and stimulated absorption and emission by a radiation field produced by a nearby star, by warm dust, or by the molecules themselves.
水的发射或吸收线的强度取决于望远镜光束中分子的数量，对于气态水而言，还取决于各个能级的能级的粒子数。这些粒子数又由能级的碰撞激发和辐射激发和去激发之间的平衡决定。辐射过程涉及到自发辐射以及由附近恒星、温暖尘埃或分子本身产生的辐射场的刺激吸收和辐射。

The main collisional partner in interstellar clouds is H₂. Accurate state-to-state collisional rate coefficients, $C_{u\ell}$, of H₂O with both ortho- and para-H₂ over a wide range of temperatures have recently become available thanks to a dedicated chemical physics study (Daniel et al., 2011). Other collision partners such as H, He and electrons are generally less important. In cometary atmospheres, water itself provides most of the collisional excitation.
星际云中的主要碰撞伴侣是 H ₂ 。最近，由于一项专门的化学物理研究（Daniel 等，2011 年），已经获得了 H ₂ O 与正 H 和反 H ₂ 在广泛温度范围内的准确状态对状态碰撞速率系数 $C_{u\ell}$ 。其他碰撞伴侣如 H、He 和电子通常不那么重要。在彗星大气中，水本身提供了大部分碰撞激发。

Astronomers traditionally analyze molecular observations through a Boltzmann diagram, in which the level populations are plotted versus the energy of the level involved. The slope of the diagram gives the inverse of the excitation temperature. If collisional processes dominate over radiative processes, the populations are in ‘local thermodynamic equilibrium’ (LTE) and the excitation temperature is equal to the kinetic temperature of the gas, $T_{\rm ex}=T_{\rm kin}$. Generally level populations are far from LTE and molecules are excited by collisions and de-excited by spontaneous emission, leading to $T_{\rm ex}<T_{\rm kin}$. The critical density roughly delineates the transition between these regimes: $n_{cr}=A_{u\ell}/C_{u\ell}$ and therefore scales with $\mu_{u\ell}^{2}\nu_{u\ell}^{3}$, where $A$ is the Einstein spontaneous emission coefficient, $\mu$ the electric dipole moment and $\nu$ the frequency of the transition $u\to\ell$. In the case of water, the combination of a large dipole moment (1.86 Debye) and high frequencies results in high critical densities of $10^{8}$–$10^{9}$ cm^-3 for pure rotational transitions.
天文学家传统上通过玻尔兹曼图来分析分子观测，其中级别的分布与所涉及级别的能量相对比。图的斜率给出了激发温度的倒数。如果碰撞过程主导辐射过程，那么分布处于‘局部热力学平衡’（LTE）状态，激发温度等于气体的动力学温度。一般来说，级别的分布远离 LTE，分子通过碰撞激发并通过自发辐射去激发，导致。临界密度大致划定了这些状态之间的过渡：，因此与成比例，其中是爱因斯坦自发辐射系数，是电偶极矩，是过渡的频率。在水的情况下，大偶极矩（1.86 Debye）和高频率的组合导致纯旋转跃迁的临界密度很高，为 - 厘米。

Analysis of water lines is much more complex than for simple molecules, such as CO, for a variety of reasons. First, because of the large dipole moment and high frequencies, the rotational transitions of water are usually highly optically thick, even for abundances as low as $10^{-10}$. Second, the water transitions couple effectively with mid- and far-infrared radiation from warm dust, which can pump higher energy levels. Third, the fact that the ‘backbone’ levels with $K_{A}$=0 or 1 have lower radiative decay rates than higher $K_{A}$ levels can lead to population ‘inversion’, in which the population in the upper state divided by its statistical weight exceeds that for the lower state (i.e., $T_{\rm ex}$ becomes negative). Infrared pumping can also initiate this inversion. The result is the well-known maser phenomenon, which is widely observed in several water transitions in star-forming regions (e.g., Furuya et al., 2003; Neufeld et al., 2013; Hollenbach et al., 2013). The bottom line is that accurate analysis of interstellar water spectra often requires additional independent constraints, for example from H${}_{2}^{18}$O or H${}_{2}^{17}$O isotopologues, whose abundances are reduced by factors of about 550 and 2500, respectively, and whose lines are more optically thin. At infrared wavelengths, lines are often spectrally unresolved, which further hinders the interpretation.
水线的分析比对简单分子（如 CO）要复杂得多，原因有很多。首先，由于水具有较大的偶极矩和高频率，水的旋转跃迁通常具有很高的光学厚度，即使对于丰度低至 $10^{-10}$ 。其次，水的跃迁与来自暖尘埃的中远红外辐射有效耦合，可以激发更高能级。第三，具有 $K_{A}$ =0 或 1 的‘骨干’能级的辐射衰减速率低于更高 $K_{A}$ 能级，可能导致人口‘反转’，即上态的人口除以其统计权重超过下态的情况（即 $T_{\rm ex}$ 变为负值）。红外泵浦也可以引发这种反转。结果就是众所周知的激射现象，在星形成区域的几个水跃迁中广泛观察到（例如，Furuya 等人，2003 年；Neufeld 等人，2013 年；Hollenbach 等人，2013 年）。 准确分析星际水谱往往需要额外的独立约束条件，例如来自 H ${}_{2}^{18}$ O 或 H ${}_{2}^{17}$ O 同位素的约束，它们的丰度分别降低约 550 倍和 2500 倍，其谱线更薄。在红外波长下，谱线通常是未解析的，这进一步阻碍了解释。

Herschel-HIFI and PACS data show strong and broad water profiles characteristic of shocks associated with embedded protostars, from low to high mass. In fact, for low-mass protostars this shocked water emission completely overwhelms the narrower lines from the bulk of the collapsing envelope, even though the shocks contain less than 1% of the mass of the system. Maps of the water emission around solar-mass protostars such as L1157 reveal water not only at the protostellar position but also along the outflow at ‘hot spots’ where the precessing jet interacts with the cloud (Nisini et al., 2010). Thus, water traces the currently shocked gas at positions, which are somewhat offset from the bulk of the cooler entrained outflow gas seen in the red- and blue-shifted lobes of low-$J$ CO lines (Tafalla et al., 2013; Lefloch et al., 2010).
赫歇尔-HIFI 和 PACS 数据显示出与嵌入式原恒星相关的冲击特征强烈而宽广的水谱线，从低质量到高质量。实际上，对于低质量原恒星，这种受到冲击的水发射完全压倒了来自坍缩包层大部分的较窄线，即使这些冲击所含的质量不到系统总质量的 1%。围绕太阳质量原恒星的水发射图，如 L1157，显示出水不仅存在于原恒星位置，还存在于与云层相互作用的“热点”处的流出口（Nisini 等人，2010 年）。因此，水迹象目前受到冲击的气体在位置上的存在，这些位置与在低 $J$ CO 线的红移和蓝移叶瓣中看到的较冷流出气体的大部分有些偏移（Tafalla 等人，2013 年；Lefloch 等人，2010 年）。

Determinations of the water abundance in shocks vary from values as low as $10^{-7}$ to as high as $10^{-4}$ (see van Dishoeck et al., 2013, for summary). In non-dissociative shocks, the temperature reaches values of a few thousand K and all available oxygen is expected to be driven into water (Kaufman and Neufeld, 1996). The low values likely point to the importance of UV radiation in the shock chemistry and shock structure. For the purposes of this chapter, the main point is that even though water is rapidly produced in shocks at potentially high abundances, the amount of water contained in the shocks is small, and, moreover, most of it is lost to space through outflows.
在冲击中确定的水丰度值从低至 $10^{-7}$ 到高至 $10^{-4}$ 不等（见 van Dishoeck 等人，2013 年，摘要）。在非解离性冲击中，温度达到几千 K 的数值，所有可用的氧都预计会转化为水（Kaufman 和 Neufeld，1996 年）。低值可能指向 UV 辐射在冲击化学和冲击结构中的重要性。对于本章的目的，主要观点是，即使水在冲击中迅速产生，且可能丰度很高，但冲击中含有的水量很少，而且大部分通过流出失去到太空中。

As the cloud collapses to form a protostar in the center, the water-ice coated grains created in the natal molecular cloud move inward, feeding the growing star and its surrounding disk (Fig. 4). The water ice abundance can be measured directly through infrared spectroscopy of various water ice bands toward the protostar itself. Close to a hundred sources have been observed, from very low luminosity objects (‘proto-brown dwarfs’) to the highest mass protostars (Gibb et al., 2004; Pontoppidan et al., 2004; Boogert et al., 2008; Zasowski et al., 2009; Öberg et al., 2011). Inferred ice abundances with respect to H₂ integrated along the line of sight are (0.5–1)$\times$$10^{-4}$.
当云坍缩形成中心的原恒星时，出生分子云中形成的水冰包覆颗粒向内移动，为不断增长的恒星及其周围的盘提供营养（图 4）。水冰丰度可以通过直接测量朝向原恒星本身的各种水冰带的红外光谱来确定。已观测到近百个源，从非常低亮度的物体（“原棕矮星”）到最高质量的原恒星（Gibb 等，2004 年；Pontoppidan 等，2004 年；Boogert 等，2008 年；Zasowski 等，2009 年；Öberg 等，2011 年）。沿视线积分的相对于 H ₂ 的推断冰丰度为（0.5-1） $\times$ $10^{-4}$ 。

The water vapor abundance in protostellar envelopes is probed through spectrally-resolved Herschel-HIFI lines. Because the gaseous water line profiles are dominated by broad outflow emission (Fig. 3), this component needs to be subtracted, or an optically thin water isotopologue needs to be used to determine the quiescent water. Clues to the water vapor abundance structure can be obtained through narrow absorption and emission features in so-called (inverse) P-Cygni profiles (see NGC 1333 IRAS 4A in Fig. 3). The analysis of these data proceeds along the same lines as for pre-stellar cores. The main difference is that the dust temperature now increases inwards, from a low value of 10–20 K at the edge to a high value of several hundred K in the center of the core (Fig. 4). In the simplest spherically symmetric case, the density follows a power-law $n\propto r^{-p}$ with $p$=1–2. As for pre-stellar cores, the data require the presence of a photodesorption layer at the edge of the core with a decreasing water abundance at smaller radii, where gaseous water is maintained by the cosmic ray induced photodesorption of water ice (Coutens et al., 2012; Mottram et al., 2013). Analysis of the combined gaseous water and water ice data for the same source shows that the ice/gas ratio is at least $10^{4}$ (Boonman and van Dishoeck, 2003). Thus, the bulk of the water stays in the ice in this cold part, at a high abundance of $\sim 10^{-4}$ as indicated by direct measurements of both the water ice and gas.
通过光谱分辨的赫歇尔-高保真光谱线 (Herschel-HIFI) 探测原恒星包层中的水蒸气丰度。由于气态水谱线轮廓主要由宽流出发射（图 3）构成，因此需要减去该成分，或者使用光学薄的水同位素体来确定静止水。水蒸气丰度结构的线索可以通过所谓的（逆）天鹅座 P 星云轮廓（见图 3 中的 NGC 1333 IRAS 4A）中较窄的吸收和发射特征获得。这些数据的分析与前恒星核的分析方法相同。主要区别在于，尘埃温度现在向内升高，从边缘的 10-20 K 的低值到核心中心的几百 K 的高值（图 4）。在最简单的球对称情况下，密度遵循幂律 $n\propto r^{-p}$ ，其中 $p$ =1-2。对于恒星前核心，数据要求在核心边缘存在光解吸层，在较小的半径处水丰度会减少，其中气态水由宇宙射线诱导的水冰光解吸来维持（Coutens 等人，2012 年；Mottram 等人，2013 年）。 对同一源的混合气态水和水冰数据的分析显示，冰/气体比至少为 $10^{4}$ （Boonman 和 van Dishoeck，2003）。因此，在这个寒冷部分，大部分水保留在冰中，其丰度高达 $\sim 10^{-4}$ ，如直接测量的水冰和气体所示。

