Bi-temporal ALS assessment of vertical-horizontal forest structures and their structural association in a temperate urban forest

This study utilized high-density, bi-temporal airborne laser scanning (ALS) data (late-autumn and early-spring) to quantify vertical and horizontal structural features of four vegetation layers (shrub, subtree, tree, and big-tree) in urban forests of Gwacheon-si, Republic of Korea. Field validation against 1,074 polygons showed that ALS-derived subtree coverage had the lowest bias, with overestimation in the tree layer and underestimation in the shrub layer. Clear species- and season-specific patterns emerged: deciduous species such as Castanea crenata showed substantial canopy loss in the early-spring period, whereas evergreen species like Pinus densiflora remained relatively stable across seasons. Significant positive associations between ALS-derived vertical point distribution and horizontal coverage were observed, particularly in the tree and big-tree layers, and these persisted after controlling for mean canopy height (partial r = 0.75-0.98 in the big-tree layer). Structural features consistently differentiated into two functional groups - an upper canopy group (tree/big-tree) and a lower strata group (subtree/shrub) - each with distinct correlation patterns and seasonal sensitivities. By resolving these associations across species, seasons, and layers, this study demonstrates the value of bi-temporal ALS for species-specific structural assessment in temperate urban forests.