Dr. Eric S. Kasischke
Radar Science Laboratory
ERIM
P.O. Box 8618
Ann Arbor, MI 48107
Co-Investigator:
Norman Christensen, Duke University
Estimation of Total Aboveground Biomass in Southern United States Old-Field Pine
OBJECTIVES
The overall goal of the NASA-sponsored research project is to develop methods to use
spaceborne
SAR
imagery to monitor patterns of carbon storage in the pine forests
found in the southeastern United States. This project is part of an overall program
being carried out by scientists at
ERIM
and Duke to develop techniques to use satellite-based
remote sensors to monitor carbon flux in this region. Towards this end, funding
has also been secured from the Environmental Protection Agency to pursue research
beyond the scope of the
SIR-C
project. Specifically,
EPA
funding is being used to obtain
and analyze Landsat
TM
and
MSS
imagery over the test site to monitor changes in forest
cover in this region from 1975 to present, to develop models describing patterns
of below-ground carbon storage, and to develop an integrated approach using information
derived from
SIR-C
and Landsat with ground-based models to study landscape scale
patterns of carbon storage and flux.
Specific objectives of the overall program are to:
Develop optimum algorithms to estimate aboveground biomass/carbon in pine forests
of the southeast U.S. using multichannel, spaceborne
SAR
data (NASA funded).
Develop models which link aboveground biomass/carbon with ground-layer carbon in pine
forests in the southeast U.S. (EPA funded).
Based on (1) and (2) above, estimate patterns of carbon storage in the pine forests
surrounding Durham, North Carolina at the times of the
SIR-C
overflights
(NASA funded).
Evaluate the utility of using
SIR-C/X-SAR
data to improve maps of forest cover in
the study region (NASA funded).
Develop methods to monitor changes in forest cover from 1975 to 1994 in the study
site using Landsat
TM
and
MSS
data (EPA funded).
Develop methods to combine the information derived from Landsat and
SIR-C
data to
study patterns of carbon flux over the past 20 years (EPA/NASA funded).
PROGRESS
Prior to March of 1994, our activities focused on investigations of
AIRSAR
and
ERS-1SAR
data to develop approaches in using radar imagery to monitor changes in biomass
in southern U.S. pines. Towards this goal, a set of 80 test sites were established
in the Duke Forest. These test sites were measured and techniques developed to estimate
patterns of biomass present in these forests based on in situ measurements as well
as allometric equations available in the literature. The utility of
SAR
data to
monitor biomass patterns was clearly established during these initial experiments. These
data were also used in theoretical modeling studies to better understand the relationships
between
EM
scattering and tree characteristics. This research resulted in eight published journal articles (see reference list).
The activities pursued during the past eighteen months were first focused on collection
of ground-truth data sets to support the analysis of
SIR-C/X-SAR
data. These ground
truth data included measurement of stand characteristics for 60 different pine stands, collection of soil moisture measurements during the
SIR-C
overflights in April
and October of 1994, collection of local meteorologic conditions during the
SIR-C
overflights, and collection of canopy hemispherical photographs. The support of
two
SIR-C
overflights within a seven-month period stretched the financial resources available
at Duke University to the fullest, and required expending funds not received until
November of 1994. Thus, we were not able to begin to fully reduce the ground data
until the spring of 1995, when additional funding was received. In addition to supporting
the experiments themselves, the
PI
for this experiment (E. Kasischke) also served
as Co-Chairman of the Ecology Discipline Panel for the National Research Council's
review of the utility of Spaceborne
SAR
for earth science applications. This activity
required considerable efforts from the
PI
during the time period from November 1994
through February 1995, and severely strained the funding resources available for
analysis of
SIR-C
data until a further funding increment was received in March of 1995.
Since March of 1995, we have begun to seriously analyze the
SIR-C/X-SAR
data sets.
Our activities during this time period have focused on reducing the ground-truth
data and on extracting signatures from the radar data sets. The data are currently
being used in two separate analyses, each of which should result in a journal article. The
first analyses is being conducted jointly with Frank Davis (UCSB) and Yong Wang (ECU).
This is a modeling study on the sensitivity of radar backscatter to variations in
surface moisture and changes in incidence angle. The second analyses is exploring
the effects variations in imaging geometry and seasonal variations (e.g., leaf on
versus leaf off, soil moisture) on the ability of
SAR
to estimate aboveground biomass.
The approach being used is through a multistage approach, where canopy biomass is estimated
from radar imagery, which in turn is used to estimate total stand. These journal
articles are currently in preparation and should be submitted by the end of the year.
