Les micro-bolomètres non-refroidis bénéficient de développements technologiques récents qui se traduisent par des matrices de plus en plus grandes (1024*768 pixels), pour des pixels de plus en plus petits (17 µm) et de plus en plus sensibles bien que non-refroidis (NETD<60 mK @300 K). L'objectif de cette thèse est d'étudier les détecteurs micro-bolométriques non-refroidis afin de tester leurs capacités en imagerie et en spectroscopie en vue de leur utilisation dans le cadre d'une mission spatiale vers les petits corps du système solaire, dont en particulier la mission Marco Polo R. J'ai étudié le fonctionnement de ces détecteurs en prenant pour exemple le détecteur Nano640Etm de la société Ulis, représentatif de la technologie des micro-bolomètres non-refroidis disponible actuellement. J'ai mesuré la stabilité dans le temps de son signal, l'homogénéité de la matrice du détecteur, et l'influence de différents paramètres de fonctionnement (température du plan focal, temps d'intégration des pixels, gain). La réponse du détecteur est linéaire en fonction de la température de scène à la puissance 4. Les résultats de cette caractérisation m'ont permis de mesurer le NETD du détecteur dans le cadre de cette expérience. J'ai ensuite testé les capacités du détecteur Nano640Etm en imagerie radiométrique calibrée afin d'estimer les erreurs maximales de calibration pour un, deux, ou trois points de calibration placés différemment dans la gamme de température observée / The recent progress of the uncooled micro-bolometer technology has lead to larger detector matrices (1024*768 pixels), with smaller pixel sizes (17 µm) and a higher sensitivity although it is an uncooled technology (NETD<60 mK @300 K). The objective of this thesis is to study uncooled micro-bolometer detectors in order to assess their performance for imagery and spectroscopy within the framework of a space mission towards small bodies of the solar system, including the Marco Polo R mission. I have studied these detectors through the example of the Nano640Etm detector of the Ulis society, which represents well the currently available uncooled micro-bolometer technology. I have measured its signal stability through time, the detector's matrix homogeneity, and the influence of various operating parameters (focal plane temperature, integration time of the pixels, gain). The detector's response is linear as a function of the scene temperature to the power of 4. The results of this characterization allowed me to measure the detector's NETD within this experiment. Then, I tested the Nano640Etm detector's ability to perform calibrated radiometric images in order to estimate the maximum calibration error for one, two, or three calibration points chosen differently within the observed temperature range. Lastly, I assessed the detector's performances for dispersive spectroscopic measurements, using its signal to noise ratio, as a function of the observation wavelength, the scene temperature, and the spectral resolution. The results of these tests are very positive.

Une période intéressante de l'histoire précoce du système solaire est celle du disque, i.e. la période pendant laquelle se déroule la plupart des processus qui vont conduire du mélange de gaz et de poussières nébulaires à des grains et des planétésimaux, qui seront à un stade ultérieur le matériel de départ pour la formation des embryons planétaires et des planètes. Les témoins de cette époque sont les constituants des météorites primitives (chondrites), principalement les inclusions réfractaires (CAIs) et les chondres. Une des questions centrales dans la compréhension des processus à l'origine des CAIs et des chondres est celle du temps. Les travaux récents de Johan Villeneuve ont permis de démontrer que l'26Al et les isotopes du Mg étaient distribués dans le disque d'accrétion à un niveau d'homogénéité permettant d'utiliser le système 26Al-26Mg comme le chronomètre le plus précis qui soit des évènements qui se sont déroulés lors des 2 ou 3 premiers millions d'années du disque. Le but de cette thèse a été de reprendre toute l'étude de l'26Al avec des mesures de plus haute précision que les mesures existantes, en associant les mesures in-situ (sonde ionique) et en roche totale (HR-MC-ICPMS). Les développements analytiques mis en place pour mesurer les compositions isotopiques en Mg ont été appliqués à l'étude d'olivines réfractaires et de chondres de la météorite Allende, et d'olivines de la pallasite Eagle Station. L'ensemble des données a permis d'apporter des éléments de réponse sur les âges de formation des chondres et de leurs précurseurs, et sur l'origine des olivines réfractaires riches en Mg et la