Consequences of a Change in the Galactic Environment of the Sun G.P. Zank Bartol Research Institute, University of Delaware, Newark, DE 19716 and 9 9 9 P.C. Frisch1 1 n Department of Astronomy and Astrophysics, University of Chicago, IL 60637 a J 0 2 Received 18 November 1998; accepted 23 December 1998 1 v 9 7 2 1 0 9 9 / h p - o r t s a : v i X r a 1CurrentlyattheUniversityofCalifornia,BerkeleyAstronomyDepartment,Berkeley,California 94720-3411 – 2 – ABSTRACT The interaction of the heliosphere with interstellar clouds has attracted interest since the late 1920’s, both with a view to explaining apparent quasi-periodic climate “catastrophes” as well as periodic mass extinctions. Until recently, however, models describingthesolarwind-localinterstellar medium(LISM)interaction self-consistently had not been developed. Here, we describe the results of a two-dimensional (2D) simulation of the interaction between the heliosphere and an interstellar cloud with the same properties as currently, except that the Ho density is increased from the present value of n(Ho)∼0.2 cm−3 to 10 cm−3. The mutual interaction of interstellar neutral hydrogen and plasma is included. The heliospheric cavity is reduced considerably in size (approximately 10 – 14 au to the termination shock in the upstream direction) and is highly dynamical. The interplanetary environment at the orbit of the Earth changes markedly, with the density of interstellar Ho increasing to ∼2 cm−3. The termination shock itself experiences periods where it disappears, reforms and disappears again. Considerable mixing of the shocked solar wind and LISM occurs due to Rayleigh-Taylor-like instabilities at the nose, driven by ion-neutral friction. Implications for two anomalously high concentrations of 10Be found in Antarctic ice cores 33 kya and 60 kya, and the absence of prior similar events, are discussed in terms of density enhancements in the surrounding interstellar cloud. The calculation presented here supports past speculation that the galactic environment of the Sun moderates the interplanetary environment at the orbit of the Earth, and possibly also the terrestrial climate. Subject headings: ISM — structure: ISM — general: solar system — interplanetary medium: solar system — general: stars — mass-loss: Sun — solar wind – 3 – 1. Introduction The solar system today is embedded in a warm low density interstellar cloud (T∼7000 K, n(Ho+H+)∼0.3 cm−3), which flows through the solar system with a relative Sun-cloud velocity of ∼26 km s−1. Neutral interstellar gas penetrates the charged solar wind of the heliosphere 2 – 98% of the diffuse material in the heliosphere is interstellar gas, and the densities of neutral interstellar gas and the solar wind are equal at approximately the orbit of Jupiter. The galactic environment of the Sun is regulated by the properties of the interstellar cloud surrounding the solar system. 3 However, when the surrounding cloud is of low density, the solar wind prevents most interstellar gas and dust from reaching 1 au, the location of the Earth. The discovery of small scale structure with column densities ≥ 3 1018 cm−2 in cold interstellar matter (Dieter et al. 1976, Faison et al. 1998, Frail et al. 1994, Meyer and Blades 1996, Watson and Meyer 1996, Heiles 1997), and the structured nature of the interstellar cloud surrounding the solar system, allow the possibility that the spatial density of the interstellar cloud surrounding the solar system may change within the next 104–106 years (Frisch 1995,1997a,1997b,1998; hereafter referred to as FR). Over the past century, many conjectures have appeared in the scientific literature linking encounters with dense interstellar clouds to possible climate changes on Earth (e.g. Shapely 1921; McCrea 1975; Begelman and Rees 1976; Fahr 1968; Reid et al. 1976; McKay and Thomas 1978; Scoville and Sanders 1986; Thaddeus 1986; Frisch 1993; Bzowski et al. 1996, Frisch 1998). For these suggestions to have substance, however, it must first be shown that the interplanetary environment of the Earth varies with changing properties of the surrounding interstellar cloud. It has been shown that in the past, the galactic environment of the Sun has changed as a function of time, and that the cloud complex sweeping past the Sun now has an order-of-magnitude more nearby interstellar gas in the upwind than the downwind directions (Frisch and York 1986,FR). 2The heliopause boundsthe heliosphere, which is the region of space occupied by the solar wind with a radius of roughly 100 pc minimum. 