When the infalling parcel enters the radius at which the dust temperature reaches $\sim$100 K, the gaseous water abundance jumps from a low value around $10^{-10}$ to values as high as $10^{-4}$ (e.g., Boonman et al., 2003; Herpin et al., 2012; Coutens et al., 2012). The 100 K radius scales roughly as $2.3\times 10^{14}\sqrt{(}L/L_{\odot})$ cm (Bisschop et al., 2007), and is small, 100 AU, for low-mass sources and a few thousand AU for high-mass protostars. The precise abundance of water in the warm gas is still uncertain, however, and can range from $<$$10^{-6}$–$10^{-4}$ depending on the source and analysis (Emprechtinger et al., 2013; Visser et al., 2013). A high water abundance would indicate that all water sublimates from the grains in the ‘hot core’ before the material enters the disk; a low abundance the opposite.
当下落的包裹进入尘埃温度达到 $\sim$ 100 K 的半径时，气态水的丰度从约 $10^{-10}$ 的低值跳升至高达 $10^{-4}$ 的值（例如，Boonman 等，2003 年；Herpin 等，2012 年；Coutens 等，2012 年）。100 K 半径大致按 $2.3\times 10^{14}\sqrt{(}L/L_{\odot})$ 厘米缩放（Bisschop 等，2007 年），对于低质量源为 100 AU，对于高质量原恒星为几千 AU。然而，温暖气体中水的确切丰度仍然不确定，可以在 $<$ $10^{-6}$ – $10^{-4}$ 之间变化，取决于源和分析（Emprechtinger 等，2013 年；Visser 等，2013 年）。高水丰度将表明所有水在物质进入盘之前从颗粒中升华在“热核心”中；低丰度则相反。

The fate of water in protostellar envelopes on scales of the size of the embedded disk is currently not well understood, yet it is a crucial step in the water trail from clouds to disks. To probe the inner few hundred AU, a high excitation line of a water isotopolog line not dominated by the outflow or high angular resolution is needed: ground-based millimeter interferometry of the H${}_{2}^{18}$O $3_{13}-2_{20}$ ($E_{u}=204$ K) line at 203 GHz (Persson et al., 2012) and deep Herschel-HIFI spectra of excited H${}_{2}^{18}$O or H${}_{2}^{17}$O lines, such as the $3_{12}-3_{03}$ ($E_{u}=249$ K) line at 1095 GHz have been used. Two main problems need to be faced in the analysis. First, comparison of ground-based and Herschel lines for the same source show that the high frequency HIFI lines can be optically thick even for H₂¹⁸O and H${}_{2}^{17}$O, because of their much higher Einstein $A$ coefficients. Second, the physical structure of the envelope and embedded disk on scales of a few hundred AU is not well understood (Jørgensen et al., 2005), so that abundances are difficult to determine since the column of warm H₂ is poorly constrained. Compact flattened dust structures are not necessarily disks in Keplerian rotation (Chiang et al., 2008) and only a fraction of this material may be at high temperatures.
目前，在嵌入盘状结构尺度的原恒星包层中，水的命运尚不十分清楚，但它是水从云层到盘状结构轨迹中的关键一步。为了探测内部几百个天文单位，需要一条不受外流或高角分辨率支配的水同位素谱线的高激发线：已经使用了地面毫米波干涉测量 203 GHz 的 H ${}_{2}^{18}$ O $3_{13}-2_{20}$ ( $E_{u}=204$ K) 线 (Persson 等人，2012) 和激发 H ${}_{2}^{18}$ O 或 H ${}_{2}^{17}$ O 线的深 Herschel-HIFI 谱，例如 1095 GHz 的 $3_{12}-3_{03}$ ( $E_{u}=249$ K) 线。分析中需要面对两个主要问题。首先，通过对比同一源的地面谱线和赫歇尔谱线，我们发现即使是 H ₂ ¹⁸ O 和 H ${}_{2}^{17}$ O，其高频 HIFI 谱线也可能具有光学厚度，因为它们的爱因斯坦 $A$ 系数要高得多。其次，几百天文单位尺度的包层和嵌入盘的物理结构尚不清楚（Jørgensen et al., 2005），因此由于暖 H ₂ 的约束条件较差，其丰度难以确定。 紧凑扁平的尘埃结构不一定是开普勒旋转中的盘状结构（Chiang 等人，2008 年），而且这些材料中只有一小部分可能处于高温状态。

Jørgensen and van Dishoeck (2010b) and Persson et al. (2012) measure water columns and use H₂ columns derived from continuum interferometry data on the same scales ($\sim$1^′′) to determine water abundances of $\sim 10^{-8}-10^{-5}$ for three low-mass protostars, consistent with the fact that the bulk of the gas on these scales is cold and water is frozen. From a combined analysis of the interferometric and HIFI data, using C¹⁸O 9–8 and 10–9 data to determine the warm H₂ column, Visser et al. (2013) infer water abundances of $2\times 10^{-5}-2\times 10^{-4}$ in the $\geq$100 K gas, as expected for the larger-scale hot cores.
Jørgensen 和 van Dishoeck（2010b）以及 Persson 等人（2012 年）测量水柱，并使用从相同尺度上的连续干涉数据导出的 H ₂ 柱密度（ $\sim$ 1 ^′′ ）来确定三颗低质量原恒星的水丰度，这与这些尺度上大部分气体是冷的且水结冰的事实一致。通过对干涉和 HIFI 数据的联合分析，使用 C ¹⁸ O 9-8 和 10-9 数据来确定温暖的 H ₂ 柱密度，Visser 等人（2013 年）推断出 100K 气体中的水丰度为 $2\times 10^{-5}-2\times 10^{-4}$ ，这与更大尺度热核心的预期相符。

The important implication of these results is that the bulk of the water stays as ice in the inner few hundred AU and that only a few % of the dust may be at high enough temperatures to thermally sublimate H₂O . This small fraction of gas passing through high-temperature conditions for ice sublimation is consistent with 2D models of collapsing envelope and disk formation, which give fractions of $<1-20$% depending on initial conditions (Visser et al., 2009, 2011; Ilee et al., 2011; Harsono et al., 2013; Hincelin et al., 2013).
这些结果的重要含义是，大部分水保持为冰在内部几百 AU，只有少数%的尘埃可能温度足够高以使 H2O 热力升华。这小部分气体通过高温条件进行冰升华与 2D 模型的坍缩包层和盘形成一致，这些模型根据初始条件给出不同的百分比，取决于（Visser 等人，2009 年，2011 年；Ilee 等人，2011 年；Harsono 等人，2013 年；Hincelin 等人，2013 年）。

The fact that only a small fraction of the material within a few hundred AU radius is at
在几百 AU 半径范围内，只有很小一部分物质处于 100K（§4.3），这意味着大部分水以冰的形式存在，并仍在向内运动（图 4）。然而，在某个半径处，高速下降的包必须遇到低速嵌入式盘，导致边界处的冲击。这种冲击导致冲击前方的尘埃温度比恒星加热所达到的温度更高（参见；参见简单的拟合公式），也可以溅射冰。在早期，吸积速度很高，所有冰都会升华或经历足够强的冲击以诱导溅射。然而，这些物质通常最终进入恒星而不是盘中，因此对当前故事并不感兴趣。人们认为盘的大部分是通过后来坍缩过程中下落的包层吸积而成的，这些包层在大半径进入盘中，冲击要弱得多。$\geq$ 100 K (§ 4.3) implies that most of the water is present as ice and is still moving inwards (Fig. 4). At some radius, however, the high-velocity infalling parcels must encounter the low-velocity embedded disk, resulting in a shock at the boundary. This shock results in higher dust temperatures behind the shock front than those achieved by stellar heating (
在几百 AU 半径范围内，只有很小一部分物质处于 100K（§4.3），这意味着大部分水以冰的形式存在，并仍在向内运动（图 4）。然而，在某个半径处，高速下降的包必须遇到低速嵌入式盘，导致边界处的冲击。这种冲击导致冲击前方的尘埃温度比恒星加热所达到的温度更高（参见；参见简单的拟合公式），也可以溅射冰。在早期，吸积速度很高，所有冰都会升华或经历足够强的冲击以诱导溅射。然而，这些物质通常最终进入恒星而不是盘中，因此对当前故事并不感兴趣。人们认为盘的大部分是通过后来坍缩过程中下落的包层吸积而成的，这些包层在大半径进入盘中，冲击要弱得多。Neufeld and Hollenbach 1994; see   参见Visser et al. 2009 for a simple fitting formula) and can also sputter ices. At early times, accretion velocities are high and all ices would sublimate or experience a shock strong enough to induce sputtering. However, this material normally ends up in the star rather than in the disk, so it is not of interest for the current story. The bulk of the disk is thought to be made up through layered accretion of parcels that fall in later in the collapse process, and which enter the disk at large radii, where the shock is much weaker
用于简单拟合公式的更高尘埃温度，也可以溅射冰。在早期，吸积速度很高，所有冰都会升华或经历足够强的冲击以诱导溅射。然而，这些物质通常最终进入恒星而不是盘中，因此对当前故事并不感兴趣。人们认为盘的大部分是通过后来坍缩过程中下落的包层吸积而成的，这些包层在大半径进入盘中，冲击要弱得多。(Visser et al., 2009)  (Visser 等，2009 年). Indeed, the narrow line widths of H
。事实上，H 的窄线宽${}_{2}^{18}$O of only 1 km s^-1 seen in the interferometric data
O 仅为 1 km s ^-1 在干涉数据中可见(Jørgensen and van Dishoeck, 2010b)
(Jørgensen 和 van Dishoeck，2010b) argue against earlier suggestions, based on Spitzer data, of large amounts of hot water going through an accretion shock in the embedded phase, or even being created through high-temperature chemistry in such a shock
根据斯皮策数据反驳了早期关于大量热水通过内嵌相中的凝聚冲击，甚至是通过这种冲击中的高温化学反应而产生的建议(Watson et al., 2007)  (Watson 等人，2007). This view that accretion shocks do not play a role also contrasts with the traditional view in the solar system community that all ices evaporate and recondense when entering the disk
。这种认为凝聚冲击不起作用的观点也与太阳系社区中的传统观点相矛盾，即所有冰进入盘时都会蒸发和重新凝结(Lunine et al., 1991; Owen and Bar-Nun, 1993)
(Lunine 等人，1991；Owen 和 Bar-Nun，1993).