possibilité que celles-ci figurent parmi les précurseurs des chondres / The disk history is a very interesting period of the early Solar System history, i.e. the period during which occurred most of the processes leading from the mixing of nebular gas and dust to grains, and then planetesimals, which will then constitute the starting material for formation of planetary embryos and terrestrial planets. The witnesses of this period that we have in the laboratory are the chondrite constituents, mostly refractory inclusions (CAIs) and chondrules. One of the central questions regarding the processes leading to the formation of CAIs and chondrules is the timing. Villeneuve et al. (2009, 2011) have shown that 26Al and Mg isotopes were homogeneously distributed in the accretion disk, at a level allowing the use of the 26Al-26Mg system as the most precise short-lived chronometer to constrain the 2 or 3 first million years of the disk history. My PhD project aimed at reconsidering the 26Al study with more precise measurements, and by combining in-situ (by MC-SIMS) and bulk (by HR-MC-ICPMS) analyses. The analytical developments we set for Mg-isotope measurements (because of the high precision needed) were then applied to a set of extraterrestrial materials, including Mg-rich isolated olivines and Mg-rich olivines in porphyritic type I chondrules from the Allende CV3 meteorite, as well as chondrules from the same meteorite, and olivines from the Eagle Station pallasite. The whole data set allowed to answer questions such as (i) the origin of chondrule precursor materials and the time of chondrule formation, as well as (ii) the origin of Mg-rich refractory olivines, and the possibility that the latters were part of chondrule precursors

O Sistema Solar é um dos temas da Astronomia mais abordados em sala de aula por professores de Ciências. Reconhecendo sua importância, nesta pesquisa buscamos investigar uma proposta de ensino-aprendizagem no tema \"A forma e os movimentos dos planetas no Sistema Solar\", que teve como eixo central a relação entre a observação cotidiana e os modelos científicos atualmente aceitos. Para ajudar na construção dessa proposta, analisamos os Cadernos de Ciências da Proposta Curricular do Estado de São Paulo e as teses e dissertações produzidas no Brasil nas últimas décadas, procurando investigar como estes materiais propõem atividades para o ensino-aprendizagem de conceitos de astronomia. Dialogamos com as ideias de Paulo Freire ao refletir sobre os significados da problematização, do diálogo, da atividade de Extensão Universitária e da importância da formação de sujeitos críticos para compreender o mundo que os cercam. Construímos uma proposta de ensino-aprendizagem com 12 atividades estruturadas a partir da metodologia dos Três Momentos Pedagógicos e realizamos intervenções em um curso para a formação de professores, com o caráter de extensão universitária, oferecido pela Universidade de São Paulo. O objetivo da proposta elaborada foi desenvolver o senso crítico ao relacionar a observação cotidiana com os modelos científicos atualmente aceitos, que geralmente são ensinados em sala de aula. Os dados obtidos na aplicação desta proposta foram analisados com base em três grandes categorias gerais que tinham como focos principais avaliar a percepção das limitações da observação imediata e ingênua e a contradição entre esta e o modelo científico. Os resultados parecem indicar que uma parte dos professores não percebeu a contradição aparente entre a forma da Terra observada no cotidiano e a forma descrita no modelo. Os que percebiam, forneciam argumentos com base em noções de referencial e de escalas e proporção. Embora metade dos professores de nossa amostra tenha percebido esta contradição, a maior parte deles não consegue explicar esta percepção com argumentos científicos ou astronômicos. Na problematização dos movimentos observáveis, todos os professores perceberam a contradição aparente entre o movimento do Sol e os movimentos da Terra aceitos atualmente no modelo. Ainda que apresentassem muitas dificuldades em justificar as razões disso, os professores usaram noções de observação de outros astros e planetas para justificar o modelo. Ao final do curso, percebemos que os professores apresentaram nos debates do tipo Júri Simulado uma série de argumentos trabalhados ao longo do curso, o que parece indicar que, após as atividades, eles passaram a argumentar de maneira menos ingênua sobre as relações entre a observação cotidiana e o modelo. Embora reconheçamos as limitações da proposta apresentada, esta parece se mostrar uma alternativa de grande potencial para intervenções na formação de professores que busquem