@article{fdi:010085149,
  title = {{I}mproving heterogeneous forest height maps by integrating {GEDI}-based forest height information in a multi-sensor mapping process},
  author = {{M}orin, {D}. and {P}lanells, {M}. and {B}aghdadi, {N}. and {B}ouvet, {A}lexandre and {F}ayad, {I}. and {L}e {T}oan, {T}. and {M}ermoz, {S}. and {V}illard, {L}.},
  editor = {},
  language = {{ENG}},
  abstract = {{F}orests are one of the key elements in ecological transition policies in {E}urope. {S}ustainable forest management is needed in order to optimise wood harvesting, while preserving carbon storage, biodiversity and other ecological functions. {F}orest managers and public bodies need improved and cost-effective forest monitoring tools. {R}esearch studies have been carried out to assess the use of optical and radar images for producing forest height or biomass maps. {T}he main limitations are the quantity, quality and representativeness of the reference data for model training. {T}he {G}lobal {E}cosystem {D}ynamics {I}nvestigation ({GEDI}) mission (full waveform {L}i{DAR} on board the {I}nternational {S}pace {S}tation) has provided an unprecedented number of forest canopy height samples from 2019. {T}hese samples could be used to improve reference datasets. {T}his paper aims to present and validate a method for estimating forest dominant height from open access optical and radar satellite images ({S}entinel-1, {S}entinel-2 and {ALOS}-2 {PALSAR}-2), and then to assess the use of {GEDI} samples to replace field height measurements in model calibration. {O}ur approach combines satellite image features and dominant height measurements, or {GEDI} metrics, in a {S}upport {V}ector {M}achine regression algorithm, with a feature selection process. {T}he method is tested on mixed uneven-aged broadleaved and coniferous forests in {F}rance. {U}sing dominant height measurements for model training, the cross-validation shows 7.3 to 11.6% relative {R}oot {M}ean {S}quare {E}rror ({RMSE}) depending on the forest class. {W}hen using {GEDI} height metrics instead of field measurements for model training, errors increase to 12.8-16.7% relative {RMSE}. {T}his level of error remains satisfactory; the use of {GEDI} could allow the production of dominant height maps on large areas with better sample representativeness. {F}uture work will focus on confirming these results on new study sites, improving the filtering and processing of {GEDI} data, and producing height maps at regional or national scale. {T}he resulting maps will help forest managers and public bodies to optimise forest resource inventories, as well as allow scientists to integrate these cartographic data into climate models.},
  keywords = {forest inventory ; {S}entinel ; {ALOS} {PALSAR} ; {GEDI} ; canopy height ; support ; vector ; temperate forest},
  booktitle = {},
  journal = {{R}emote {S}ensing},
  volume = {14},
  numero = {9},
  pages = {2079 [25 p.]},
  year = {2022},
  DOI = {10.3390/rs14092079},
  URL = {https://www.documentation.ird.fr/hor/fdi:010085149},
}