Jump to Main Content
Characterization of canopy fuels using ICESat/GLAS data
- García, Mariano, Popescu, Sorin, Riaño, David, Zhao, Kaiguang, Neuenschwander, Amy, Agca, Muge, Chuvieco, Emilio
- Remote sensing of environment 2012 v.123 pp. 81-89
- biomass, bulk density, canopy, data collection, energy, fire behavior, fuels (fire ecology), leaf area index, leaves, lidar, models, remote sensing
- This study aimed to estimate canopy fuel properties relevant for crown fire behavior using ICESat/GLAS satellite LiDAR data. GLAS estimates were compared to canopy fuel products generated from airborne LiDAR data, which had been previously validated against field data. The geolocation accuracy of the data was evaluated by comparing ground elevation on both datasets, showing an offset of 1 pixel (20m). Canopy cover (CC) was estimated as the ratio of the canopy energy to the total energy of the waveform. Application of a canopy base height threshold (CBH) to compute the canopy energy increased the accuracy of CC estimates (R²=0.89; RMSE=16.12%) and yielded a linear relationship with airborne LiDAR estimates. In addition, better agreement was obtained when the CC derived from airborne LiDAR data was estimated using the intensity of the returns. An empirical model, based on the CC and the leading edge (LE), was derived to estimate leaf area index (LAI) using stepwise regression providing good agreement with the reference data (R²=0.9, RMSE=0.15). Canopy bulk density (CBD) was estimated using an approach based on the method developed by Sando and Wick (1972) to derive CBD from field measurements, and adapted to GLAS data. Thus, foliage biomass was distributed vertically throughout the canopy extent based on the distribution of canopy material and CBD was estimated as the maximum 3m-deep running mean considering layers with a thickness of 15cm, which is the vertical resolution of the GLAS data. This approach gave a coefficient of determination of 0.78 and an RMSE of 0.02kgm⁻³.