Figure 5 shows the history of water molecules in disks at the end of the collapse phase at $t_{\rm acc}=2.5\times 10^{5}$ yr for a standard model with an initial core mass of 1 M_⊙, angular momentum $\Omega_{0}=10^{-14}$ s^-1 and sound speed $c_{s}=0.26$ km s^-1 (Visser et al., 2011). The material ending up in zone 1 is the only water that is completely ‘pristine’, i.e., formed as ice in the cloud and never sublimated, ending up intact in the disk. Material ending up in the other zones contains water that sublimated at some point along the infalling trajectory. In zones 2, 3 and 4, close to the outflow cavity, most of the oxygen is in atomic form due to photodissociation, with varying degrees of subsequent reformation. In zones 5 and 6, most oxygen is in gaseous water. Material in zone 7 enters the disk early and comes close enough to the star to sublimate. This material does not end up in the star, however, but is transported outward in the disk to conserve angular momentum, re-freezing when the temperature becomes low enough. The detailed chemistry and fractions of water in each of these zones depend on the adopted physical model and on whether vertical mixing is included (Semenov and Wiebe, 2011), but the overall picture is robust.
图 5 显示了在标准模型中，初始核心质量为 1 M _⊙ ，角动量 $\Omega_{0}=10^{-14}$ s ^-1 和声速 $c_{s}=0.26$ km s ^-1 （Visser 等，2011 年）的情况下，在坍缩阶段结束时盘中水分子的历史， $t_{\rm acc}=2.5\times 10^{5}$ 年。最终进入区域 1 的物质是唯一完全“原始”的水，即在云中形成并从未升华的冰，最终完整地进入盘中。进入其他区域的物质包含在某个时刻沿着向内运动轨迹升华的水。在 2、3 和 4 区，靠近流出腔的地方，大部分氧以原子形式存在，由于光解而有不同程度的随后再形成。在 5 和 6 区，大部分氧以气态水形式存在。进入区域 7 的物质早期进入盘中，并且靠近恒星足够接近以升华。然而，这种物质并未最终进入恒星，而是在盘中向外运输以保持角动量，在温度足够低时重新冻结。 这些区域中水的详细化学和分数取决于采用的物理模型以及是否包括垂直混合（Semenov 和 Wiebe，2011 年），但总体情况是稳健的。

With increasing wavelengths, regions further out and deeper into the disk can be probed. The surface layers of the inner few AU of disks are probed by near- and mid-IR observations. Spitzer-IRS detected a surprising wealth of highly-excited pure rotational lines of warm water at 10–30
随着波长增加，可以探测到更远和更深处的区域。通过近红外和中红外观测可以探测到盘的内部几 AU 的表面层。Spitzer-IRS 探测到了大量高度激发的暖水纯旋转线，在 10-30$\mu$m (Carr and Najita, 2008; Salyk et al., 2008)
(Carr 和 Najita，2008; Salyk 等，2008), and these lines have since been shown to be ubiquitous in disks around low-mass T Tauri stars
，这些线后来被证明在低质量 T Tauri 恒星周围的盘中是普遍存在的(Pontoppidan et al., 2010a; Salyk et al., 2011)
(Pontoppidan 等，2010a; Salyk 等，2011), with line profiles consistent with a disk origin
，具有与盘源一致的线型剖面(Pontoppidan et al., 2010b)
（Pontoppidan 等人，2010b）. Typical water excitation temperatures are
。典型的水激发温度为$T_{\rm ex}$$\approx$450 K. Spectrally resolved ground-based near-IR vibration-rotation lines around 3
450 K。在地面附近的近红外振动-转动线约 3 的光谱分辨率。$\mu$m show that in some sources the water originates in both a disk and a slow disk wind
显示一些来源的水起源于盘和缓慢的盘风(Salyk et al., 2008; Mandell et al., 2012)
(Salyk 等，2008; Mandell 等，2012). Abundance ratios are difficult to extract from the observations, because the lines are highly saturated and, in the case of Spitzer data, spectrally unresolved. Also, the IR lines only probe down to moderate height in the disk until the dust becomes optically thick. Nevertheless, within the more than an order of magnitude uncertainty, abundance ratios of H₂O/CO
丰度比很难从观测数据中提取，因为谱线高度饱和，在 Spitzer 数据的情况下，谱线无法分辨。此外，红外线谱线只能探测到盘中的中等高度，直到尘埃变得光学厚。尽管如此，在超过一个数量级的不确定性范围内，已经推断出了 H ₂ O/CO 的丰度比为 1-10，对应的辐射半径高达几个天文单位$\sim$1–10 have been inferred for emitting radii up to a few AU
1-10 已被推断出用于发射半径高达几个天文单位(Salyk et al., 2011; Mandell et al., 2012)
(Salyk 等人，2011 年; Mandell 等人，2012 年). This indicates that the inner disks have high water abundances of order
这表明内部盘具有高水含量的数量级$\sim 10^{-4}$ and are thus not dry, at least not in their surface layers. The IR data show a clear dichotomy in H₂O detection rate between disks around the lower-mass T Tauri stars and higher-mass, hotter A-type stars
并且因此至少在其表层不干燥。红外数据显示了低质量 T Tauri 恒星周围盘与高质量、热 A 型恒星之间 H ₂ O 检测率的明显二分法(Pontoppidan et al., 2010a; Fedele et al., 2011)
(Pontoppidan 等人，2010a 年; Fedele 等人，2011 年). Also, transition disks with inner dust holes show a lack of water line emission. This is likely due to more rapid photodissociation by stars with higher
此外，具有内部尘埃孔的过渡盘显示出缺乏水线发射。这可能是由于更快的光解作用，因为更高的恒星辐射更强$T_{*}$, and thus stronger UV radiation, in regions where the molecules are not shielded by dust.
，因此在分子未被尘埃屏蔽的区域中，UV 辐射更强。

Moving to longer wavelengths, Herschel-PACS spectra probe gas at intermediate radii of the disk, out to 100 AU. Far-IR lines from warm water have been detected in a few disks (Rivière-Marichalar et al., 2012; Meeus et al., 2012; Fedele et al., 2012, 2013). As for the inner disk, the abundance ratios derived from these data are highly uncertain. Sources in which both H₂O and CO far-infrared lines have been detected (only a few) indicate H₂O/CO column density ratios of $10^{-1}$, suggesting a water abundance of order $10^{-5}$ at intermediate layers, but upper limits in other disks suggest values that may be significantly less. Again the disks around T Tauri stars appear to be richer in water than those around A-type stars (Fig. 6).
转移到更长波长，赫歇尔-PACS 光谱探测盘的中间半径处的气体，延伸到 100 天文单位。已在一些盘中检测到来自温暖水的远红外线。至于内盘，从这些数据推导出的丰度比高度不确定。已检测到 H ₂ O 和 CO 远红外线的源（仅有少数）表明 H ₂ O/CO 柱密度比为 $10^{-1}$ ，表明中间层的水丰度约为 $10^{-5}$ ，但其他盘的上限值表明可能明显较少。再次，围绕 T Tauri 星的盘似乎比围绕 A 型星的盘富含水（图 6）。

In principle, the pattern of water lines with wavelength should allow the transition from the gaseous water-rich to the water-poor (the snow line) to be probed. As shown by LTE excitation disk models, the largest sensitivity to the location of the snow line is provided by lines in the 40–60 $\mu$m region, which is exactly the wavelength range without observational facilities except for SOFIA (Meijerink et al., 2009). For one disk, that around TW Hya, the available shorter and longer wavelength water data have been used to put together a water abundance profile across the entire disk (Zhang et al., 2013). This disk has a dust hole within 4 AU, within which water is found to be depleted. The water abundance rises sharply to a high abundance at the inner edge of the outer disk at 4 AU, but then drops again to very low values as water freezes out in the cold outer disk.
原则上，水线的波长模式应该允许从富含水的气态到贫水（雪线）的过渡得以探测。正如 LTE 激发盘模型所示，对雪线位置的最大敏感性由 40-60 米区域的线提供，这正是除了 SOFIA（Meijerink 等人，2009）之外没有观测设施的波长范围。对于一个盘，即围绕 TW Hya 的盘，已经利用可用的较短和较长波长的水数据来组合整个盘上的水丰度剖面（Zhang 等人，2013）。这个盘在 4 AU 内有一个尘域，其中发现水被耗尽。水丰度在外盘的内边缘（4 AU 处）急剧上升至高丰度，但随后在寒冷的外盘中水结冰后再次降至极低值。

The cold gaseous water reservoir beyond 100 AU is uniquely probed by Herschel-HIFI data of the ground rotational transitions at 557 and 1113 GHz. Weak, but clear detections of both lines have been obtained in two disks, around the nearby T Tau star TW Hya (Hogerheijde et al., 2011) and the Herbig Ae star HD 100546 (Hogerheijde et al., in prep.) (Fig. 3). These are the deepest integrations obtained with the HIFI instrument, with integration times up to 25 hr per line. Similarly deep integrations on 5 other disks do not show detections of water at the same level, nor do shallower observations of a dozen other disks of different characteristics. One possible exception, DG Tau (Podio et al., 2013), is a late class I source with a well-known jet and a high X-ray flux. The TW Hya detection implies abundances of gaseous water around $10^{-7}$ in the intermediate layer of the disk, with the bulk of the oxygen in ice on grains at lower layers. Quantitatively, 0.005 Earth oceans of gaseous water and a few thousand oceans of water ice have been detected (1 Earth ocean = 1.4$\times 10^{24}$ gr=0.00023 M_Earth). While this is plenty of water to seed an Earth-like planet with water, a single Jovian-type planet formed in this ice-rich region could lock up the bulk of this water.
在距离 100 天文单位之外的冷气态水库被 Herschel-HIFI 数据独特地探测到，这些数据涉及 557 和 1113 GHz 的地面旋转跃迁。在两个盘中分别获得了这两条线的微弱但清晰的探测结果，一个盘围绕附近的 T Tau 星 TW Hya（Hogerheijde 等人，2011）和 Herbig Ae 星 HD 100546（Hogerheijde 等人，准备中）（图 3）。这些是使用 HIFI 仪器获得的最深度的积分，每条线的积分时间长达 25 小时。对其他 5 个盘进行的同样深度的积分未显示出相同水平的水探测结果，对另外十几个具有不同特征的盘的较浅观测也未显示出水的探测结果。一个可能的例外是 DG Tau（Podio 等人，2013），它是一个具有众所周知的喷流和高 X 射线通量的晚 I 类源。TW Hya 的探测结果意味着在盘的中间层周围存在气态水，氧的大部分存在于较低层的颗粒冰中。定量上，已经探测到了 0.005 个地球海洋的气态水和几千个海洋的水冰（1 个地球海洋=1.4 个克=0.00023 M）。 尽管这足以在类地行星上播种水，但在这个富含冰的区域形成的单个类木星型行星可能会锁住大部分水。

