The camera was down for the early part of 2020, but otherwise it seems like the understory count of # of pixels with gcc > 0.4 doesn't really lead the overstory gcc_90 in the deciduous broadleaf (DB) region of interest. Will test different thresholds next between 0.3 and 0.4.
Here is a comparison of different gcc thresholds. 0.35 and 0.36 look especially promising for detecting the lead between understory and canopy. Should I to other NEON forest sites to confirm?
The camera was down for the early part of 2020, but otherwise it seems like the understory count of # of pixels with gcc > 0.4 doesn't really lead the overstory gcc_90 in the deciduous broadleaf (DB) region of interest. Will test different thresholds next between 0.3 and 0.4.