The scattering trans-Neptunian Objects ( TNOs ) can be measured to smaller sizes than any other distant small-body population .
We use the largest sample yet obtained , 68 discoveries , primarily by the Outer Solar System Origins Survey ( OSSOS ) , to constrain the slope of its luminosity distribution , with sensitivity to much fainter absolute H magnitudes than previous work .
Using the analysis technique in Shankman et al .
( 54 ) , we confirm that a single slope for the H -distribution is not an accurate representation of the scattering TNOs and Centaurs , and that a break in the distribution is required , in support of previous conclusions .
A bright-end slope of \alpha _ { b } = 0.9 transitioning to a faint-end slope \alpha _ { f } of 0.4-0.5 with a differential number contrast c from 1 ( a knee ) to 10 ( a divot ) provides an acceptable match to our data .
We find that break magnitudes H _ { b } of 7.7 and 8.3 , values both previously suggested for dynamically hot Kuiper belt populations , are equally non-rejectable for a range of \alpha _ { f } and c in our statistical analysis .
Our preferred divot H -distribution transitions to \alpha _ { f } = 0.5 with a divot of contrast c = 3 at H _ { b } = 8.3 , while our preferred knee H -distribution transitions to \alpha _ { f } = 0.4 at H _ { b } = 7.7 .
The intrinsic population of scattering TNOs required to match the OSSOS detections is 3 \times 10 ^ { 6 } for H _ { r } < 12 , and 9 \times 10 ^ { 4 } for H _ { r } < 8.66 ( D \gtrsim 100 km ) , with Centaurs having an intrinsic population two orders of magnitude smaller .