trabalhar além do conhecimento presente nos materiais didáticos, promovendo um diálogo constante entre o conhecimento astronômico e os aspectos vivenciáveis no cotidiano. / The Solar System is one of the most discussed topics of Astronomy in science classes. Recognizing its importance, this study investigates a teaching-learning proposal about \"the shape and movements of the planets in the Solar System\" which focus on the relationship between daily observation and scientific models currently accepted. We analyze the sciences curriculum proposal of the State of São Paulo and theses and dissertations produced in Brazil in recent decades, describing how these materials propose activities for teaching and learning about astronomy. We have dialogued with the ideas of Paulo Freire in order to reflect on the meanings of \"problematization\", dialogue, university extension and the importance of educating critical subjects to understand the world around them. We created a teaching and learning proposal with 12 structured activities based on the methodology of Three Pedagogical Moments and interventions conducted in a course for the training of teachers, with the character of university extension, offered by the University of São Paulo. The objective of this proposal was develop a critical sense of the subjects relate to everyday observation with currently accepted scientific models, which are usually taught in the classroom. The data obtained in the implementation of this proposal were analyzed based on three main general categories that had as main focus to evaluate the perception of contradiction and limitations of immediate and naive observation and the scientific model. The results seem to indicate that some of the teachers do not realize the apparent contradiction between the Earth\'s shape observed in everyday life and as described in the model. Those who perceived provided arguments based on notions of reference and scales and proportion. Although half of the teachers in our sample have noticed this contradiction, most of them can not explain this perception with astronomical or scientific arguments. In questioning the observable movements, all teachers realized the contradiction between the apparent motion of the Sun and the Earth in the movements currently accepted model. Although presented many difficulties in justifying the reasons that many teachers have used notions of observation of other astronomical bodies and planets to justify the model. At the end of the course, we realize that the teachers presented in the discussions of the type Simulated Jury worked a series of arguments along the course, which seems to indicate that after the activities, they began to argue less naive way relations between everyday observation model. While we recognize the limitations of the proposal, this seems to show a great potential for alternative interventions in teacher seeking work beyond the present knowledge in textbooks, promoting an ongoing dialogue between the astronomical knowledge and aspects in daily life.

Neste trabalho, desenvolvemos um mapeamento simplético que nos permite estudar o comportamento dinâmico de ressonâncias asteroidais no âmbito do problema dos três corpos restrito, elíptico, espacial. Para obter este mapeamento, combinamos um esquema simplético similar ao desenvolvido por Hadjidemetriou (1986) junto com o desenvolvimento assimétrico da função perturbadora (Ferraz-Mello, 1987), que leva em conta as inclinações do perturbado e do perturbador como sendo referidas a um plano invariante (Roig et al., 1997). Este mapeamento é aplicado aos casos das ressonâncias asteroidais 2/1 e 3/2. Estudam-se um grande número de condições iniciais no espaço de fase, de forma a conseguir tirar conclusões de tipo estatístico sobre os processos envolvidos na geração de mecanismos difusivos que podem agir nessas ressonâncias. / In this work, we developed a symplectic mapping which allow us to study the dynamical behaviour of asteroidal resonances in the frame of the non-planar elliptic restricted three-body problem. To obtain such a mapping we combine a symplectic scheme similar to that of Hadjidemetriou (1986) together with an asymmetric expansion of the disturbing funtion (Ferraz-Mello, 1987) which takes into account the inclinations of both the perturber and the disturbed bodies (Roig et al., 1997). This mapping is applied to the 2/1 and 3/2 mean motion resonances in the asteroidal belt. We explore a wide range of initial conditions in the phase space in order to get a large number of results which allow us to make some statistical conclusions about the generation of diffusion mechanisms acting in these resonances.

