Abstract
ABSTRACTAirborne laser scanning (ALS) data and digital hemispherical photos (DHP) from 93 sample plots in Laeva test site, Estonia, were used to study effects of phenology and scan angle on the ALS-based canopy cover (CCALS) estimates. The relative share of first returns (P1/A) for 6185 forest stands was analysed. The CCALS was calculated using different height thresholds and echoes, and was compared with the CC estimates based on DHP (CCDHP) and crown model (CCRCrown). The first of many echoes-based canopy cover estimate (CCALS,1.3_1) saturated at values greater than 80%. The strongest correlation of CCDHP was found with CCALS,1.3_A using all echoes and a 1.3 m height break (R2 = 0.81, RMSE = 11.8%). Correcting the estimate for view nadir angle did not improve the correlation of CCALS,1.3_A with CCDHP. The CCRCrown had a weak correlation (R2 < 0.25) with CCALS and with CCDHP. The P1/A was not influenced by tree species composition, but by phenology, stand relative density and forest height; however, CCALS was not dependent on stand height. Foliage phenology had a substantial effect on CCALS and CCDHP. In dense mixed-species forests, we recommend to use all returns for canopy cover estimation.
Highlights
Airborne laser scanning (ALS) has become a widely used remote sensing method for assessing forest structure variables in operational forestry (Andersen, McGaughey & Reutebuch, 2005)
When after the final leaf unfolding stage (aFLU) and before bud swelling with leaves off (bBS) measurements were tested separately, the smallest root mean square error (RMSE) was found for CCALS,1.3_A and canopy cover (CC) estimates based on DHP (CCDHP) of aFLU flight and the R2 corresponded to a moderate correlation
We analysed discrete-return ALS data from dense deciduous broadleaf-dominated mixed forests for CC estimation using references obtained from tree crown models and digital hemispherical photos (DHP)
Summary
Airborne laser scanning (ALS) has become a widely used remote sensing method for assessing forest structure variables in operational forestry (Andersen, McGaughey & Reutebuch, 2005). Vertical canopy cover (CC) is the key factor for most of these estimates and when combined with other forest structure parameters it is used for leaf-area index (LAI) mapping (Korhonen & Morsdorf, 2014; Solberg et al, 2009) or could be used for planning of thinnings in forest management (Vastaranta et al, 2011). Forest canopy structure is described with mean tree height (H), crown length and the ratio of crown length to H, crown cover, effective CC and angular canopy closure. Distinguishing between different CC estimates is essential for forest structure studies (Jennings, Brown, & Sheil, 1999; Korhonen, Korpela, Heiskanen, & Maltamo, 2011).
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