Direct detections of water ice are complicated by the fact that IR absorption spectroscopy requires a background light source, and thus a favorable near edge-on orientation of the disk. In addition, care has to be taken that foreground clouds do not contribute to the water ice absorption (Pontoppidan et al., 2005). The 3 $\mu$m water ice band has been detected in only a few disks (Terada et al., 2007; Honda et al., 2009). To measure the bulk of the ice, one needs to go to longer wavelengths, where the ice features can be seen in emission. Indeed, the crystalline H₂O features at 45 or 60 $\mu$m have been detected in several sources with ISO-LWS (Malfait et al., 1998, 1999; Chiang et al., 2001) and Herschel-PACS (McClure et al., 2012, Bouwman et al., in prep). Quantitatively, the data are consistent with 25–50% of the oxygen in water ice on grains in the emitting layer.
水冰的直接探测受到了许多复杂因素的影响，因为红外吸收光谱需要一个背景光源，因此需要一个有利的近边缘方向的盘面。此外，必须小心确保前景云不会对水冰吸收产生影响（Pontoppidan 等人，2005 年）。仅在少数几个盘面中检测到了 3 $\mu$ m 水冰带（Terada 等人，2007 年；Honda 等人，2009 年）。要测量冰的大部分，需要转向更长波长，这样冰的特征可以在发射中看到。事实上，ISO-LWS（Malfait 等人，1998 年，1999 年；Chiang 等人，2001 年）和 Herschel-PACS（McClure 等人，2012 年，Bouwman 等人，准备中）已在几个源中检测到了 45 或 60 $\mu$ m 处的结晶 H ₂ O 特征。定量上，数据与在发射层中颗粒上的水冰中的氧的 25-50%一致。

The ISO-LWS far-infrared spectra also suggested a strong signature of hydrated silicates in at least one target (Malfait et al., 1999). Newer Herschel-PACS data show no sign of such a feature in the same target (Bouwman, priv. comm.). An earlier claim of hydrated silicates at 2.7 $\mu$m in diffuse clouds has now also been refuted (Whittet et al., 1998; Whittet, 2010). Moreover, there is no convincing detection of any mid-infrared feature of hydrated silicates in hundreds of Spitzer spectra of T Tauri (e.g., Olofsson et al., 2009; Watson et al., 2009), Herbig Ae (e.g., Juhász et al., 2010) and warm debris (e.g., Olofsson et al., 2012) disks. Overall, the strong observational consensus is that the silicates prior to planet formation are ‘dry’.
ISO-LWS 远红外光谱还表明至少一个目标中存在水合硅酸盐的强烈特征（Malfait 等人，1999 年）。新的 Herschel-PACS 数据显示在同一目标中没有这种特征（Bouwman，私人通讯）。早期声称弥散云中存在 2.7μm 水合硅酸盐的说法现已被驳斥（Whittet 等人，1998 年；Whittet，2010 年）。此外，在数百个 T Tauri（例如，Olofsson 等人，2009 年；Watson 等人，2009 年）、Herbig Ae（例如，Juhász 等人，2010 年）和温暖碎片（例如，Olofsson 等人，2012 年）盘的 Spitzer 光谱中没有任何水合硅酸盐的中红外特征的令人信服的检测。总体而言，观测结果一致表明在行星形成之前的硅酸盐是“干燥的”。

The observations of gaseous water discussed in § 5.1 indicate the presence of both rotationally hot $T_{\rm ex}\approx 450$ K and cold ($T_{\rm ex}<50$ K) water vapor, with abundances of $\sim 10^{-4}$ and much lower values, respectively. Based on the chemistry of water vapor discussed in § 2.4, we expect it to have a relatively well understood distribution within the framework of the disk thermal structure, potentially modified by motions of the various solid or gaseous reservoirs. This is broadly consistent with the observations.
在第 5.1 节讨论的气态水的观测表明存在旋转热 $T_{\rm ex}\approx 450$ K 和冷（ $T_{\rm ex}<50$ K）水蒸气，丰度分别为 $\sim 10^{-4}$ 和较低的值。根据第 2.4 节讨论的水蒸气化学，我们预计它在盘状热结构框架内具有相对清晰的分布，可能会受到各种固体或气态储库的运动的影响。这与观测结果基本一致。

Traditionally, the snow line plays a critical role in the distribution of water, representing the condensation or sublimation front of water in the disk, where the gas temperatures and pressures allow water to transition between the solid and gaseous states (Fig. 7). For the solar nebula disk, there is a rich literature on the topic (Hayashi, 1981; Sasselov and Lecar, 2000; Podolak and Zucker, 2004; Lecar et al., 2006; Davis, 2007; Dodson-Robinson et al., 2009). Within our modern astrophysical understanding, this dividing line in the midplane is altered when viewed within the framework of the entire disk physical structure. There are a number of recent models of the water distribution that elucidate these key issues (Glassgold et al., 2009; Woitke et al., 2009b; Bethell and Bergin, 2009; Willacy and Woods, 2009; Gorti et al., 2011; Najita et al., 2011; Vasyunin et al., 2011; Fogel et al., 2011; Walsh et al., 2012; Kamp et al., 2013).
传统上，雪线在水的分布中起着关键作用，代表盘中水的冷凝或升华前缘，在那里气体温度和压力允许水在固体和气态之间转变（图 7）。对于太阳星云盘，关于这个主题有丰富的文献（Hayashi, 1981; Sasselov and Lecar, 2000; Podolak and Zucker, 2004; Lecar et al., 2006; Davis, 2007; Dodson-Robinson et al., 2009）。在我们现代天体物理学的理解中，这个分界线在中面盘内被改变，当在整个盘物理结构的框架内观察时。有许多最近的水分布模型阐明了这些关键问题（Glassgold et al., 2009; Woitke et al., 2009b; Bethell and Bergin, 2009; Willacy and Woods, 2009; Gorti et al., 2011; Najita et al., 2011; Vasyunin et al., 2011; Fogel et al., 2011; Walsh et al., 2012; Kamp et al., 2013）。

Since the outer asteroid belt is located outside the Hayashi snow line, it provides a natural reservoir of icy bodies in the solar system. This part of the belt is dominated by so-called C, P and D class asteroids with sizes up to a few 100 km at distances of $\sim$3, 4 and $\geq$4 AU, respectivily, characterized by their particularly red colors and very low albedos, ${{}_{<}\atop{{}^{\sim}}}$0.1 (Bus and Binzel, 2002). Because of their spectroscopic similarities to the chemically primitive carbonaceous chondrites found as meteorites on Earth, C-type asteroids have been regarded as largely unaltered, volatile-rich bodies. The P- and D-types may be even richer in organics.
由于外部小行星带位于哈亚希雪线之外，它为太阳系提供了一个天然的冰体储库。这部分小行星带主要由所谓的 C、P 和 D 类小行星主导，其大小可达几百公里，在距离分别为 3、4 和 4 天文单位的位置上，其特点是特别红色和非常低的反照率，为 0.1（Bus and Binzel, 2002）。由于它们在光谱上与地球上发现的化学原始的含碳球粒陨石相似，C 型小行星被认为是基本未改变的、富含挥发性物质的天体。P 型和 D 型可能甚至更富含有机物质。

The water content in these objects has been studied though IR spectroscopy of the 3 $\mu$m band. In today’s solar system, any water ice on the surface would rapidly sublimate at the distance of the belt, so only water bonded to the rocky silicate surface is expected to be detected. Hydrated minerals can be formed if the material has been in contact with liquid water. The majority of the C-type asteroids show hydrated silicate absorption, indicating that they indeed underwent heating and aqueous alteration episodes (Jones et al., 1990, and refs. therein). However, only 10% of the P and D-type spectra show weak water absorption, suggesting that they have largely escaped this processing and that the abundance of hydrated silicates gradually declines in the outer asteroid belt. Nevertheless, asteroids that do not display water absorption on their surfaces (mostly located beyond 3.5 AU) may still retain ices in their interior. Indeed, water fractions of 5–10% of their total mass have been estimated. This is consistent with models that show that buried ice can persist in the asteroid belt within the top few meters of the surface over billions of years, as long as the mean surface temperature is less than about 145 K (Schorghofer, 2008). Water vapor has recently been detected around the dwarf planet Ceres at 2.7 AU in the asteroid belt, with a production rate of at least $10^{26}$ mol s^-1, directly confirming the presence of water (Küppers et al., 2014).
这些天体中的水含量已通过 3 $\mu$ m 带的红外光谱研究。在今天的太阳系中，表面上的任何水冰在带的距离处会迅速升华，因此预计只有与岩石硅酸盐表面结合的水才能被检测到。如果物质接触过液态水，则可以形成含水矿物。大多数 C 型小行星显示出含水硅酸盐吸收，表明它们确实经历了加热和水蚀变事件（Jones 等人，1990 年，以及其中的参考文献）。然而，只有 10%的 P 型和 D 型光谱显示出弱的水吸收，这表明它们在很大程度上逃脱了这种处理，而含水硅酸盐的丰度在外部小行星带中逐渐下降。然而，表面上不显示水吸收的小行星（主要位于 3.5 AU 之外）可能仍然在其内部保留着冰。事实上，它们的总质量中已估计出 5-10%的水分。 这与模型一致，该模型显示，在表面顶部的几米深处，只要平均表面温度低于约 145K（Schorghofer，2008 年），在数十亿年内，埋藏的冰可以在小行星带中持续存在。最近在小行星带 2.7 天文单位处探测到了水蒸气，其产生速率至少为 $10^{26}$ 摩尔每秒 ^-1 ，直接证实了水的存在（Küppers 等，2014 年）。

Hsieh and Jewitt (2006) discovered a new population of small objects in the main asteroid belt, displaying cometary characteristics. These so-called main belt comets (ten are currently known) display clearly asteroidal orbits, yet have been observed to eject dust and thus satisfy the observational definition of a comet. These objects are unlikely to have originated elsewhere in the solar system and to have subsequently been trapped in their current orbits. Instead, they are intrinsically icy bodies, formed and stored at their current locations, with their cometary activity triggered by some recent event.
Hsieh 和 Jewitt（2006 年）发现了主小行星带中一群新的小天体，显示出彗星特征。这些所谓的主带彗星（目前已知十个）明显显示出小行星的轨道，但已观察到它们喷射尘埃，因此满足了彗星的观测定义。这些天体不太可能起源于太阳系中的其他地方，然后被困在它们当前的轨道中。相反，它们是固有的冰体，在它们当前的位置形成并储存，其彗星活动是由某些最近的事件触发的。

