From \citet{Cheng_2017}

New Horizons re- sults showed that the 1988 occultation light curve inflection is pri- marily explained by the thermal gradient model: the steep near- surface temperature gradient ∼2–5 K/km is present, the Pluto ra-dius is 1189.9 ±0.4 km, and the vertical optical depth of Pluto haze is only ∼0.013 ( Gladstone et al. 2016 )
"The haze in Pluto’s atmosphere is optically thin, with the
brightness approximated as
I/F = ∼ P ( θ ) n σs/4 √ (2 πRH) (1)
( Gladstone et al., 2016 ) where P ( θ) is scattering phase function at phase angle θ, n is number density of haze particles, σs is the scat- tering cross section, R is Pluto radius, and H is scale height.
The phase function is a key diagnostic of physical properties including size, shape, and composition (refractive index). The brightness ver- sus phase angle is considered here, by comparing I / F obtained in back scatter and forward scatter geometries as listed in Table 1 . As haze can be accurately measured only off the limb of Pluto, these observations were made over various locations on Pluto dur- ing the 19 h time span of Table 1 . The phase function was com- piled using only haze observations over the day side limb at simi- lar northern latitudes ( > 40 °)." 
Comment: So the haze phase function obtained here could be accpeted to be in consistent to a degree with itself at different phase angles, because as discussed in this paper there is spatial variation in the haze distribution in plutos atmosphere.
From Table 4 and equation 1 the haze properties are calulcated based on the assumption that they are either spherical mie particles or fractals aggregates of spheres. In eand equation 1 the haze properties are calulcated based on the assumption that they are either spherical mie particles or fractals aggregates of spheres.