Testing different methods of forest height and aboveground biomass estimations from ICESat/GLAS data in Eucalyptus plantations in Brazil
Baghdadi, N. ; Le Maire, G. ; Fayad, I. ; Bailly, J.S ; Nouvellon, Y. ; Lemos, C. ; Hakamada, R.
Type de document
Article de revue scientifique à comité de lecture
Affiliation de l'auteur
IRSTEA MONTPELLIER UR LCMO UMR TETIS FRA ; CIRAD MONTPELLIER PERSYST UMR ECO&SOLS FRA ; IRSTEA MONTPELLIER UR LCMO UMR TETIS FRA ; AGROPARISTECH UMR TETIS MONTPELLIER FRA ; CIRAD MONTPELLIER PERSYST UMR ECO&SOLS FRA ; INTERNATIONAL PAPER DO BRASIL BRA ; INTERNATIONAL PAPER DO BRASIL BRA
Résumé / Abstract
The Geoscience Laser Altimeter System (GLAS) has provided a useful dataset for estimating forest heights in many areas of the globe. Most of the studies on GLAS waveforms have focused on natural forests and only a few were conducted over forest plantations. This work set out to estimate the stand-scale dominant height and aboveground biomass of intensively managed Eucalyptus plantations in Brazil using the most commonly used models developed for natural forests. These forest plantations are valuable case studies, with large and numerous stands that are very uniform, in which field measurements are precise compared to natural forests. The height of planted Eucalyptus forest stands estimated from waveforms acquired by GLAS were compared with in situ measurements in order to determine the model that produced the best forest height estimates. For our slightly sloping study site (slope<7°), the direct method defined as the difference between the signal begin and the ground peak provided forest height estimates with an accuracy of 2.2 m. The use of statistical models based on waveform metrics and digital elevation models provided slightly better results (1.89 m accuracy) in comparison with the direct method and the most relevant metrics proved to be the trailing edge extent and the waveform extent. Moreover, a power law model was used to fit in situ aboveground biomass to in situ forest height. The results using this model with GLAS-derived heights showed an accuracy for biomass of 16.1Mg/ha.
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 7, num. 1, p. 290 - 299