Space Research & Planetary Sciences Division Web Site

DRAG - Dynamics of Rarefied Gas

We are particularly interested in advancing our understanding of dynamic phenomena occurring on planetary or cometary surfaces containing ices and volatile species. In support to the Rosetta mission, we perform numerical modelling of dusty-gas outflow from the surfaces of comets using Direct Simulation Monte Carlo methods.

Cometary Science

Dusty-Gas Outflow from Cometary Nuclei

In general, numerical simulations of the outflow from the cometary nucleus tend to be simplified assuming axisymmetric geometry with respect to the Sun and solving the Navier-Stokes equations to derive a gas and dust distribution (e.g. Keller et al., 1994; Crifo and Rodionov, 1999). Because of its small size (typically <10 km in diameter) the cometary nucleus and the innermost coma is best investigated by spacecraft. In 2001, NASA's Deep Space 1 spacecraft flew within 2000 km of the nucleus of comet 19P/Borrelly and acquired images with the MICAS imaging system (Soderblom et al., 2002). These data immediately showed the limitations of current models. The nucleus was highly irregular in shape and the dust emission highly non-isotropic. The data set was limited to only a few good images of the dust coma. Constraining model parameters under these conditions is difficult. Ho et al. (2005) have studied these images and concluded that improved fits to the inner coma distribution are only possible if an acceleration profile for the dust similar to that found in the numerical models of Tenishev and Combi (2003 and priv. communication) is used. Ho (2005) has also attempted to link the structures seen in the coma to those seen in simultaneous ground-based observations. A numerical model of the dust emission has been established which allows us to input the emission parameters seen at the source and extrapolate them to simulate their appearance to a ground-based observer. However, from this work it is evident that further progress can only be made by detailed modelling of the outflow in close collaboration with image analysis (cf Crifo et al., 2002b). The dust scattering properties also play an important role which we ourselves have recently been addressing  (Bertini et al., 2006).

For the gas outflow, modellers have moved towards using the Direct Simulation Monte Carlo (DSMC) approach (Bird, 1994; Davidsson and Skorov, 2004; Crifo et al., 2002a, 2005). This has significant advantages in that low gas outflow rates can be examined. This will be of particular importance for the Rosetta mission which will observe comet Churyumov-Gerasimenko from an almost inert state through to perihelion covering a wide range of gas outflow regimes.

We are developing coma simulations to help us interpret measurements from Rosetta. In the past 2 years, a sensitivity analysis method with a minimum number of simulation runs has been developed. The “Polynomial Chaos Expansion” (PCE) method was developed for uncertainty propagation but was recently used for sensitivity analyses. The method allows one to make statements about surface production rates if the position and size of the interaction region (between strong source and weak background) can be established by in-situ measurements. An approach to using PCE to establish non-linear relations between source parameters and the resulting flow field has been published (Finklenburg and Thomas, 2014).

Effects of the complex topography of comet nuclei on the flow field can only be studied in a 3D DSMC simulation. A collaboration with the group of J.-S. Wu (National Chiao Tung Uni., Taiwan) has been initiated and boosted considerably through an SNSF-funded exploratory workshop in Bern in January 2013. They have developed a 3D DSMC code, which runs on a parallel computer (PDSC+). The shape model of Comet 9P/Tempel 1 (Thomas et al., 2007) has been used as a first test case for PDSC+ in Bern. The results show that a production rate proportional to the local surface temperature does not explain the observed gas distribution. Non-uniform emission models have been constructed that agree better with observation (Finklenburg et al., 2014).

Relevant Publications:

Liao Y., C.C. Su, R. Marschall, J.S. Wu, M. Rubin, I.L. Lai, W.-H. Ip, H.U. Keller, J. Knollenberg, E. Kührt, Y. Skorov, and N. Thomas, (2016), 3D Direct Simulation Monte Carlo Modelling of the Inner Gas Coma of Comet 67P/Churyumov-Gerasimenko: A Parameter Study, Earth, Moon, and Planets, accepted, 28 February 2016.

Marschall, R., C.C. Su, Y. Liao, N. Thomas, K. Altwegg, H. Sierks, W.-H. Ip, H.U. Keller, J. Knollenberg, E. Kührt, I.L. Lai, M. Rubin, Y. Skorov, J.S. Wu, L. Jorda, F. Preusker, F. Scholten, A. Gracia Berná, A. Gicquel, G. Naletto, X. Shi, and J.-B. Vincent, (2016), Modelling of the observations of the inner gas and dust coma of comet 67P/Churyumov-Gerasimenko – First results, Astron. Astrophys., accepted , 27 February 2016

Liao, Y., C.C. Su, S. Finklenburg, M. Rubin, W.-H. Ip, H.U. Keller, J. Knollenberg, E. Kührt, L.I. Lai, Y. Skorov, N. Thomas, J.S. Wu, and Y.S. Chen, (2014), 3-D DSMC Simulations of Comet 67P/Churyumov-Gerasimenko, Lunar and Planetary Institute Science Conference Abstracts, 45, 1764.