3The interstellar cloud surrounding the heliosphere is sometimes referred to as the “local interstellar cloud”, or LIC. – 4 – Therefore the sensitivity of the heliosphere to variations in the boundary conditions imposed by the LISM justify closer examination. It is the purpose of this paper to show that even a moderate alteration in the density of the cloud surrounding the solar system can yield substantial variations to the interplanetary environment in the inner heliosphere. Early studies investigating a heliosphere embedded in a dense interstellar cloud considered the relative ram pressures of the solar wind and surrounding interstellar cloud to estimate the heliopause location (e.g. Holzer 1972, Holzer 1989). Contemporary models consider the interaction of the solar wind and interstellar medium (ISM) self-consistently, by including the effects of resonant charge exchange between the ionized and neutral gases. In the supersonic solar wind itself, charge-exchange can lead to a significant deceleration of the wind due to the freshly ionized interstellar neutrals extracting momentum from the solar wind. The concomitant reduction in solar wind ram pressure can lead to a significant reduction in the size of the heliospheric cavity. In the boundary region separating the solar wind from the ISM (the “heliosheath”), neutral hydrogen charge exchange with decelerated interstellar plasma acts to partially divert, heat and filter the Ho before it enters the heliosphere. This filtration of Ho in the heliosheath can reduce the number density of inflowing Ho by almost half. The rather complicated nonlinear coupling of plasma and Ho in the vicinity of a stellar wind is now captured in modern models (Baranov and Malama, 1993, Zank et al., 1996a, or see e.g., Zank 1998a for a review). The weak coupling of neutral hydrogen gas and plasma via resonant charge exchange affects both distributions in important ways. This implies that the self-consistent coupling of plasma and neutral hydrogen is necessary for modelling the interaction of the solar wind with the ISM. We employ a self-consistent two-dimensional (2D) numerical simulation to evaluate heliospheric structure and properties when the heliosphere is embedded in a neutral interstellar cloud whose number density is some thirty times greater than at present. – 5 – Table 1: Parameters for Current and Modelled LISM1 “LIC” Solar Wind Model Model Plasma Neutral H ISM (1 au) ISM ISM n(Ho) (cm−3) 0.2 10 n(p+) (cm−3) 0.1–0.25 5.0 0.1 n(e−) (cm−3) 0.1–0.25 5.0 0.1 u (km s−1) 26 400 –26 –26 T(K) 7000 105 8000 8000 M 7.6 1.75 2.48 1“LIC” refers to the current values of the LISM. The remaining columns list the model parameters used in the simulation. See Frisch et al. 1998 for a discussion of LIC parameters. – 6 – 2. Basic Model A multi-fluid model is used here to model the interaction of the solar wind with cloud of enhanced density. The need for modelling the neutrals as a multi-fluid stems from the variation in the charge exchange mean-free-path for H in different regions of the heliosphere and ISM. Large anisotropies are introduced in the neutral gas distribution by charge exchange with the solar wind plasma (both sub- and supersonic regions) and the multi-fluid approach represents an attempt to capture this characteristic in a tractable and computationally efficient manner. We consider the interaction of the heliosphere with an interstellar cloud similar to the cloud now surrounding the solar system, but with neutral number densities increased to 10 cm−3. The parameters of the enhanced density cloud, and the solar wind at 1 au, are given in Table 1. The relatively high neutral density ensures that the Ho distribution is essentially collisional. Williams et al. (1997), using the results of Dalgarno (1960), fitted the function σ = 3.2×10−15E−0.11cm2 0.1 < E < 100 (1) HH eV eV to describe the cross-section for Ho-Ho collisions (E is the neutral atom energy in electronvolts). eV The collisional mean-free-path for Ho with the given input parameters ranges from less than 2 au in the ISM to less than 1 au in the heliospheric boundary regions (below) to less than 2.5 au in the heliosphere itself. This suggests that the multi-fluid description outlined below is suitable on scales larger than a few au for these moderate ISM densities, and this, as illustrated below, is shown to be the case. The heliosphere-ISM environment can be described in terms of three thermodynamically distinct regions; the supersonic solar wind (region 