Since main belt comets are optically faint, it is not known whether they display hydration spectral features that could point to the presence of water. Activity of main belt comets is limited to the release of dust and direct outgassing of volatiles, like for Ceres, has so far not been detected. The most stringent indirect upper limit for the water production rate derived from CN observations is that in the prototypical main belt comet 133P/Elst-Pizarro $<1.3\times 10^{24}$ mol s^-1 (Licandro et al., 2011), which is subject to uncertainties in the assumed water-to-CN abundance ratio. For comparison, this is five orders of magnitude lower than the water production rate of comet Hale-Bopp. Its mean density is 1.3 gr cm^-3 suggesting a moderately high ice fraction (Hsieh et al., 2004). Herschel provided the most stringent direct upper limits for water outgassing in 176P/LINEAR ($<4\times 10^{25}$ mol s^-1, 3$\sigma$; de Val-Borro et al. 2012) and P/2012 T1 PANSTARRS ($<8\times 10^{25}$ mol s^-1; O’Rourke et al. 2013).
由于主带彗星在光学上很暗淡，目前尚不清楚它们是否显示出可能指向水存在的水合物光谱特征。主带彗星的活动仅限于释放尘埃和挥发物的直接放气，就像矮行星谷神星一样，迄今为止尚未检测到。从 CN 观测中推导出的水产生速率的最严格的间接上限是原型主带彗星 133P/Elst-Pizarro 的 $<1.3\times 10^{24}$ mol s ^-1 （Licandro 等，2011 年），这取决于假定的水对 CN 丰度比的不确定性。作为比较，这比哈雷-博普彗星的水产生速率低五个数量级。其平均密度为 1.3 gr cm ^-3 ，表明具有适度高的冰含量（Hsieh 等，2004 年）。赫歇尔提供了 176P/LINEAR（ $<4\times 10^{25}$ mol s ^-1 ，3 $\sigma$ ；de Val-Borro 等，2012 年）和 P/2012 T1 PANSTARRS（ $<8\times 10^{25}$ mol s ^-1 ；O’Rourke 等，2013 年）水放气的最严格的直接上限。

Another exciting discovery is the direct spectroscopic detection of water ice on the asteroid 24 Themis (Campins et al., 2010; Rivkin and Emery, 2010), the largest (198 km diameter) member of the Themis dynamical family at $\sim$3.2 AU, which also includes three main belt comets. The 3.1 $\mu$m spectral feature detected in Themis is significantly different from those in other asteroids, meteorites and all plausible mineral samples available. Campins et al. (2010) argue that the observations can be accurately matched by small ice particles evenly distributed on the surface. A subsurface ice reservoir could also be present if Themis underwent differentiation resulting in a rocky core and an ice mantle. Jewitt and Guilbert-Lepoutre (2012) find no direct evidence of outgassing from the surface of Themis or Cybele with a 5$\sigma$ upper limit for the water production rate $1.3\times 10^{28}$ mol s^-1, assuming a cometary water-to-CN mixing ratio. They conclude that any ice that exists on these bodies should be relatively clean and confined to a 10% fraction of the Earth-facing surface.$<$
另一个令人兴奋的发现是直接光谱检测到小行星 24 Themis 上的水冰（Campins 等，2010 年；Rivkin 和 Emery，2010 年），它是 Themis 动力家族中最大的成员（直径 198 公里），位于 3.2 AU 处，该家族还包括三颗主带彗星。在 Themis 上检测到的 3.1 米光谱特征与其他小行星、陨石和所有可行的矿物样品中的特征明显不同。Campins 等人（2010 年）认为，这些观测结果可以通过表面均匀分布的小冰颗粒准确匹配。如果 Themis 经历了分化，形成了岩质核心和冰外壳，那么地下冰库也可能存在。Jewitt 和 Guilbert-Lepoutre（2012 年）未发现来自 Themis 或 Cybele 表面的气体排放的直接证据，假设彗星水与 CN 混合比的上限为 5 个 mol/s，他们得出结论，这些天体上存在的任何冰应该相对干净，并且局限于地球朝向表面的 10%部分。

Altogether, these results suggest that water ice may be common below asteroidal surfaces and widespread in asteroidal interiors down to smaller heliocentric distances than previously expected. Their water contents are clearly much higher than those of meteorites that originate from the inner asteroid belt, which have only 0.01% of their mass in water (Hutson and Ruzicka, 2000).
总的来说，这些结果表明，在小行星表面以下，水冰可能很常见，并且在小行星内部的分布范围比先前预期的更小。它们的含水量明显比源自内小行星带的陨石高得多，后者只有 0.01%的质量是水（Hutson 和 Ruzicka，2000）。

Comets are small solar system bodies with radii less than 20 km that have formed and remained for most of their lifetimes at large heliocentric distances. Therefore, they likely contain some of the least-processed, pristine ices from the solar nebula disk. They have often been described as ‘dirty snowballs’, following the model of Whipple (1950), in which the nucleus is visualized as a conglomerate of ices, such as water, ammonia, methane, carbon dioxide, and carbon monoxide, combined with meteoritic materials. However, Rosetta observations during the Deep Impact encounter (Küppers et al., 2005) suggest a dust-to-gas ratio in excess of unity in comet 9P/Tempel 1. Typically, cometary ice/rock ratios are of order unity with an implied porosity well over 50% (A’Hearn, 2011).
彗星是太阳系中半径小于 20 公里的小天体，它们在大半径处形成并在大部分寿命中保持。因此，它们可能含有太阳星云盘中一些未经加工的原始冰。它们经常被描述为“脏雪球”，遵循 Whipple（1950）的模型，其中核被视为由水、氨、甲烷、二氧化碳和一氧化碳等冰以及陨石材料组成的凝聚物。然而，罗塞塔在深度撞击事件期间的观测（Küppers 等，2005）表明，彗星 9P/Tempel 1 中的尘埃与气体比例超过了单位。通常，彗星的冰/岩比例约为单位数量级，暗示孔隙度远远超过 50%（A’Hearn，2011）。

The presence and amount of water in comets is usually quantified through their water production rates, which are traditionally inferred from radio observations of its photodissociation product, OH, at 18 cm (Crovisier et al., 2002). Measured rates vary from $10^{26}$ to $10^{29}$ mol s^-1. The first direct detection of gaseous water in comet 1P/Halley, through its $\nu_{3}$ vibrational band at 2.65 $\mu$m, was obtained using the KAO (Mumma et al., 1986). The 557 GHz transition of ortho-water was observed by SWAS (Neufeld et al., 2000) and $Odin$ (Lecacheux et al., 2003), whereas Herschel provided for the first time access to multiple rotational transitions of both ortho- and para-water (Hartogh et al., 2011). These multi-transition mapping observations show that the derived water production rates are sensitive to the details of the excitation model used, in particular the ill-constrained temperature profile within the coma, with uncertainties up to 50% (Bockelée-Morvan et al., 2012).
彗星中水的存在和数量通常是通过其水产生速率来量化的，这些速率传统上是通过其光解产物 OH 的无线电观测来推断的，观测波长为 18 厘米（Crovisier 等，2002 年）。测量速率从 0 到 1 摩尔每秒不等。通过其 2.65 微米处的 3 个振动带，首次直接探测到 1P/哈雷彗星中的气态水，使用了 KAO（Mumma 等，1986 年）。SWAS（Neufeld 等，2000 年）和 $Odin$ （Lecacheux 等，2003 年）观测到了正水的 557 GHz 跃迁，而赫歇尔首次提供了对正水和顺水的多个旋转跃迁的访问（Hartogh 等，2011 年）。这些多跃迁映射观测表明，得出的水产生速率对所使用的激发模型的细节敏感，特别是对于尾气中温度剖面的不确定性高达 50%（Bockelée-Morvan 等，2012 年）。

Dynamically, comets can be separated into two general groups: short-period, Jupiter-family comets and long-period comets (but see Horner et al. 2003 for a more detailed classification). Short-period comets are thought to originate from the Kuiper belt, or the associated scattered disk, beyond the orbit of Neptune, while long-period comets formed in the Jupiter-Neptune region and were subsequently ejected into the Oort cloud by gravitational interactions with the giant planets. In reality, the picture is significantly more complex due to migration of the giant planets in the early solar system (see below). In addition, recent simulations (Levison et al., 2010) suggest that the Sun may have captured comets from other stars in its birth cluster. In this case, a substantial fraction of the Oort-cloud comets, perhaps in excess of 90%, may not even have formed in the Sun’s protoplanetary disk. Consequently, there is increasing emphasis on classifying comets based on their chemical and isotopic composition rather than orbital dynamics (Mumma and Charnley, 2011). There is even evidence for heterogeneity within a single comet, illustrating that comets may be built up from cometesimals originating at different locations in the disk (A’Hearn, 2011).
动态地，彗星可以分为两个一般性群体：短周期、木星家族彗星和长周期彗星（但请参见 Horner 等人 2003 年的更详细分类）。短周期彗星被认为起源于冥王星轨道外的柯伊伯带或相关的散射盘，而长周期彗星形成于木星-海王星区域，随后通过与巨行星的引力相互作用被抛入奥尔特云。实际上，由于早期太阳系中巨行星的迁移（见下文），情况要复杂得多。此外，最近的模拟（Levison 等人，2010 年）表明太阳可能从其诞生星团中捕获了其他恒星的彗星。在这种情况下，奥尔特云彗星的相当一部分，也许超过 90%，甚至可能并非在太阳的原行星盘中形成。因此，越来越强调根据彗星的化学和同位素组成而不是轨道动力学对彗星进行分类（Mumma 和 Charnley，2011）。 甚至有证据表明单个彗星内部存在异质性，说明彗星可能是由来自盘中不同位置的彗核物质组成的（A'Hearn, 2011）。

Traditionally, the ortho-to-para ratio in water and other cometary volatiles has been used to contrain the formation and thermal history of the ices. Recent laboratory experiments suggest that this ratio is modified by the desorption processes, both thermal sublimation and photodesorption, and may therefore tell astronomers less about the water formation location than previously thought (see discussions in van Dishoeck et al. 2013 and Tielens 2013).
传统上，水和其他彗星挥发物中的正对位比已被用来限制冰的形成和热史。最近的实验室实验证明，这种比率受到脱附过程的影响，包括热升华和光解吸，因此可能告诉天文学家的信息比之前认为的关于水形成位置少（请参见 van Dishoeck 等人 2013 年和 Tielens 2013 年的讨论）。

Water is a significant or major component of almost all moons of the giant planets for which densities or spectral information are available. Jupiter’s Galilean moons exhibit a strong gradient from the innermost (Io, essentially all rock, no ice detected) to the outermost (Callisto, an equal mixture of rock and ice). Ganymede has nearly the same composition and hence rock-to-ice ratio as Callisto. Assuming that the outermost of the moons reflects the coldest part of the circumplanetary disk out of which the moons formed, and hence full condensation of water, Callisto’s (uncompressed) density matches that of solar-composition material in which the dominant carbon-carrier was methane rather than carbon monoxide (Wong et al., 2008). The fact that Callisto has close to (but not quite) the full complement of water expected based on the solar oxygen abundance implies that the disk around Jupiter had a different chemical composition (CH₄-rich, CO-poor) from that of the solar nebula disk, which was CO-dominated (Prinn and Fegley, 1981).
水是几乎所有巨大行星的卫星中的一个重要或主要组成部分，只要有密度或光谱信息可用。 木星的伽利略卫星显示出从最内侧（Io，基本上全是岩石，未检测到冰）到最外侧（木卫一，岩石和冰的混合物）的强烈梯度。 木卫三的成分几乎与木卫一相同，因此岩石与冰的比率也相同。 假设卫星中最外侧反映了卫星形成的环行星盘的最冷部分，因此水完全凝结，木卫一（未压缩）的密度与太阳组成物质的密度相匹配，其中主要的碳载体是甲烷而不是一氧化碳（Wong 等，2008 年）。 木卫一几乎具有预期的太阳氧丰度所期望的全部水含量，这意味着木星周围的盘具有不同的化学组成（富含 CH ₄ ，贫含 CO），而不是太阳星云盘的 CO 主导（Prinn 和 Fegley，1981 年）。