Finklenburg, S., N. Thomas, C.C. Su , and J.-S. Wu, (2014), The spatial distribution of water in the inner coma of comet 9P/Tempel 1: Comparison between models and observations, Icarus, accepted, 24 March 2014.

Finklenburg, S. and N. Thomas, (2014) Relating in situ gas measurements to the surface outgassing properties of cometary nuclei, Planetary and Space Science, accepted, 11 February 2014, doi:10.1016/j.pss.2014.02.005.

Finklenburg, S., N. Thomas, J. Knollenberg, and E. Kührt, (2011), Comparison of DSMC and Euler Equations Solutions for Inhomogeneous Sources on Comets, 27th International Symposium on Rarefied Gas Dynamics, 2010 AIP Conf. Proc. 1333, 1151-1156, doi: 10.1063/1.3562799.

Thomas, N. (2009) The nuclei of Jupiter family comets: A critical review of our present knowledge, Planetary and Space Science, 57, 1106-1117. 

Cometary models for the Rosetta target

The C-G shape model of Lowry et al. (2012) has been imported and first runs with a homogeneous outgassing indicate that the code is running successfully within a manageable time (see Figure). Considerable time has been invested in optimizing the grid and reducing the collision distance to mfp ratio by implementation of the transient adaptive subcell method (Su et al., 2010).

Optical remote-sensing instruments on cometary missions probe the structure of the inner coma by measuring sunlight scattered by emitted dust. Modelling of the ejection and acceleration of gas and dust can provide the link between the observable dust distribution and processes occurring at the surface of the nucleus that are driven by the sublimation of ice when it is exposed to solar irradiation.

It is expected that ESAs (European Space Agency) spacecraft Rosetta will rendezvous with Comet 67P/Churyumov-Gerasimenko and observe the comet’s nucleus and coma from mid-2014 to the beginning of 2016. During the mission, the spacecraft will accompany the comet from around 4 AU (astronomical units) from the Sun (when gas production rates are expected to be low) through to perihelion (at about 1.3 AU) when production rates may reach 1028 molecules per second. In order establish observation and data analysis strategies for Rosetta and to interpret the resulting measurements, a model for the gas-dust interaction of Comet 67P is necessary. Preliminary simulations of the outgassing of Comet 67P are done based on DSMC (Direct Simulation Monte Carlo) method. By setting up models and implementing simulations, the properties of gas and dust (ex. the gas flow-field in the innermost coma and the dust acceleration by gas drag) of the comet can be thus well studied.

  • Some Text goes here as a caption.

The Surface Layer

For the comet rendezvous phase, the dusty-gas outflow (within the first few 100 m) and its relationship to the surface layer (including surface topography and the temperature structure) is of prime importance. There is now considerable debate about the structure of the upper layer. The latest data from the Deep Impact spacecraft (A'Hearn et al., 2005) showed that 9P/Tempel 1 has significant inhomogeneity in shape, surface appearance, local topography, and colour. In addition, the first observations of the gas distribution directly above the nucleus surface have been reported (Feaga et al., 2007). The H2O column density was at a maximum close to the sub-solar point with a distribution which suggests emission proportional to the solar irradiance. However, this work has also provided the remarkable conclusion that the distribution of CO2 above the nucleus differs markedly from that of H2O. The surface temperature of 9P/Tempel 1 at 1.5 AU [Groussin et al., 2007] showed a maximum temperature of 336±7 K. Even in regions where water ice was detected on the surface of 9P/Tempel 1 by IR spectroscopy [Sunshine et al., 2006], the observed temperature exceeded 270 K at an image scale of 120 m px-1. The data have allowed an estimate of the thermal inertia of <50 W m-2 s-1/2 K-1 [Groussin et al., 2007] although this has been challenged by Davidsson et al. [2009]. The differences have major implications for heat transfer in the upper layers of the nucleus [Thomas, 2009]. A thermal model has therefore been developed and linked to the PDSC++ inputs to test the implications for the inner coma gas distribution. 

People working in DRAG

Space Research & Planetary Sciences Division Web Site