<div>Planetary rings are, and have been, a common feature throughout the solar system.</div><div>Rings have been observed orbiting each of the giant planets, several Trans-Neptunian Objects, and debris rings are thought to have orbited both Earth and Mars.</div><div>The bright, massive planetary rings orbiting Saturn have been observed for centuries, and the Cassini Mission has given researchers a recent and extensive closeup view of these rings.</div><div>The Saturn ring system has served as a natural laboratory for scientists to understand the dynamics of planetary ring systems, as well as their influence on satellites orbiting nearby.</div><div>Researchers have shown that planetary ring systems and nearby satellites can be tightly-coupled systems.</div><div><br></div><div>In this work, I discuss the physics which dominate the dynamical evolution of planetary ring systems, as well as the interactions with any nearby satellites.</div><div>Many of these dynamics have been incorporated into a one-dimensional mixed Eulerian-Lagrangian numerical model that I call "RING-MOONS," to simulate the long-term evolution of tightly coupled satellite-ring systems.</div><div>In developing RING-MOONS, I have discovered that there are three evolution regimes for tightly-coupled satellite-ring systems which I designate as the "Boomerang," "Torque-Dependent," and "Slingshot" regimes.</div><div>Each regime may be defined using the rotation period of the primary body and the bulk density of the ring material.</div><div><br></div><div>The slow rotation period of Mars places it in the Boomerang regime.</div><div>I hypothesize that a giant impact with Mars ejected material into orbit, forming a debris ring around the planet.</div><div>Using RING-MOONS, I demonstrate how Lindblad torques cause satellites which form at the edge of the ring to initially migrate away from the ring, but over time as the mass of the ring decreases, tidal torques always cause the satellites to migrate inwards.</div><div>Assuming the satellites rapidly tidally disrupt upon migrating to the rigid Roche limit, a new ring is formed.</div><div>I show that debris material cycles between orbiting Mars as a planetary ring, or as discrete satellites, and that Phobos may be a product of a repeated satellite-ring cycle.</div><div>Uranus, which has a faster rotation rate falls within the Torque-Dependent regime.</div><div>Hypothesizing that a massive ring once orbited Uranus, I use RING-MOONS to demonstrate how the satellite Miranda may have formed from such a ring, and migrated outwards to its current orbit, but that any other satellites would have migrated inwards overtime.</div><div><br></div><div>Lastly, I examine Trans-Neptunian Objects (TNOs) in binary systems.</div><div>Tidal torques exerted on each body can decrease the mutual semi-major axis of the system.</div><div>I outline the conditions for which a fully synchronous system may experience a complete decay of the mutual orbit due to tidal torques.</div><div>As the semi-major axis decreases, it is possible for the smaller of the two bodies to shed mass before coming into contact with the more massive to form a contact binary.</div><div>I hypothesize that Chariklo and Chiron are contact binaries that formed via the tidal collapse of a binary TNOs system, and demonstrate how mass shedding may have occurred to form the rings observed today.</div>

Orientador: Ernesto Vieira Neto / Resumo: O estudo da formação e evolução do Sistema Solar é uma fonte de informação para entender sob quais condições a vida poderia surgir e evoluir. Nós apresentamos, nesta Tese de doutorado, um estudo numérico da fase final de acresção dos planetas gigantes do Sistema Solar durante e após a fase do disco de gás protoplanetário. Em nossas simulações, utilizamos um modelo recente e confiável para a formação de Urano e Netuno para esculpir as propriedades do disco trans-Netuniano original (Izidoro et al. , 2015a). Nós fizemos este estudo de uma maneira autoconsistente considerando os efeitos do gás e da evolução dos embriões planetários que formam Urano e Netuno por colisões gigantescas. Consideramos diferentes histórias de migração de Júpiter, devido a incerteza de como Júpiter migrou, durante a fase de gás. As nossas simulações permitiram obter pela primeira vez as propriedades orbitais do disco trans-Netuniano original. Então, calculamos o tempo de instabilidade dos planetas gigantes a partir de sistemas planetários que formam similares Urano e Netuno. Nossos resultados indicam fortemente que a instabilidade dos planetas gigantes acontecem cedo em até 500 milhões de anos e mais provável ainda ter acontecido em 136 milhões de anos após a dissipação do gás. Nós também realizamos simulações para discutir alguns efeitos dinâmicos que acontecem na região do cinturão de Kuiper. Estes efeitos acontecem quando Netuno esteve em alta excentricidade durante a instabilidade planetária. Para es... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: A study of the formation and evolution of the Solar System is a source of information for an understanding of what conditions life could arise and evolve. We present a numerical study of the final stage of accretion of the giant planets of the Solar System during and after the protoplanetary gas disc phase. In our simulations, we use a recent and reliable model for the formation of Uranus and Neptune to sculpt the properties of the original trans-Neptunian disk (Izidoro et al. , 2015a). We have done this study in a self-consistent way considering the effects of gas and the evolution of planetary embryos which form Uranus and Neptune by mutual giant collisions. We considered different Jupiter migration stories due to the uncertainty of how Jupiter’s migration was during the gas phase. Our simulations provide for the first time to obtain the orbital properties of the original trans-Neptunian disk. We then calculate the instability time of the giant planets from planetary systems which form similar Uranus and Neptune. Our results strongly indicate that the instability of the giant planets occurs early within 500 million years and even more likely to happen at 136 million years after gas dissipation. We also perform simulations to discuss some dynamical effects that happen in the Kuiper belt region. These effects happen when Neptune was in high eccentricity during planetary instability. For this problem, we use the simulations performed by Gomes et al. (2018) who investigated the... (Complete abstract click electronic access below) / Doutor