The Saturnian satellites are very different. For satellites large enough to be unaffected by porosity, but excluding massive Titan, the ice-to-rock ratio is higher than for the Galilean moons (Johnson and Estrada, 2009). However, Titan—by far the most massive moon—has a bulk density and mass just in between, and closely resembling, Ganymede’s and Callisto’s. Evidently the Saturnian satellite system had a complex collisional history, in which the original ice-rock ratio of the system was not preserved except perhaps in Titan. The moon Enceladus, which exhibits volcanic and geyser activity, offers the unique opportunity to sample Saturnian system water directly. Neptune’s Triton, like Pluto, has a bulk density and hence ice-rock ratio consistent with what is expected for a solar nebula disk in which CO dominated over CH₄. Its water fraction is about 15–35%. At this large distance from the Sun, N₂ can also be frozen out and Triton’s spectrum is indeed dominated by N₂ ice with traces of CH₄ and CO ices; the water signatures are much weaker than on other satellites (Cruikshank et al., 2000). The smaller Trans Neptunion Objects (TNOs) (few hundred km size) are usually found to have mean densities around 1 g cm^-3 and thus a high ice fraction. Larger TNOs such as Quaoar and Haumea have much higher densities (2.6–3.3 g cm^-3) suggesting a much lower ice content, even compared with Pluto (2.0 g cm^-3$<$) (Fornasier et al., 2013).
土星的卫星非常不同。对于足够大以不受多孔性影响的卫星，但不包括质量巨大的泰坦，其冰岩比高于伽利略卫星（Johnson and Estrada, 2009）。然而，泰坦——迄今为止最大的卫星——具有介于甘尼米德和卡利斯托之间的体积密度和质量，非常类似。显然，土星的卫星系统经历了复杂的碰撞历史，原始冰岩比可能仅在泰坦中得以保留。展示火山和间歇泉活动的卫星土卫二提供了直接采样土星系统水的独特机会。海王星的特里同和冥王星一样，具有符合太阳星云盘中 CO 主导 CH 的体积密度和因此冰岩比的特征。其水分约为 15-35%。在离太阳如此遥远的地方，氮也可能被冻结，特里同的光谱确实以 N2 冰为主，带有少量 CH3 和 CO 冰；水的特征要比其他卫星弱得多。 较小的距海王星天体（TNOs）（几百公里大小）通常具有平均密度约为 1 克/立方厘米，因此含有很高的冰含量。较大的 TNOs，如夸奥尔和哈美亚，具有更高的密度（2.6-3.3 克/立方厘米），表明冰含量较低，甚至与冥王星（2.0 克/立方厘米）相比也是如此（Fornasier 等人，2013 年）。

In summary, the water ice content of the outer satellites varies with position and temperature, not only as a function of distance from the Sun, but also from its parent planet. Ice fractions are generally high ($\geq$50%) in the colder parts and consistent with solar abundances depending on the amount of oxygen locked up in CO. However, collisions can have caused a strong reduction of the water ice content.
总的来说，外卫星的水冰含量随位置和温度变化，不仅与距离太阳的距离有关，还与其母行星有关。在较冷的地区，冰含量通常较高（50%左右），并且与太阳丰度一致，取决于 CO 中氧的锁定量。然而，碰撞可能导致水冰含量大幅减少。

The largest reservoirs of what was once water ice in the solar nebula disk are presumably locked up in the giant planets. If the formation of giant planets started with an initial solid core of 10–15 Earth masses with subsequent growth from a swarm of planetesimals of ice and rock with solar composition, the giant planets could have had several Earth masses of oxygen, some or much of which may have been in water molecules in the original protoplanetary disk. The core also gravitationally attracts the surrounding gas in the disk consisting mostly of H and He with the other elements in solar composition. The resulting gas giant planet has a large mass and diameter, but a low overall density compared with rocky planets. The giant planet atmosphere is expected to have an excess in heavy elements, either due to the vaporization of the icy planetesimals when they entered the envelopes of the growing planet during the heating phase, or due to partial erosion of the original core, or both (Encrenaz, 2008; Mousis et al., 2009). These calculations assume that all heavy elements are equally trapped within the ices initially, which is a debatable assumption, and that the ices fully evaporate, with most of the refractory material sedimenting onto the core. In principle, the excesses provide insight into giant planet formation mechanisms and constraints on the composition of their building blocks.
太阳星云盘中曾经的水冰最大储量可能被锁定在巨大行星中。如果巨大行星的形成始于一个初始固态核心，质量为 10-15 个地球质量，随后从冰和岩石的行星体群中以太阳组成的方式增长，那么巨大行星可能拥有数个地球质量的氧气，其中一部分或大部分可能以水分子的形式存在于原始原行星盘中。核心还通过引力吸引盘中的周围气体，该气体主要由 H 和 He 组成，其他元素以太阳组成的方式存在。由此产生的气态巨行星拥有较大的质量和直径，但与岩石行星相比，整体密度较低。预计巨行星大气中会有过量的重元素，这可能是由于当它们在加热阶段进入不断增长的行星的包层时，冰质行星体蒸发，或者由于原始核心的部分侵蚀，或两者兼而有之。 这些计算假设所有重元素最初均被均匀地困在冰中，这是一个有争议的假设，并且这些冰完全蒸发，其中大部分不挥发物质沉积到了核心上。原则上，这些过剩物提供了对巨行星形成机制以及对它们构成基础的约束的见解。

The D/H ratio of water potentially provides a unique fingerprint of the origin and thermal history of water. Figure 11 summarizes the various measurements of solar system bodies, as well as interstellar ices and protostellar objects.
水的 D/H 比率可能提供了水的起源和热史的独特指纹。图 11 总结了太阳系天体以及星际冰和原恒星物体的各种测量结果。

The primordial [D]/[H] ratio set shortly after the Big Bang is $2.7\times 10^{-5}$. Since then, deuterium has been lost due to nuclear fusion in stars, so the deuterium abundance at the time of our solar system formation, 4.6 billion years ago (redshift of about $z\approx 1.4$), should be lower than the primordial value, but higher than the current day interstellar [D]/[H] ratio. The latter value has been measured in the diffuse local interstellar medium from UV absorption lines of atomic D and H, and varies from place to place but can be as high as $2.3\times 10^{-5}$ (Prodanović et al., 2010). These higher values suggest that relatively little deuterium has been converted in stars, so a [D]/[H] value of the solar nebula disk only slightly above the current ISM value is expected at the time of solar system formation. The solar nebula [D]/[H] ratio can be measured from the solar wind composition as well as from HD/H₂ in the atmospheres of Jupiter and Saturn and is found to be $(2.5\pm 0.5)\times 10^{-5}$ (Robert et al., 2000, and references cited), indeed in-between the primordial and the current ISM ratios.⁴⁴4Note that measured HD/H₂ and HDO/H₂O ratios are 2$\times$(D/H)_X, the D/H ratios in these molecules.
在大爆炸后不久设定的原始[D]/[H]比例为 $2.7\times 10^{-5}$ 。此后，氘在恒星核聚变中丢失，因此在我们太阳系形成时，即 46 亿年前（红移约 $z\approx 1.4$ ），氘的丰度应低于原始值，但高于当前星际[D]/[H]比值。后者的数值已通过原子 D 和 H 的紫外吸收线在弥散的局部星际介质中测量，不同地方有所变化，但可以高达 $2.3\times 10^{-5}$ （Prodanović等，2010）。这些较高的数值表明相对较少的氘已在恒星中转化，因此预计在太阳系形成时，太阳星云盘的[D]/[H]值仅略高于当前 ISM 值。太阳星云[D]/[H]比值可以从太阳风组成以及木星和土星大气中的 HD/H ₂ 中测量，发现为 $(2.5\pm 0.5)\times 10^{-5}$ （Robert 等，2000 年，以及引用的参考文献），确实介于原始值和当前 ISM 比值之间。 ⁴⁴4Note that measured HD/H₂ and HDO/H₂O ratios are 2$\times$(D/H)_X, the D/H ratios in these molecules.

The D/H ratio of Earth’s ocean water is $1.5576\times 10^{-4}$ (‘Vienna Standard Mean Ocean Water’, VSMOW or SMOW). Whether this value is representative of the bulk of Earth’s water remains unclear, as no measurements exist for the mantle or the core. It is thought that recycling of water in the deep mantle does not significantly change the D/H ratio. In any case, the water D/H ratio is at least a factor of 6 higher than that of the gas out of which our solar system formed. Thus, Earth’s water must have undergone fractionation processes that enhance deuterium relative to hydrogen at some stages during its history, of the kind described in § 2.4 for interstellar chemistry at low temperatures.
地球海洋水的 D/H 比例为 $1.5576\times 10^{-4}$ （'维也纳标准海洋水'，VSMOW 或 SMOW）。地球水的大部分是否代表这个值仍不清楚，因为地幔或地核没有测量数据。人们认为，深部地幔中的水循环不会显著改变 D/H 比例。无论如何，水的 D/H 比例至少比太阳系形成时的气体高出 6 倍。因此，地球的水在其历史某些阶段必须经历了增强重氢相对于氢的分馏过程，这类过程在低温下的星际化学中描述在第 2.4 节。

Which solar system bodies show D/H ratios in water similar to those found in Earth’s oceans? The highest [D]/[H] ratios are found in two types of primitive meteorites (Robert et al., 2000): LL3, an ordinary chondrite that may have the near-Earth asteroid 433 Eros as its parent body (up to a factor of 44 enhancement compared to the protosolar ratio in H₂) and some carbonaceous chondrites (a factor of 15 to 25 enhancement). Most chondrites show lower enhancements than the most primitive meteorites. Pre-Herschel observations of six Oort-cloud or long-period comets give a D/H ratio in water of $\sim 3\times 10^{-4}$, a factor of 12 higher than the protosolar ratio in H₂ (Mumma and Charnley, 2011).
哪些太阳系天体显示出与地球海洋中发现的水的 D/H 比类似的比例？最高的[D]/[H]比例出现在两种原始陨石中（Robert 等人，2000 年）：LL3，一种可能具有近地小行星 433 Eros 作为其母体的普通球粒陨石（与 H 中的原始太阳比例相比增加了 44 倍），以及一些含碳球粒陨石（增加了 15 到 25 倍）。大多数球粒陨石显示出比最原始的陨石更低的增强效应。对六颗奥尔特云或长周期彗星的赫歇尔之前的观测显示，水的 D/H 比例为 1，比 H 中的原始太阳比例高 12 倍（Mumma 和 Charnley，2011）。

