The size-frequency distribution of H>13 NEOs and ARM targets detected by Pan-STARRS1

We determine the absolute magnitude (H) distribution (or size-frequency distribution, SFD; $N(H) propto 10^{alpha H}$ where $alpha$ is the slope of the distribution) for near-Earth objects (NEO) with $13<H<30$ and Asteroid Retrieval Mission (ARM) targets with $27<H<31$ that were detected by the 1st telescope of the Panoramic Survey Telescope and Rapid Response System - Pan-STARRS1 (e.g. Kaiser et al. 2002, Kaiser 2004, Hodapp et al. 2004). The NEO and ARM target detection efficiencies were calculated using the Greenstreet et al. (2012) NEO orbit distribution. The debiased Pan-STARRS1 NEO absolute magnitude distribution is more complex than a single slope power law - it shows two transitions - at H$sim$16 from steep to shallow slope, and in the $21<H<23$ interval from a shallow to steep slope, which is consistent with other recent works (e.g. Mainzer et al. 2011c, Brown et al. 2013, Harris and D`Abramo 2015). We fit $alpha = 0.48pm0.02$ for NEOs with $13<H<16$, $alpha = 0.33pm0.01$ for NEOs with $16<H<22$, and $alpha = 0.62pm0.03$ for the smaller objects with $H>22$. There is also another change in slope from steep to shallow around H=27. The three ARM target candidates detected by Pan-STARRS1 in one year of surveying have a corrected SFD with slope $alpha = 0.40^{+0.33}_{-0.45}$. We also show that the window for follow up observations of small (H$gtrsim$22) NEOs with the NASA IRTF telescope and Arecibo and Goldstone radars are extremely short - on order of days, and procedures for fast response must be implemented in order to measure physical characteristics of small Earth-approaching objects. CFHTs MegaCam and Pan-STARRS1 have longer observing windows and are capable of following-up more NEOs due to their deeper limiting magnitudes and wider fields of view.