1. Dynamical modeling of the DI dust ejecta cloud Tanyu Bonev (Institute of Astronomy and National Astronomical Observatory, Bulgaria) and the ESO DI observing team Publication: Deep Impact as a World Observatory Event: Synergies in Space, Time, and Wavelength, ESO Astrophysics Symposia. ISBN 978-3-540-76958-3. Springer Berlin Heidelberg, 2009, p. 177

2. The ESO DI observing team

3. Outline 1. Images of the ejecta cloud. 2. Initial guess for the velocity size dependence. 3.The location of the impact site. 4.Monte Carlo model. 5.Derivation of the particles size distribution and of the total mass in the ejecta plume. 6.Comparison with photometry from the first hours. 7.Conclusions

4. 4.972 (+17.8) 5.995 (+42.4) 6.993 (+66.3)7.955 (+89.4) Post-impact minus pre-impact images FORS2@VLT R-band Scale: 0.25 arcsec/px corresponds to: 162 km/px – 05.07. 167 km/px – 08.07.

5. Velocities and accelerations: initial guess v = a t v = √(2 a d) סּ d: apex distance a: radiation pressure acceleration v: initial velocity

7. Constraints on the impact site location Position angle of the ejecta plume motion. Cometary equator. Rotation axis orientation: R.A. = 293.8 deg DEC = 72.6 deg latitude of impact (cometocentric coordinates) M. A’Hearn, priv. communication (Thomas et al., Icarus)

8. Selection of the “best” impact site location

9. The model 1 million dust particles are emitted for a period of 20 minutes starting at the moment of the impact. These particles are distributed in time, space and particle sizes as follows: 100 emission events 200 emission directions randomly distributed in a cone with full opening angle of 180 degree. particles of radii in the range from 0.1 to 100 micrometer are used, distributed logarithmically in 51 bins. In a process of trial and error the position of the source (impact site) and the velocity law are adjusted. Final task: to find the particle size distribution and the related quantities.

10. Solution by linear regression: B(x,y) = ∑ K i * S i (x,y)) i = 0, 50 S i (x,y) is the scattering area produced by the particles of one particular size, i. Initially, S is calculated with an adopted particle size distribution. The solution, Ki, represents the final PSD. The shown particle distributions are for the first post-impact observation.

11. Observation Model solution Simulation of the ejecta plume observed 18 hours after the impact

13. Parameters found from the modeling (the results are derived from the fit of the model to the first post-impact observation, +18 hours) 4600 ton water were created by the impact ( Kueppers et al. 2005, Nature, Vol. 437 ). Dust to water ratio approximately 3.

14. Comparison of our results with photometry from the first post-impact hours CFTH + Megacam data, Jana Pittichova et al., 2005, ACM’2005. Brightness decrease with the the velocity law used in the Monte Carlo model, calculated for 4 different particle size distributions. The data are normalized to their maxima and scaled to the Megacam measurements.

15. Comparison of our results with photometry from the first post-impact hours Brightness decrease with the velocity law used in the Monte Carlo model, calculated for 4 different particle size distributions. The data are normalized to their maxima. Light curve of the cometary dust obtained with OSIRIS. Kueppers et al. 2005, Nature, Vol. 437

16. 1.The ejected dust plume is described by a dynamical model 2.The total amount of dust derived is ≈ 12000 ton. 3.The particle size distribution can be described by a power law with index -3 +/- 0.2 4.The velocities used in the model and their size distribution indicate acceleration of the particles by the gas in the coma. 5. The results derived from dynamical modeling of the ejecta cloud days after the impact are consistent with the photometry from the first hours. Conclusions

17. 1.The ejected dust plume is described by a dynamical model 2.The total amount of dust derived is ≈ 12000 ton. 3.The particle size distribution can be described by a power law with index -3 +/- 0.2 4.The velocities used in the model and their size distribution indicate acceleration of the particles by the gas in the coma. 5. The results derived from dynamical modeling of the ejecta cloud days after the impact are consistent with the photometry from the first hours.