Context. The issue of the long term dynamics of Jupiter family comets (JFCs) involves uncertain assumptions about the physical evolution and lifetimes of these comets. Contrary to what is often assumed, real effects of secular dynamics cannot be excluded and therefore merit investigation. Aims. We use a random sample of late heavy bombardment cometary projectiles to study the long-term dynamics of JFCs by a Monte Carlo approach. In a steady-state picture of the Jupiter family, we investigate the orbital distribution of JFCs, including rarely visited domains like retrograde orbits or orbits within the outer parts of the asteroid main belt. Methods. We integrate 100 000 objects over a maximum of 100 000 orbital revolutions including the Sun, a comet, and four giant planets. Considering the steady-state number of JFCs to be proportional to the total time spent in the respective orbital domain, we derive the capture rate based on observed JFCs with small perihelia and large nuclei. We consider a purely dynamical model and one where the nuclei are eroded by ice sublimation. Results. The JFC inclination distribution is incompatible with our erosional model. This may imply that a new type of comet evolution model is necessary. Considering that comets may live for a long time, we show that JFCs can evolve into retrograde orbits as well as asteroidal orbits in the outer main belt or Cybele regions. The steady-state capture rate into the Jupiter family is consistent with ∼1 × 10 9 scattered disk objects with diameters D > 2 km. Conclusions. Our excited scattered disk makes it difficult to explain the JFC inclination distribution, unless the physical evolution of JFCs is more intricate than assumed in standard, erosional models. Independent of this, the population size of the Jupiter family is consistent with a relatively low-mass scattered disk.

Context. Unraveling the events that took place in the solar system during the period known as the late heavy bombardment requires the interpretation of the cratered surfaces of the Moon and terrestrial planets. This, in turn, requires good estimates of the statistical impact probabilities for different source populations of projectiles, a subject that has received relatively little attention, since the works of Öpik (1951, Proc. R. Irish Acad. Sect. A, 54, 165) and Wetherill (1967, J. Geophys. Res., 72, 2429. Aims. We aim to work around the limitations of the Öpik and Wetherill formulae, which are caused by singularities due to zero denominators under special circumstances. Using modern computers, it is possible to make good estimates of impact probabilities by means of Monte Carlo simulations, and in this work, we explore the available options. Methods. We describe three basic methods to derive the average impact probability for a projectile with a given semi-major axis, eccentricity, and inclination with respect to a target planet on an elliptic orbit. One is a numerical averaging of the Wetherill formula; the next is a Monte Carlo super-sizing method using the target's Hill sphere. The third uses extensive minimum orbit intersection distance (MOID) calculations for a Monte Carlo sampling of potentially impacting orbits, along with calculations of the relevant interval for the timing of the encounter allowing collision. Numerical experiments are carried out for an intercomparison of the methods and to scrutinize their behavior near the singularities (zero relative inclination and equal perihelion distances). Results. We find an excellent agreement between all methods in the general case, while there appear large differences in the immediate vicinity of the singularities. With respect to the MOID method, which is the only one that does not involve simplifying assumptions and approximations, the Wetherill averaging impact probability departs by diverging toward infinity, while the Hill sphere method results in a severely underestimated probability. We provide a discussion of the reasons for these differences, and we finally present the results of the MOID method in the form of probability maps for the Earth and Mars on their current orbits. These maps show a relatively flat probability distribution, except for the occurrence of two ridges found at small inclinations and for coinciding projectile/target perihelion distances. Conclusions. Our results verify the standard formulae in the general case, away from the singularities. In fact, severe shortcomings are limited to the immediate vicinity of those extreme orbits. On the other hand, the new Monte Carlo methods can be used without excessive consumption of computer time, and the MOID method avoids the problems associated with the other methods.

Context. The orbits of one-apparition comets discovered in the early part of the last century have formerly been determined with very different numerical methods and assumptions on the model of the solar system, including the number of planets taken into account. Moreover, observations of the comet-minus-star-type sometimes led to determination of the comet position that are less precise than what we can derive today by using a more modern star catalogue. Aims. We aim to provide a new catalogue of cometary orbits that are derived using a completely homogeneous data treatment, accurate numerical integration, and a modern model of the solar system. Methods. We collected the complete sets of observations for investigated comets from the original publications. Then we recalculated the cometary positions for the comet-minus-star-type of observations using the Positions and Proper Motions Star Catalogue, and applied a uniform method for the data selection and weighting. As a final result, new osculating orbits were determined. Secondly, dynamical calculations were performed to the distance of 250 AU from the Sun to derive original and future barycentric orbits for evolution backward and forward in time. These numerical calculations for a given object start from a swarm of virtual comets constructed using our osculating (nominal) orbit. In this way, we obtained the orbital element uncertainties of original and future barycentric orbits.Results. We present homogeneous sets of orbital elements for osculating, original, and future orbits for 38 one-apparition comets. Non-gravitational orbits are derived for thirteen of them.

Context. The Nice model predicts that the trans-planetary planetesimal disk made a large or even dominant contribution to the cratering in the inner solar system during the late heavy bombardment (LHB). In the presence of evidence that lunar craters and mare basins may be mainly of asteroidal origin, there is a dilemma of the missing comets that is not yet resolved. Aims. We aim to revisit the problem of cometary impact rates on the Moon and the terrestrial planets during the LHB with a flexible model, allowing us to study the influences of physical destruction of comets, the mass of the primordial disk, and the distribution of this mass over the entire size range. Methods. We performed a Monte Carlo study of the dynamics of the cometary LHB projectiles and derive the impact rates by calculating individual collision probabilities for a huge sample of projectile orbits. We used Minimum Orbit Intersection Distances (MOIDs) according to a new scheme introduced here. Different calculations were performed using different models for the physical evolution of comet nuclei and for the properties of the primordial, trans-planetary disk. Results. Based on the capture probability of Jupiter Trojans, we find a best fit radius of the largest LHB comet impacting the Moon for a low-mass primordial disk. For this disk mass, the LHB cratering of the Moon, Mercury and Mars were dominated by asteroids. However, some smaller lunar maria were likely preceded by comet impacts. The volatile delivery to the Earth and Mars by LHB comets was much less than their water inventories. Conclusions. There is no excessive cometary cratering, if the LHB was caused by a late planetary instability in the Nice Model. The Earth and Mars obtained their water very early in their histories. The Noachian water flows on Mars cannot be attributed to the arrival of LHB-related H 2 O or CO 2 .