Data on the UT 2005 June 14 mini-outburst of comet 9P/Tempel 1 taken from different viewpoints have been examined for morphological differences and parallax. The data were taken with the Hubble Space Telescope (HST), from the Deep Impact (DI) spacecraft, and from the Calar Alto observatory, Spain. The mini-outburst source region was found to be located near 218+/-6E, 6+/-5N on the Deep Impact nucleus shape model. The mini-outburst occurred at ~12 pm local solar time. The distribution of light in the mini-outburst is similar to that expected for an ejecta curtain. 

The method and software used to determine the surface location was checked using position angles of the impact ejecta plume as seen from DI and HST. The general region of impact was recovered and a downrange tilt of the ejecta curtain axis of 10.2 deg from the surface normal was found. 

We computed tracks of possible source regions for nine other mini-outbursts seen from DI. Five of these tracks converge on the 2005 June 14 event location. Three of the tracks converge at a second location near (60E, 20S), well separated from the first. Multiple mini-outbursts arise at each location either from a single source or from a few sources in close proximity. The mini-outbursts occur both at night and during the day indicating at most weak, if any, control by direct sunlight. The times of outburst are non-random with a preference for early-afternoon, dusk and midnight. None of the mini-outbursts occurred near dawn. They occur at low latitudes (between +/- 40 deg) near the points where the principal axis of minimum moment of inertia cuts the surface. These regions are furthest from the center of figure and have the lowest effective surface gravity.

We use these results to develop a conceptual model of the mini-outburst process and make comparisons with the theoretical calculations. We find that the tensile strength of the sub-surface material must be very low (e.g., ~102 103 dynes.cm-2) and, on the basis of features imaged on the western facet of the nucleus, suggest that inflation of the subsurface may be occurring. Our model makes specific predictions about the kind of surface morphology that should result from mini-outburst activity. We show that one of the isolated rimless depressions and the close-packed depressions found in the Deep Impact images have the properties needed and identify them as possible sites of past and current mini-outburst activity.