THE LUNAR ROCK SIZE FREQUENCY DISTRIBUTION DERIVED FROM THERMAL INFRARED OBSERVATIONS (Invited Presentation)

The present-day rock size frequency distribution (SFD) on the Moon offers a snapshot of the evolution of the lunar surface over time. Bombardment of the surface by impactors of all sizes has been the dominant process modifying the lunar surface for billions of years. This both excavates rocks from the subsurface and degrades and/or buries rocks on the surface ultimately leading to the formation of fine-grained regolith. Current remote sensing methods are only sensitive to surface rocks larger than 1 m, so little is known about the timescale or processes by which ~1 m rocks degrade into fine-grained regolith. Knowledge of rocks smaller than 1 m is also important for assessing hazards at landing sites.

Here I present a new approach that enables both the detection of smaller rocks than previously possible and the relative distribution of rock sizes. This is possible both because of the comprehensive temporal coverage by the Diviner Lunar Radiometer Experiment (Diviner) on the Lunar Reconnaissance Orbiter (LRO) and through the use of three-dimensional (3D) thermal modeling of rocks on the lunar surface. Coherent blocks of rock have a higher thermal inertia than fine-grained regolith, so rocks cool more slowly over the course of the night. Warm temperatures have a greater radiance contribution at shorter wavelengths, so multi-spectral observations enable the calculation of subpixel rock abundance. Previous work used rock temperatures from a 1D thermal model to derive rock abundance and regolith temperature from Diviner data (Bandfield et al., 2011). Ten years later, Diviner has now observed nearly every location on the Moon at every hour of the lunar night, so the cooling behavior of the surface can be used as an additional constraint on models. Furthermore, small rocks are able to cool more efficiently than large rocks due to their higher surface area to volume ratio. I conducted thermal modeling in 3D to derive the nighttime cooling of rocks ranging in size from 1 mm to 1 m. Then I assumed the rock SFD follows an exponential curve where rock frequency decreases with increasing rock size and using Diviner observations, I found the best fit total rock abundance and rate of drop-off at large rock sizes. Preliminary results reveal many small rocks in the lunar highlands that had previously evaded detection, and show that lunar maria have more large rocks than the highlands consistent with their more recent resurfacing.