@Article{NeeffDutrSantFrei:2005:TrFoBi,
author = "Neeff, Till and Dutra, Luciano Vieira and Santos, Jo{\~a}o
Roberto dos and Freitas, Corina Costa",
affiliation = "University of California, Berkeley, CA, United States; University
of Freiburg, Biometry Department, Tennenbacher Strasse 4, 79085
Freiburg Brsg., Germany",
title = "Tropical forest biomass measurement by interferometric height
modeling and P-band radar backscatter",
journal = "Forest Science",
year = "2005",
volume = "51",
number = "6",
pages = "585--593",
keywords = "Amazon, P-band, backscatter saturation, biomass, environmental
management, forest, forest height, forest management, forest
resources, forestry, forestry research, forestry science,
interferometric height, natural resource management, natural
resources, radar.",
abstract = "A new approach to tropical forest biomass monitoring with airborne
interferometric X and P-band synthetic aperture radar (SAR) data
is presented. Forest height, basal area, and aboveground biomass
are modeled from remote sensing data for a study site in the
Brazilian Amazon. Radar data quality has improved: A novel digital
model of vegetation height from X- and P-band interferometry is
available along with the usual P-band backscatter information. The
digital vegetation height model is derived from the
interferometric surface models of the ground (from P-band) and the
forest canopy (from X-band). The difference between the surface
models is called interferometric height, and is used as a measure
of vegetation height. Interferometric height is shown to relate to
a subset of the forest trees that changes according to the forest
successional stages. The suitability of radar backscatter and
interferometric height as a means for forest and biomass
monitoring was explored by relating forest parameters as measured
in the field to remote sensing data. Basal area and biomass were
related to radar backscatter with limited precision of r 2 = 0.19
and r 2 = 0.34, respectively. Mean forest height is shown to
relate to interferometric height with good precision (r 2 = 0.83,
RMSE = 4.1 m). A statistical model for forest biomass as a
function of both P-band backscatter and interferometric height
information not only arrives at high values of precision (with r 2
= 0.89 and a RMSE from cross-validation of only 46.1 t/ha), but
also overcomes the well-known issue of backscatter saturation.
This research shows that tropical forest biomass can be quantified
and mapped over large areas for a range of forest structures with
reasonably tight and similar errors.",
issn = "0015-749X",
language = "en",
urlaccessdate = "03 maio 2024"
}