In addition to submitting the two journal articles mentioned above, during FY96/97
we plan the following activities (each of which will result in a separate journal
article):
1. Analysis of Different Approaches to Estimation of Aboveground Biomass Using SIR-C/X-
SAR
Data
Several different methodologies have been proposed to estimate aboveground biomass
in forests. In addition to the two-step approach developed by this program, Jon
Ranson uses a polarization ratio to estimate biomass, while Craig Dobson uses a multiple-step approach where longer wavelength
SAR
data are used to estimate tree height and basal
area and shorter wavelength data are used to estimate canopy biomass. Under this
study, we will compare all three approaches to determine which approach is best suited
for the forest conditions found in the southeast U.S.
2. Regional Estimates on Patterns of Carbon Storage in a Southeast U.S. Pine Forest
Complex using Spaceborne
SAR
Data
This study will take the results from the algorithm from the previous analyses and
use them to estimate patterns of biomass present in the pine forests in the Duke
Forest Region. This analysis will also utilize the below-ground carbon models developed
by the
EPA
study to develop a total carbon budget for these forests.
3. Analysis of the Effects of Imaging and Scene Parameters for Detection of Floodingin
a Bald Cypress Forest Wetland
Duke Forest contains a test stand of Bald Cypress which was continuously flooded during
the
SIR-C
overflights. In the April flight, the leaves on the trees of this stand
had just begun to flush, while during the October flight, the leaves were fully flushed. This study will focus on determining how well the different wavelength/polarizations
of the
SIR-C/X-SAR
detected this flooded forest wetland as a function of incidence
angle and leaf on/off conditions. The study will employ both empirical analyses
as well as theoretical scattering models.
4. Mapping Forest Cover using Combined Landsat
TM
and
SIR-C/X-SAR
Data
This study will investigate whether imaging radar's unique capabilities in mapping
forest structural characteristics can be used to improve maps of forest/land cover
which are traditionally derived from
MSS
data.
5. Patterns of Carbon Storage and Flux in a Southern U.S. Forest from 1974 to 1994
This is the culmination of the overall program. This study will combine information
(within a
GIS
context) derived from three separate sources (e.g., ground-based models,
SIR-C/X-SAR,
and Landsat) to look at how patterns of forest cover change have influenced carbon flux in the forests surrounding Durham, North Carolina.
PUBLICATIONS
Dobson, M. C., F. T., Ulaby, T. Le Toan, A. Beaudoin, E. S. Kasischke, and N. C. Christensen,
Dependence of radar backscatter on conifer forest biomass, IEEE Trans. Geosci. Remote Sens.,
30, pp. 412-415, 1992.
Kasischke, E. S. and N. L. Christensen, Jr., Connecting forest ecosystem and microwave
backscatter models, Int. J. Remote Sens.
, 11, pp. 1277-1298, 1990.
Kasischke, E. S., L. L. Bourgeau-Chavez, N. L. Christensen, Jr., and E. Haney, Observations
on the sensitivity of
ERS-1SAR
image intensity to changes in aboveground biomass
in young loblolly pine forests, Intern. J. Remote Sens.
, 15, pp. 3-16, 1994.
Kasischke, E. S., Christensen, N. L., Jr., and E. Haney, Modeling of geometric properties
of loblolly pine tree and stand characteristics for use in radar backscatter models,
IEEE Trans. Geosci. Remote Sensing
, 32, pp. 800-822, 1994.
Kasischke, E. S., Christensen, N. L., Jr., and L. L. Bourgeau-Chavez, Correlating
radar backscatter with components of biomass in loblolly pine forests. IEEE Trans. Geosci. Remote Sensing
, 33, 643-659, 1995.
Ustin, S. L., C. A. Wessman, B. Curtiss, E. Kasischke, J. Way, and V. Vanderbilt,
Opportunities for Using the
EOS
Imaging Spectrometers and Synthetic Aperture Radar
in Ecological Models, Ecology
, 72, pp. 1934-1945, 1991.
Wang, Y., F. W. Davis, J. M. Melack, E. S. Kasischke and N. L. Christensen, Jr.,
The Effects of Changes in Forest Biomass on Radar Backscatter from Tree Canopies,
Int. J. Remote Sens.
, 16, pp. 503-513, 1995.
Wang, Y., E. S. Kasischke, F. W. Davis, J. M. Melack, and N. L. Christensen, Jr.,
The Effects of Changes in Loblolly Pine Biomass and Soil Moisture Variations on
ERS-1SAR
Backscatter - a Comparison of Observations with Theory, Remote Sens. Env.
, 49, pp. 25-31, 1994.