<p>Magnetic reconnection is a universal process occurring at boundaries between magnetized plasmas, where changes in the topology of the magnetic field lead to the transport of charged particles across the boundaries and to the conversion of electromagnetic energy into kinetic and thermal energy of the particles. Reconnection occurs in laboratory plasmas, in solar system plasmas and it is considered to play a key role in many other space environments such as magnetized stars and accretion disks around stars and planets under formation. Magnetic reconnection is a multi-scale plasma process where the small spatial and temporal scales are strongly coupled to the large scales. Reconnection is initiated rapidly in small regions by microphysical processes but it affects very large volumes of space for long times. The best laboratory to experimentally study magnetic reconnection at different scales is the near-Earth space, the so-called Geospace, where Cluster spacecraft <i>in situ</i> measurements are available. The European Space Agency Cluster mission is composed of four-spacecraft flying in a formation and this allows, for the first time, simultaneous four-point measurements at different scales, thanks to the changeable spacecraft separation. In this thesis Cluster observations of magnetic reconnection in Geospace are presented both at large and at small scales. </p><p>At large temporal (a few hours) and spatial (several thousands km) scales, both fluid and kinetic evidence of reconnection is provided. The evidence consist of ions accelerated and transmitted across the Earth’s magnetopause. The observations show that component reconnection occurs at the magnetopause and that reconnection is continuous in time. </p><p>The microphysics of reconnection is investigated at smaller temporal (a few ion gyroperiods) and spatial (a few ion gyroradii) scales. Two regions are important for the microphysics: the X-region, around the X-line, where reconnection is initiated and the separatrix region, away from the X-line, where most of the energy conversion occurs. Observations of a separatrix region at the magnetopause are shown and the microphysics is described in detail. The separatrix region is shown to be highly structured and dynamic even away from the X-line.</p><p>Finally the discovery of magnetic reconnection in turbulent plasma is presented by showing, for the first time, <i>in situ</i> evidence of reconnection in a thin current sheet found in the turbulent plasma downstream of the quasi-parallel Earth’s bow shock. It is shown that turbulent reconnection is fast and that electromagnetic energy is converted into heating and acceleration of particles in turbulent plasma. It is also shown that reconnecting current sheets are abundant in turbulent plasma and that reconnection can be an efficient energy dissipation mechanism.</p>

Magnetic reconnection is a universal process occurring at boundaries between magnetized plasmas, where changes in the topology of the magnetic field lead to the transport of charged particles across the boundaries and to the conversion of electromagnetic energy into kinetic and thermal energy of the particles. Reconnection occurs in laboratory plasmas, in solar system plasmas and it is considered to play a key role in many other space environments such as magnetized stars and accretion disks around stars and planets under formation. Magnetic reconnection is a multi-scale plasma process where the small spatial and temporal scales are strongly coupled to the large scales. Reconnection is initiated rapidly in small regions by microphysical processes but it affects very large volumes of space for long times. The best laboratory to experimentally study magnetic reconnection at different scales is the near-Earth space, the so-called Geospace, where Cluster spacecraft in situ measurements are available. The European Space Agency Cluster mission is composed of four-spacecraft flying in a formation and this allows, for the first time, simultaneous four-point measurements at different scales, thanks to the changeable spacecraft separation. In this thesis Cluster observations of magnetic reconnection in Geospace are presented both at large and at small scales. At large temporal (a few hours) and spatial (several thousands km) scales, both fluid and kinetic evidence of reconnection is provided. The evidence consist of ions accelerated and transmitted across the Earth’s magnetopause. The observations show that component reconnection occurs at the magnetopause and that reconnection is continuous in time. The microphysics of reconnection is investigated at smaller temporal (a few ion gyroperiods) and spatial (a few ion gyroradii) scales. Two regions are important for the microphysics: the X-region, around the X-line, where reconnection is initiated and the separatrix region, away from the X-line, where most of the energy conversion occurs. Observations of a separatrix region at the magnetopause are shown and the microphysics is described in detail. The separatrix region is shown to be highly structured and dynamic even away from the X-line. Finally the discovery of magnetic reconnection in turbulent plasma is presented by showing, for the first time, in situ evidence of reconnection in a thin current sheet found in the turbulent plasma downstream of the quasi-parallel Earth’s bow shock. It is shown that turbulent reconnection is fast and that electromagnetic energy is converted into heating and acceleration of particles in turbulent plasma. It is also shown that reconnecting current sheets are abundant in turbulent plasma and that reconnection can be an efficient energy dissipation mechanism.

