PONDS ON EROS: PERT NEAR BUT NOT PLUMB?
Here, we examine the pixel scales of images available at the pond locations, and compare the observed distribution of ponds on Eros to that of the image pixel scale. We find a strong correlation between the density of ponds observed and the pixel scale in the best MSI images available at a given region. Over half the ponds are in regions imaged to 2.0 m/px or better, in total comprising only one eighth of the surface area. We suggest that there is a significant observational bias in the data at small pond sizes (< 30 m diameter), and that there may be many more unidentified small ponds in the regions imaged to coarser pixel scales.
We also investigate the topography of these ponds on Eros using a new shape model derived from stereophotoclinometric analysis, and validated against altimetry from the NEAR Laser Rangefinder, in order to constrain the mode of pond formation from existing models. We update the locations of 55 pond “candidates” identified in images registered to the new shape model. We classify the flatness of these features according to the behavior of the first and second derivatives of the topography. We find that less than half of pond candidates have clearly flat floors. Based on the pond topography, we favor an external origin for the ponds deposits. We suggest that fine dust may be transported into bounding depressions by electrostatic levitation, but may adhere to slopes, and that seismic shaking may not be sufficient to bring the deposits to an equipotential surface. Disaggregation of a central boulder should result in an obvious break in slope, and this is only observed in roughly half the pond candidates.