3), the very hot subsonic solar wind (region 2), and the ISM itself (region 1). Each region acts a source of secondary Ho atoms whose distribution reflects that of the plasma distribution in the region. Accordingly, the neutral distribution resulting from the interaction of the solar wind with the surrounding interstellar cloud may be approximated by three distinct neutral components originating from each region (Zank et al. 1996a). Each of these three neutral components is represented by a distinct Maxwellian distribution function appropriate to the characteristics of the source distribution in the multi-fluid – 7 – models. This approximation allows the use of simpler production and loss terms for each neutral component. The complete highly non-Maxwellian H distribution function is then the sum over the three components, 3 f(x,v,t) = f (x,v,t). (2) i Xi=1 In principle, for each component, an integral equation must be solved (Hall 1992; Zank et al. 1996b). Instead, Zank et al. use (2) to obtain three Boltzmann equations corresponding to each neutral component. This is an extension of the procedure developed in Pauls et al. (1995). For component 1, both losses and gains in the interstellar medium need to be included, but only losses are needed in the heliosheath and solar wind. Similarly for components 2 and 3. Thus, for each of the neutral hydrogen components i (i = 1,2 or 3) ∂fi +v·∇f =  P1+P2+P3−(βex +βph)fi region i , (3) i ∂t  −(β +β )f otherwise ex ph i  and P means that the production or source term P is to be evaluated for the parameters 1,2,3 ex of components 1, 2, or 3 respectively. The β and β terms describe losses by either charge ex ph exchange or photoionization respectively. Under the assumption that each of the neutral component distributions is approximated adequately by a Maxwellian, one obtains immediately from (3) an isotropic hydrodynamic description for each neutral component, ∂ρ i +∇·(ρ u ) = Q ; (4) i i ρi ∂t ∂ (ρ u )+∇·[ρ u u +p I] = Q ; (5) i i i i i i mi ∂t ∂ 1 p 1 γ ρ u2+ i +∇· ρ u2u + u p = Q . (6) ∂t (cid:18)2 i i γ−1(cid:19) (cid:20)2 i i i γ−1 i i(cid:21) ei The source terms Q are listed in Pauls et al. (1995) and Zank et al. (1996a). The subscript i above refers to the neutral component of interest (i = 1,2,3), ρ , u , and p denote the neutral i i i component i density, velocity, and isotropic pressure respectively, I the unit tensor and γ (= 5/3) the adiabatic index. The plasma is described similarly by the 2D hydrodynamic equations ∂ρ +∇·(ρu) = Q ; (7) ρp ∂t – 8 – ∂ (ρu)+∇·[ρuu+pI] = Q ; (8) mp ∂t ∂ 1 p 1 γ ρu2+ +∇· ρu2u+ up = Q , (9) ep ∂t (cid:18)2 γ−1(cid:19) (cid:20)2 γ−1 (cid:21) where Q denote the source terms for plasma density, momentum, and energy. These terms (ρ,m,e),p are also defined in Pauls et al. (1995) and Zank et al. (1996a). The remaining symbols enjoy their usual meanings. The proton and electron temperatures are assumed equal in the multi-fluid models. The coupled multi-fluid system of equations (4) – (9) are solved numerically as described in Pauls et al. (1995) and Zank et al. (1996a). 3. Simulation Results 3.1. Global Heliosphere Configuration The global structure of the heliosphere embedded in a high density environment is obtained by integrating the coupled time-dependent system of hydrodynamic equations (4) – (6) and (7) – (9) numerically in two spatial dimensions. The integration begins at time t=0 using as an initial condition the heliosphere embedded in a low density cloud (Zank et al. 1996a). The LISM neutral number density is increased to 10 cm−3 to mimic the “encounter” of the heliosphere with a denser cloud, and it is shown below that the heliosphere then fails to settle into a steady-state or equilibrium configuration. A time sequence of the 2D global plasma structure is illustrated in Plate 1. The time dependence of both the plasma and neutrals is presented separately below. Four successive figures are shown, each separated from the preceding in time by ∼ 66 days. The color depicts the plasma temperature. The Sun is located at the origin and the interstellar wind is assumed to Fig. 1.