Magnetized planets, such as Earth, are strongly influenced by the solar wind. The Sun is very dynamic, releasing varying amounts of energy, resulting in a fluctuating energy and momentum exchange between the solar wind and planetary magnetospheres. The efficiency of this coupling is thought to be controlled by magnetic reconnection occurring at the boundary between solar wind and planetary magnetic fields. One of the main tasks in space physics research is to increase the understanding of this coupling between the Sun and other solar system bodies. Perhaps the most important aspect regards the transfer of energy from the solar wind to the terrestrial magnetosphere as this is the main source for driving plasma processes in the magnetosphere-ionosphere system. This may also have a direct practical influence on our life here on Earth as it is responsible for Space Weather effects. In this thesis I investigate both the global scale of the varying solar-terrestrial coupling and local phenomena in more detail. I use mainly the European Space Agency Cluster mission which provide unprecedented three-dimensional observations via its formation of four identical spacecraft. The Cluster data are complimented with observations from a broad range of instruments both onboard spacecraft and from groundbased magnetometers and radars. A period of very strong solar driving in late October 2003 is investigated. We show that some of the strongest substorms in the history of magnetic recordings were triggered by pressure pulses impacting a quasi-stable magnetosphere. We make for the first time direct estimates of the local energy flow into the magnetotail using Cluster measurements. Observational estimates suggest a good energy balance between the magnetosphere-ionosphere system while empirical proxies seem to suffer from over/under estimations during such extreme conditions. Another period of extreme interplanetary conditions give rise to accelerated flows along the magnetopause which could account for an enhanced energy coupling between the solar wind and the magnetosphere. We discuss whether such conditions could explain the simultaneous observation of a large auroral spiral across the polar cap. Contrary to extreme conditions the energy conversion across the dayside magnetopause has been estimated during an extended period of steady interplanetary conditions. A new method to determine the rate at which reconnection occurs is described that utilizes the magnitude of the local energy conversion from Cluster. The observations show a varying reconnection rate which support the previous interpretation that reconnection is continuous but its rate is modulated. Finally, we compare local energy estimates from Cluster with a global magnetohydrodynamic simulation. The results show that the observations are reliably reproduced by the model and may be used to validate and scale global magnetohydrodynamic models.

The current paradigm in planet formation theory is developed around a hierarquical growth of solid bodies, from interstellar dust grains to rocky planetary cores. A particularly difficult phase in the process is the growth from meter-size boulders to planetary embryos of the size of our Moon or Mars. Objects of this size are expected to drift extremely rapid in a protoplanetary disk, so that they would generally fall into the central star well before larger bodies can form. In this thesis, we used numerical simulations to find a physical mechanism that may retain solids in some parts of protoplanetary disks long enough to allow for the formation of planetary embryos. We found that such accumulation can happen at the borders of so-called dead zones. These dead zones would be regions where the coupling to the ambient magnetic field is weaker and the turbulence is less strong, or maybe even absent in some cases. We show by hydrodynamical simulations that material accumulating between the turbulent active and dead regions would be trapped into vortices to effectively form planetary embryos of Moon to Mars mass. We also show that in disks that already formed a giant planet, solid matter accumulates on the edges of the gap the planet carves, as well as at the stable Lagrangian points. The concentration is strong enough for the solids to clump together and form smaller, rocky planets like Earth. Outside our solar system, some gas giant planets have been detected in the habitable zone of their stars. Their wakes may harbour rocky, Earth-size worlds.