Novel forest structure metrics from airborne LiDAR data for improved snow interception estimation

Abstract

A set of 57 novel canopy metrics of potential value for snow modeling were created from airborne LiDAR data. The metrics were meant to estimate size and relative location of gap openings around a point within forested areas, allowing for measures of the spatial arrangement of surrounding canopy elements. These new metrics were correlated with snow interception measured in the field (8488 manually measured snow interception points). The results were further compared to the correlation values between effective interception and traditionally used forest parameters (CC and LAI). The correspondence between all metrics was also analyzed in order to understand the potential cross correlation between each variable. LAI (average R: 0.57) demonstrated low correlations when directly compared to snow interception and further showed a large cross correlation with canopy closure (average R: 0.72). A new metric, ‘mean distance to canopy’ had the highest correlation (average R: 0.78) over all storm events to the effective interception. But in contrast to LAI, this metric did not show any cross correlation with canopy closure (CC). Likewise, ‘total gap area,’ an indirect measurement of apparent gap fraction (another new metric), also showed a high correlation to effective interception (average R: 0.72) without demonstrating a significant cross correlation to CC or to mean distance to canopy. These findings suggest that modeling forest snow processes with both CC and LAI may not be the best option due to both the low correlation of LAI as well as high cross correlation between these parameters. However, the pairing of mean distance to canopy and/or total gap area with canopy closure could give more robust estimations of snow interception within heterogeneous terrain.