— Plate 1: A time sequence of the plasma temperature distribution. The color corresponds to the Log[Temperature]. The orderingA,B,C,Dcorrespondstoatemporalseparation fromtheprecedingfigureof∼66days. Thefourfiguresshowanapproximatelyfull evolutionarycyclewhichisrepeatedona∼280dayperiod. Seetextforfurtherdetails. – 9 – flow from the right boundary (located at 1000 au; a numerical semi-circle of radius 1000 au forms the computational domain) with the parameters listed in Table 1. Since the plasma interstellar Mach number is supersonic (see Zank et al. 1996a for a discussion about a subsonic ISM), a bow shock, labelled BS in Plate 1A, forms, which decelerates, heats and diverts the interstellar plasma around the heliospheric obstacle. The plasma is now no longer collisionally equilibrated with the interstellar Ho and charge exchange between the interstellar plasma and neutrals just downstream of the bow shock leads to an effective heating/compression and diversion of the neutral interstellar H in the heliosheath region (between the bow shock and heliopause). As a result, a pile-up of decelerated, diverted and heated interstellar Ho is created in the region downstream of the bow shock, which is illustrated in Plate 2. Plate 2 shows the corresponding global distribution of Ho at the same times as used in Plate 1, except that the Ho density is plotted. Far upstream of the bow shock, the Ho number density is ∼ 10 cm−3, increasing rapidly just downstream of the bow shock in the pile-up region with densities 2–2.5 times greater than the ambient ISM density. Changes in the neutral number density, velocity, temperature and Mach number are seen more easily in the one-dimensional (1D) plots along the stagnation axis than in Plate 2 (discussed later in the context of Figure 2). The number density of neutrals entering the heliosphere at ∼ 10 au is ∼ 7 cm−3, decreasing to ∼ 2 cm−3 at 1 au. The enhanced-density ISM case can be compared to simulations using the current low-density surrounding ISM (the LIC column of Table 1). The hydrogen wall of the high density cloud model possesses a more complicated structure than that of the low density case – it possesses a weak double-peaked structure along the stagnation axis in the upstream direction, reverts to a single-peaked structure away from but in the neighbourhood of the stagnation axis, after which it again becomes double-peaked. The Ho flow is strongly filtered in the vicinity of the wall, this due essentially to the divergence of the Ho flow through charge exchange with the bow shock-diverted interstellar plasma, thus leading to an Ho number density entering the heliosphere that is much Fig. 2.— Plate 2: Atimesequence(correspondingtothatofPlate1)oftheglobaldistributionoftheneutralinterstellarHdensity. Thecolorreferstothedensitymeasuredincm−3. Thehydrogenwallintheupstreamdirectionisclearlyvisible,asistheeffectivefiltrationofH asitenterstheheliosphere. – 10 – lower than that of the ISM. As is well-known (e.g. Holzer 1972), the effect of interstellar H streaming through the supersonic solar wind, and experiencing charge exchange with solar wind protons, is to reduce the momentum of the solar wind significantly. The relatively large Ho number density entering the heliosphere reduces the solar wind ram pressure significantly, 25%–75%, from its nominal non-mediated value thereby reducing the global extent of the heliosphere. For the parameters of Table 1, the heliosphere shrinks dramatically to ∼ 10 – 14 au in the upstream direction. The density of interstellar neutral hydrogen at 1 au becomes ∼ 2 cm−3 (Figure 2), unlike the current state for which no neutral interstellar hydrogen reaches Earth orbit (e.g. Baranov and Malama, 1993; Zank et al., 1996a). In addition, the configuration becomes unstable (see below). Thereductioninsolarwindmomentumisaccompaniedbyanincreaseinthetotaltemperature of the solar wind (solar wind ions plus pickup ions), in comparison to an adiabatically expanding solar wind, since pickup ions acquire a large velocity perpendicular to both the solar wind velocity and magnetic field vectors and therefore have energies that are typically ∼ 1 keV in the unmediated solar wind (where u ∼ 400 km s−1). Accordingly, beyond the ionization cavity (a cavity created by the photoionization of interstellar neutrals by solar wind UV radiation), the solar wind temperature increases (in the absence of pickup ions, the total solar wind temperature decays adiabatically, the polytropic index depending on the nature of the specific heat sources identified to heat the wind). Radial cuts along the stagnation axis for the plasma density, velocity, temperature, and Mach number are illustrated in Figure 1. The increase in the solar wind temperature is evident in Figure 1c. By contrast, in the absence of pickup ions or other heating processes, the solar wind temperature would cool adiabatically according to r−4/3 for an adiabatic index of 5/3.