Science of Remote Sensing (Dec 2021)

Comprehensive comparison of airborne and spaceborne SAR and LiDAR estimates of forest structure in the tallest mangrove forest on earth

  • Atticus E.L. Stovall,
  • Temilola Fatoyinbo,
  • Nathan M. Thomas,
  • John Armston,
  • Médard Obiang Ebanega,
  • Marc Simard,
  • Carl Trettin,
  • Robert Vancelas Obiang Zogo,
  • Igor Akendengue Aken,
  • Michael Debina,
  • Alphna Mekui Me Kemoe,
  • Emmanuel Ondo Assoumou,
  • Jun Su Kim,
  • David Lagomasino,
  • Seung-Kuk Lee,
  • Jean Calvin Ndong Obame,
  • Geldin Derrick Voubou,
  • Chamberlain Zame Essono

Journal volume & issue
Vol. 4
p. 100034

Abstract

Read online

A recent suite of new global-scale satellite sensors and regional-scale airborne campaigns are providing a wealth of remote sensing data capable of dramatically advancing our current understanding of the spatial distribution of forest structure and carbon stocks. However, a baseline for forest stature and biomass estimates has yet to be established for the wide array of available remote sensing products. At present, it remains unclear how the estimates from these sensors compare to one another in terrestrial forests, with a clear dearth of studies in high carbon density mangrove ecosystems. In the tallest mangrove forest on Earth (Pongara National Park, Gabon), we leverage the data collected during the AfriSAR campaign to evaluate 17 state-of-the-art sensor data products across the full range of height and biomass known to exist globally in mangrove forest ecosystems, providing a much-needed baseline for sensor performance. Our major findings are: (Houghton, Hall, Goetz) height estimates are not consistent across products, with opposing trends in relative and absolute errors, highlighting the need for an adaptive approach to constraining height estimates (Panet al., 2011); radar height estimates had the lowest calibration error and bias, with further improvements using LiDAR fusion (Bonan, 2008); biomass variability and uncertainty strongly depends on forest stature, with variation across products increasing with canopy height, while relative biomass variation was highest in low-stature stands (Le Quéréet al., 2017); a remote sensing product's sensitivity to variations in canopy structure is more important than the absolute accuracy of height estimates (Mitchardet al., 2014); locally-calibrated area-wide totals are more representative than generalized global biomass models for high-precision biomass estimates. The findings presented here provide critical baseline expectations for height and biomass predictions across the full range of mangrove forest stature, which can be directly applied to current (TanDEM-X, GEDI, ICESat-2) and future (NISAR, BIOMASS) global-scale forest monitoring missions.

Keywords