Herschel provided the first measurement of the D/H ratio in a Jupiter-family comet. A low value of $(1.61\pm 0.24)\times 10^{-4}$, consistent with VSMOW, was measured in comet 103P/Hartley 2 (Hartogh et al., 2011). In additon, a relatively low ratio of $(2.06\pm 0.22)\times 10^{-4}$ was found in the Oort-cloud comet C/2009 P1 Garradd (Bockelée-Morvan et al., 2012) and a sensitive upper limit of $<2\times 10^{-4}$ (3$\sigma$) was obtained in another Jupiter-family comet 45P/Honda-Mrkos-Pajdušáková (Lis et al. 2013). The Jupiter-family comets 103P and 45P are thought to originate from the large reservoir of water-rich material in the Kuiper belt or scattered disk at 30–50 AU. In contrast, Oort cloud comets, currently at much larger distances from the Sun, may have formed closer in, near the current orbits of the giant planets at 5–20 AU, although this traditional view has been challened in recent years (see § 6.2). Another caveat is that isotopic ratios in cometary water may have been altered by the outgassing process (Brown et al., 2011). Either way, the Herschel observations demonstrate that the earlier high D/H values are not representative of all comets.
赫歇尔首次测量了木星家族彗星中的 D/H 比值。在哈特利 2 号彗星（Hartogh 等人，2011）中测得了与 VSMOW 一致的低值 $(1.61\pm 0.24)\times 10^{-4}$ 。此外，在奥尔特云彗星 C/2009 P1 Garradd（Bockelée-Morvan 等人，2012）中发现了相对较低的比值 $(2.06\pm 0.22)\times 10^{-4}$ ，并在另一颗木星家族彗星 45P/Honda-Mrkos-Pajdušáková（Lis 等人，2013）中获得了 $<2\times 10^{-4}$ （3 $\sigma$ ）的敏感上限。据认为，103P 和 45P 木星家族彗星起源于库伯带或分散盘中富含水的大量物质，距太阳 30-50 天文单位。相比之下，奥尔特云彗星目前距离太阳更远，可能在更接近的地方形成，靠近 5-20 天文单位的巨行星当前轨道，尽管这种传统观点近年来受到挑战（见第 6.2 节）。另一个警告是，彗星水中的同位素比值可能已经被气体抛射过程改变（Brown 等人，2011）。无论如何，赫歇尔的观测表明，早期高 D/H 值并不代表所有彗星。

How do these solar system values compare with interstellar and protostellar D/H ratios? Measured D/H ratios in water in protostellar envelopes vary strongly. In the cold outer envelope, D/H values for gas-phase water are as high as $10^{-2}$ (Liu et al., 2011; Coutens et al., 2012, 2013), larger than the upper limits in ices of (2–5)$<$$\times 10^{-3}$ obtained from infrared spectroscopy (Dartois et al., 2003; Parise et al., 2003). In the inner warm envelope, previous discrepancies for gaseous water appear to have been resolved in favor of the lower values, down to (3–5)$\times 10^{-4}$ (Jørgensen and van Dishoeck, 2010a; Visser et al., 2013; Persson et al., 2013, and subm.) The latter values are within a factor of two of the cometary values.
这些太阳系数值与星际和原恒星 D/H 比值相比如何？在原恒星包层中，水的 D/H 比值变化很大。在寒冷的外包层中，气相水的 D/H 值高达 $10^{-2}$ （Liu 等，2011 年；Coutens 等，2012 年，2013 年），比从红外光谱中获得的冰的上限值（2-5） $<$ $\times 10^{-3}$ 要大（Dartois 等，2003 年；Parise 等，2003 年）。在内部温暖的包层中，以前对气态水的差异似乎已经解决，有利于更低的值，降至（3-5） $\times 10^{-4}$ （Jørgensen 和 van Dishoeck，2010a；Visser 等，2013 年；Persson 等，2013 年，及提交）。后者的值在彗星值的两倍范围内。

Based on these data, the jury is still out whether the D/H ratio in solar system water was already set by ices in the early (pre-)collapse phase and transported largely unaltered to the comet-forming zone (cf. Fig. 4), or whether further alteration of D/H took place in the solar nebula disk along the lines described in § 2.4.5 (see also chapter by Ceccarelli et al.). The original D/H ratio in ices may even have been reset early in the embedded phase by thermal cycling of material due to accretion events onto the star (see chapter by Audard et al.). Is the high value of $10^{-2}$ found in cold gas preserved in the material entering the disk or are the lower values of $<10^{-3}$ found in hot cores and ices more representative? Or do the different values reflect different D/H ratios in layered ices (Taquet et al., 2013)?
根据这些数据，关于太阳系水中的 D/H 比是否已经由早期（前）坍缩阶段的冰设置，并且大部分未经改变地运输到形成彗星的区域（参见图 4），目前尚无定论，或者 D/H 的进一步改变是否发生在太阳星云盘中，沿着§2.4.5 描述的方式进行（另请参阅 Ceccarelli 等人的章节）。冰中的原始 D/H 比甚至可能在嵌入阶段早期被由于物质对恒星的吸积事件而导致的热循环重置（请参阅 Audard 等人的章节）。在进入盘中的材料中保留的冷气体中发现的 $10^{-2}$ 的高值是存在的，还是在热核和冰中发现的 $<10^{-3}$ 的较低值更具代表性？或者不同的值是否反映了分层冰中不同的 D/H 比（Taquet 等人，2013 年）？

Regardless of the precise initial value, models have shown that vertical and radial mixing within the solar nebula disk reduces the D/H ratios from initial values as high as $10^{-2}$ to values as low as $10^{-4}$ in the comet-forming zones (Willacy and Woods, 2009; Yang et al., 2012b; Jacquet and Robert, 2013; Furuya et al., 2013; Albertsson et al., 2014). For water on Earth, the fact that the D enrichment is a factor $\sim$6 above the solar nebula value rules out the warm thermal exchange reaction of H₂O + HD $\to$ HDO + H₂ in the inner nebula ($\sim$1 AU) as the sole cause. Mixing with some cold reservoir with enhanced D/H at larger distances is needed.
无论初始值是多少，模型表明，太阳星云盘内的垂直和径向混合将将 D/H 比值从最初高达 $10^{-2}$ 的值降低到彗星形成区域低至 $10^{-4}$ 的值（Willacy 和 Woods，2009 年；Yang 等，2012b 年；Jacquet 和 Robert，2013 年；Furuya 等，2013 年；Albertsson 等，2014 年）。对于地球上的水来说，D 富集是太阳星云值的 $\sim$ 6 倍，排除了 H ₂ O + HD $\to$ HDO + H ₂ 在内部星云（ $\sim$ 1 AU）中作为唯一原因的温暖热交换反应。需要与一些在较远距离具有增强 D/H 的冷贮库混合。

Figure 11 contains values for several other solar system targets. Enceladus, one of Saturn’s moons with volcanic activity, has a high D/H ratio consistent with it being built up from outer solar system planetesimals. Some measurements indicate that lunar water may have a factor of two higher hydrogen isotopic ratio than the Earth’s oceans (Greenwood et al., 2011), although this has been refuted (Saal et al., 2013).
图 11 包含了几个其他太阳系目标的数值。土卫二的土卫星之一，具有火山活动的土卫二，具有与由外太阳系小行星堆积而成的高 D/H 比值一致。一些测量表明，月球水的氢同位素比地球海洋高出两倍，尽管这一观点已被驳斥（Greenwood 等，2011 年；Saal 等，2013 年）。

Mars and Venus have interestingly high D/H ratios. The D/H ratio of water measured in the Martian atmosphere is 5.5 times VSMOW, which is interpreted to imply a significant loss of water over Martian history and associated enhancement of deuterated water. Because D is heavier than H, it escapes more slowly, therefore over time the atmosphere is enriched in deuterated species like HDO. A time-dependent model for the enrichment of deuterium on Mars assuming a rough outgassing efficiency of 50% suggests that the amount of water that Mars must have accreted is 0.04–0.4 oceans. The amount of outgassing may be tested by Curiosity Mars rover measurements of other isotopic ratios, such as ³⁸Ar to ³⁶Ar.
火星和金星的 D/H 比率非常高。在火星大气中测得的水的 D/H 比率是 VSMOW 的 5.5 倍，这被解释为意味着火星历史上水的显著流失和氘水的增加。因为 D 比 H 重，所以它逃逸得更慢，因此随着时间的推移，大气中富集了 HDO 等氘化物种。假设火星上氘的富集是一个时间相关模型，假设粗略的气体抽出效率为 50%，这表明火星必须吸收的水量为 0.04-0.4 个海洋。气体抽出的量可以通过好奇号火星探测器测量其他同位素比率来测试，例如 ³⁸ Ar 到 ³⁶ Ar。

The extremely high D/H ratio for water measured in the atmosphere of Venus, about 100 times VSMOW, is almost certainly a result of early loss of substantial amounts of crustal waters, regardless of whether the starting value was equal to VSMOW or twice that value. The amount of water lost is very uncertain because the mechanism and rate of loss affects the deuterium fractionation (Donahue and Hodges, 1992). Solar wind stripping, hydrodynamic escape, and thermal (Jeans’) escape have very different efficiencies for the enrichment of deuterium per unit amount of water lost. Values ranging from as little as 0.1% to one Earth ocean have been proposed.
金星大气中水的 D/H 比值极高，约为 VSMOW 的 100 倍，几乎可以肯定是由于大量地壳水的早期流失所致，无论起始值是否等于 VSMOW 或是其两倍。流失的水量非常不确定，因为流失的机制和速率会影响氘分馏（Donahue 和 Hodges，1992）。太阳风剥离、流体动力逃逸和热（James'）逃逸对单位流失水量的氘富集效率差异很大。已经提出了从仅有 0.1%到一个地球海洋的数值范围。

Models of hydrogen-dominated atmospheres of giant exoplanets ($T\approx 700-3000$ K; warm Neptune to hot Jupiter) solve for the molecular abundances (i.e., ratios of number densities, also called ‘mixing ratios’) under the assumption of thermochemical equilibrium. Non-equilibrium chemistry is important if transport (e.g., vertical mixing) is rapid enough that material moves out of a given temperature and pressure zone on a timescale short compared with that needed to reach equilibrium. The CH₄/CO ratio is most affected by such ‘disequilibrium’ chemistry, especially in the upper atmospheres. Since CO locks up some of the oxygen, this can also affect the H₂O abundance.
巨大外行星（ $T\approx 700-3000$ K；温暖的海王星到热木星）的氢为主的大气模型解决了分子丰度（即，数密度比，也称为“混合比率”）在热化学平衡假设下的问题。如果传输（例如，垂直混合）足够快，使得物质在短于达到平衡所需时间的时间尺度内移出给定的温度和压力区域，那么非平衡化学就很重要。CH ₄ /CO 比率在这种“非平衡”化学中受到最大影响，尤其是在上层大气中。由于 CO 锁定了一些氧气，这也会影响 H ₂ O 的丰度。

