Close Window

LC-18 Abstract

Amazon Forest Selective Logging and Phenology from Imaging Spectroscopy

Gregory Paul Asner — Carnegie Institution (US-PI)
Jose Natalino Macedo Silva — Brazilian Forest Service (SA-PI)

Research Approach

Our proposal focuses on the

detection of canopy structural, biophysical and biochemical changes of selective

logging using hyperspectral optical remote sensing data. The optical data respond primarily to

chemical and structural changes in forest canopies from both natural and human

sources of variability. Our ongoing

field work characterizing chronosequences of selective logging intensity and

regrowth in the eastern Amazon will establish the chemical and structural

effects, which will be remotely sensed with the airborne AVIRIS imaging

spectrometer. These chemical and

structural effects resulting from selective logging will need to be separated

from background phenological effects.

The goals, or final products, of this proposal are: 1) identification of

selective-logging events, 2) quantification of canopy damage, 3)

characterization and separation of phenological signatures, 4) demonstration of

approaches for scaling up the airborne results to TERRA ASTER, EO-1 Hyperion,

and Landsat ETM+, and 5) develop substantial education and inter-agency

training on the use of remote sensing for forestry applications. The science goals require quantitative

estimates of three biochemical and structural remote-sensing products. From the hyperspectral optical data, we will

estimate 1) sub-pixel exposed cover of green canopy, non-photosynthetic

vegetation (NPV), soil, and deep shadow, 2) canopy leaf area index (LAI) and

relative water content, and 3) upper canopy pigment concentration.

We predict that each stage in the

selective logging chronosequence from primary forest to the logging event to

regrowth has distinct hyperspectral signatures. A chemical-physical photon transport model will be used to

analyze the hyperspectral data. Fazenda

Cauaxi, Floresta Nacional de Tapajos, and Fazenda Fortaleza (spread across the

state of Para) are the primary study sites at which our chronosequences have

been and will continue to be characterized.

The field effort includes rotational surveys of canopy damage, wood

volume and biomass removal, and regrowth as well as intensive campaigns (during

aircraft overflights) to measure canopy biophysics, biochemistry and structure,

and atmospheric properties. AVIRIS will

be flown over these sites in the beginning (July) and end (Nov) of the dry

season. We propose that hyperspectral

remote sensing will greatly enhance the reliability and accuracy of selective

logging detection, measurement of selective logging intensity (e.g. via

structural damage), and regrowth. This

information is critically needed to assess the impacts of selective logging,

which is pertinent to forest management and carbon cycle research.

Close Window