|
Tropical Forest Structure and Biomass Estimation from Bistatic InSAR at X-band (TanDEM-X), Small-Spot Lidar, and Field Measurements in the Amazonian Basin
Robert
Neil
Treuhaft, JPL, California Institute of Technology, robert.treuhaft@jpl.nasa.gov
(Presenter)
Fabio
G.
Goncalves, Wood Hole Research Center, Oregon State University / JPL, fgoncalves@whrc.org
Joao
Roberto
dos Santos, INPE, jroberto@ltid.inpe.br
Soren
Madsen, JPL, Caltech, soren.madsen@jpl.nasa.gov
Michael
Palace, CSRC-University of New Hampshire, michael.palace@unh.edu
Michael
Keller, USDA Forest Service, mkeller.co2@gmail.com
Scott
Hensley, JPL, scott.hensley@jpl.nasa.gov
Tropical forests are complex structurally and this complicates both field efforts and remote sensing studies to estimate biophysical properties, such as forest structure and biomass. In addition, the vast expanse and logistic difficulty in conducting field based measurements highlights the importance of remote sensing methods. Limitation of optical sensors in the tropics due to cloud and aerosol contamination are issues not faced with radar sensors. TanDEM-X, horizontally and vertically polarized radar interferometric data were taken over Tapajós National Forest in Brazil with vertical wavelengths of 155, 73, and 39 m—corresponding to alpha_z (or kappa_z) of 0.04, 0.09, and 0.16 rad/m, respectively. The area under study is 55 km x 20 km. Preliminary analyses suggest that X-band interferometric coherence decreases by about a factor of 2 (from 0.95 to 0.45) with increasing field-measured vertical extent (average heights of 7-25 m) and biomass (10-430 Mg/ha), further suggesting, as has been observed at C-band, that InSAR penetrates substantially more than SAR. Unlike InSAR coherence versus biomass, SAR power at X-band versus biomass shows no trend. Airborne lidar coherence at the above alpha_z/kappa_z is also shown to decrease as a function of biomass. The lidar decrease is about 15% more than the InSAR, suggesting that lidar penetrates more than InSAR. But this is a preliminary observation, which will be clarified with a calculation showing differences in effective penetration for lidar and InSAR. The importance of these preliminary results is that X-band InSAR may be useful for structure and biomass estimation, with wall-to-wall coverage. The efficacy of InSAR depends on whether a small number of Fourier components (meaning a small number of baselines or alpha_z's) are sufficient for structure/biomass remote sensing. This preliminary analysis suggests that X-band radar is potentially applicable to tropical forest remote sensing estimates of structure and biomass. The comparison with high resolution aiborne lidar sheds light on the rather unique application of X-band radar.
Presentation Type: Poster
Session: Poster Session 2-B
(Wed 4:30 PM)
Associated Project(s):
Poster Location ID: 50
|