The two main parameters determining the water abundance profiles are the temperature-pressure combination and the C/O ratio. Giant planets often show a temperature inversion in their atmospheres, with temperature increasing with height into the stratosphere, which affects both the abundances and the spectral appearance. However, for the full range of pressures (
确定水丰度剖面的两个主要参数是温度-压力组合和 C/O 比。巨大行星的大气层通常显示温度反转，随着高度增加，温度逐渐升高进入平流层，这影响了丰度和光谱外观。然而，对于全压力范围（$10^{2}$–$10^{-5}$ bar) and temperatures (700–3000 K), water is present at abundances greater than
bar）和温度（700-3000 K），水的丰度大于$10^{-3}$ with respect to H₂ for a solar C/O abundance ratio of 0.54, i.e., most of the oxygen is locked up in H₂O. In contrast, when the C/O ratio becomes close to unity or higher, the H₂O abundance can drop by orders of magnitude especially at lower pressures and higher temperatures, since the very stable CO molecule now locks up the bulk of the oxygen
相对于 H ₂ ，对于太阳 C/O 丰度比 0.54，即大部分氧被锁定在 H ₂ O 中。相反，当 C/O 比接近或更高时，H ₂ O 的丰度可以在较低压力和较高温度下大幅下降，因为非常稳定的 CO 分子现在锁定了大部分氧(Kuchner and Seager, 2005; Madhusudhan et al., 2011b)
（Kuchner 和 Seager，2005；Madhusudhan 等，2011b). CH₄ and other hydrocarbons are enhanced as well. As noted previously, observational evidence (not without controversy) for at least one case of such a carbon-rich giant planet, WASP-12b, has been found
. CH ₄ 和其他碳氢化合物也得到增强。正如先前所指出的，至少有一个这样的碳丰富巨行星 WASP-12b 的观测证据（尽管有争议）已经被发现(Madhusudhan et al., 2011a)
(Madhusudhan 等，2011a). Hot Jupiters like WASP-12b are easier targets for measuring C/O ratios, because their elevated temperature profiles allow both water and carbon-bearing species to be measured without the interference from condensation processes that hamper measurements in Jupiter and Saturn in our own solar system, although cloud-forming species more refractory than water ice can reduce spectral contrast for some temperature ranges. Figure 13 illustrates the changing spectral appearance with C/O ratio for a hot Jupiter (
. 像 WASP-12b 这样的热木星更容易用于测量 C/O 比值，因为它们升高的温度轮廓允许测量水和含碳物质，而不受干扰来自凝结过程的影响，这些过程妨碍了在我们太阳系中的木星和土星中的测量，尽管比水冰更难挥发的云形成物质可能会降低某些温度范围的光谱对比度。图 13 说明了热木星随着 C/O 比值的变化而产生的光谱外观 ($T=2300$ K); for cooler planets (
K)；对于较冷的行星 ($<1000$ K) the changes are less obvious.
变化不太明显。

What can cause exoplanetary atmospheres to have very different C/O ratios from those found in the interstellar medium or even in their parent star? If giant planets indeed form outside the snow line at low temperatures through accretion of planetesimals, then the composition of the ices is a key ingredient in setting the C/O ratio. The volatiles (but not the rocky cores) vaporize when they enter the envelope of the planet and are mixed in the atmosphere during the homogenization process. As the giant planet migrates through the disk, it encounters different conditions and thus different ice compositions as a function of radius (Mousis et al., 2009; Öberg et al., 2011). Specifically a giant planet formed outside the CO snow line at 20–30 AU around a solar mass star may have a higher C/O ratio than that formed inside the CO (but outside the H₂O) snow line.
什么会导致外行星大气具有与星际介质甚至其母星中发现的 C/O 比率非常不同的特征？如果巨大行星确实是在低温下通过行星碎片的聚集在雪线外形成的，那么冰的组成是设定 C/O 比率的关键因素。挥发物质（但不包括岩心）进入行星的包层时蒸发，并在均质化过程中混合在大气中。当巨大行星穿越盘时，它会遇到不同的条件，因此不同的冰组成会随着半径的变化而变化（Mousis 等，2009 年；Öberg 等，2011 年）。具体来说，围绕太阳质量星体在 20-30 AU 处形成的超级行星可能比在 CO（但在 H ₂ O）雪线内部形成的行星具有更高的 C/O 比率。

The exact composition of these planetesimals depends on the adopted model. The solar system community traditionally employs a model of the solar nebula disk, which starts hot and then cools off with time allowing various species to re-freeze. In this model, the abundance ratios of the ices are set by their thermodynamic properties (‘condensation sequence’) and the relative elemental abundances. For solar abundances, water is trapped in various clathrate hydrates like NH₃-H₂O and H₂S-5.75 H₂O between 5 and 20 AU. In a carbon-rich case, no water ice is formed and all the oxygen is CO, CO₂ and CH₃OH ice (Madhusudhan et al., 2011b). The interstellar astrochemistry community, on the other hand, treats the entire disk with non-equilibrium chemistry, both with height and radius. Also, the heritage of gas and ice from the protostellar stage into the disk can be considered (Aikawa and Herbst, 1999; Visser et al., 2009; Aikawa et al., 2012; Hincelin et al., 2013, see Fig. 5). The non-equilibrium models show that indeed the C/O ratio in ices can vary depending on location and differ by factors of 2–3 from the overall (stellar) abundances (Öberg et al., 2011).
这些小行星的确切组成取决于所采用的模型。太阳系社区传统上采用太阳星云盘模型，该模型一开始很热，然后随着时间的推移冷却，使得各种物种可以重新结晶。在这个模型中，冰的丰度比由它们的热力学性质（“凝结序列”）和相对元素丰度确定。对于太阳丰度，水被困在各种类沸石水合物中，如 NH ₃ -H ₂ O 和 H ₂ S-5.75 H ₂ O 在 5 到 20 AU 之间。在富含碳的情况下，不形成水冰，所有氧都是 CO，CO ₂ 和 CH ₃ OH 冰（Madhusudhan 等人，2011b）。另一方面，星际天体化学社区对整个盘进行非平衡化学处理，无论是高度还是半径。此外，可以考虑从原恒星阶段到盘中的气体和冰的遗传（Aikawa 和 Herbst，1999；Visser 等人，2009；Aikawa 等人，2012；Hincelin 等人，2013，见图 5）。 非平衡模型表明，冰中的 C/O 比确实会因位置而异，并且与整体（恒星）丰度相差 2-3 倍（Öberg 等人，2011）。

The composition of Earth-like and super-Earth planets (up to 10 $M_{\Earth}$) is determined by similar thermodynamic arguments, with the difference that most of the material stays in solid form and no hydrogen-rich atmosphere is attracted. Well outside the snow line, at large distances from the parent star, the planets are built up largely from planetesimals that are half rock and half ice. If these planets move inward, the water can become liquid, resulting in ‘ocean planets’ or ‘waterworlds’, in which the entire surface of the planet is covered with water (Kuchner, 2003; Léger et al., 2004). The oceans on such planets could be hundreds of kilometers deep and their atmospheres are likely thicker and warmer than on Earth because of the greenhouse effect of water vapor. Based on their relatively low bulk densities (from the mass-radius relation), the super-Earths GJ 1214b (Charbonneau et al., 2009) and Kepler 22b (Borucki et al., 2012) are candidate ocean planets.
类地和超类地行星（最多可达 10 个 $M_{\Earth}$ ）的组成由类似的热力学论证确定，不同之处在于大部分材料保持固态形式，不会吸引富氢大气。在远离母恒星的大距离处的雪线外，这些行星主要由半岩石半冰的小行星堆积而成。如果这些行星向内运动，水可能变成液态，形成“海洋行星”或“水世界”，其中行星的整个表面都被水覆盖（Kuchner, 2003; Léger 等，2004）。这些行星上的海洋可能有数百公里深，它们的大气可能比地球上的大气更厚更温暖，这是由于水蒸气的温室效应。根据它们相对较低的总体密度（从质量-半径关系），超类地 GJ 1214b（Charbonneau 等，2009）和 Kepler 22b（Borucki 等，2012）是海洋行星的候选者。

Further examination of terrestrial planet models shows that there could be two types of water worlds: those which are true water rich in their bulk composition and those which are mostly rocky but have a significant fraction of their surface covered with water (Earth is in the latter category) (Kaltenegger et al., 2013). Kepler 62e and f are prototypes of water-rich planets within the habitable zone, a category of planets that does not exist in our own solar system (Borucki et al., 2013). Computing the atmospheric composition of terrestrial exoplanets is significantly more complex than that of giant exoplanets and requires consideration of many additional processes, including even plate tectonics (Meadows and Seager, 2011; Fortney et al., 2013).
对地球类行星模型的进一步研究表明，可能存在两种类型的水世界：一种是在其主要成分中富含水的真正水世界，另一种是主要为岩石但表面有大量水覆盖的水世界（地球属于后者）（Kaltenegger 等人，2013 年）。Kepler 62e 和 f 是宜居带内富含水的行星的原型，这是我们太阳系中不存在的一类行星（Borucki 等人，2013 年）。计算类地外行星的大气组成比计算巨行星的复杂得多，需要考虑许多额外的过程，甚至包括板块构造（Meadows 和 Seager，2011 年；Fortney 等人，2013 年）。

As for the terrestrial planets in our own solar system, the presence of one or more giant planet can strongly affect the amount of water on exo-(super)-Earths (Fig. 12). A wide variety of dynamical and population synthesis models on possible outcomes have been explored (e.g. Raymond et al., 2006; Mandell et al., 2007; Bond et al., 2010; Ida and Lin, 2004, 2010; Alibert et al., 2011; Mordasini et al., 2009).
至于我们太阳系中的地球类行星，一个或多个巨行星的存在可能会对外星（超级）地球上的水量产生强烈影响（见图 12）。已经探讨了各种可能结果的动力学和群体综合模型（例如 Raymond 等人，2006 年；Mandell 等人，2007 年；Bond 等人，2010 年；Ida 和 Lin，2004 年，2010 年；Alibert 等人，2011 年；Mordasini 等人，2009 年）。

Water on a terrestrial planet in the habitable zone may be cycled many times between the liquid oceans and the atmosphere through evaporation and rain-out. However, only a very small fraction of water molecules are destroyed or formed over the lifetime of a planet like Earth. The total Earth hydrogen loss is estimated to be 3 kg s^-1. Even if all the hydrogen comes from photodissociation of water, the water loss would be 27 kg s^-1. Given the total mass of the hydrosphere of $1.5\times 10^{21}$ kg, it would take $1.8\times 10^{12}$ yr to deplete the water reservoir. The vast majority of the water bonds present today were, therefore, formed by the chemistry that led to the bulk of water in interstellar clouds and protoplanetary disks.
在宜居带内的地球类行星上，水可能在液态海洋和大气层之间多次循环，通过蒸发和降雨。然而，在类似地球的行星寿命期间，只有极小部分水分子被破坏或形成。据估计，地球总氢损失量为 3 千克每秒。即使所有氢来自水的光解，水的损失量也将为 27 千克每秒。考虑到水圈的总质量为 $1.5\times 10^{21}$ 千克，将需要 $1.8\times 10^{12}$ 年才能耗尽水库。因此，今天存在的绝大多数水键是由导致星际云和原行星盘中大部分水形成的化学反应形成的。