Peer-Reviewed Articles
921 Land Cover in the Amazon Estuary: Linking of the Thematic Mapper
with Botanical and Historical Data
Eduardo Brondizio, Emilio Moran, Paul Mausel, and You Wu
Abstract
Download
Full Article
A Landsat TM scene from July, 1991 was analyzed for an area of the Amazon estuary
in Ponta de Pedras, Marajo Island, Brazil. Distinctive spectral signatures
were determined for 14 land- cover classes, including upland and floodplain
forest, three stages of secondary succession, palm forest, mangrove, pasture,
and three types of savanna. Image classification of the study area was conducted
using the 14 class spectral statistics informed by 1992 vegetation inventories
and field studies documenting historical land use. The use of field- based
information supportive of classification resulted in individual test field
class results which ranged from 81 to 100 percent individual class accuracy.
Elements of the classification were focused on addressing the difficult problem
of identifying the conversion of 'natural' to 'managed' floodplain forest.
The combination of feature classification using computer-analyzed TM data in
conjunction with detailed ground measurements/surveys permitted identification
of subtle changes in natural forest that was associated with conversion to
managed floodplain forest.
931 Change Detection at Multiple Temporal Scales: Seasonal and Annual
Variations in Landscape Variables
Eric F. Lambin
Abstract
Download
Full Article
Demonstrates the limitations of the classic change detection approach in situations
where landscape attributes are characterized by marked seasonal variations;
explores the effect of increasing the frequency of observation on the change
detection performances, using a common data aggregation approach; and tests
a method which monitors broad time-scale changes in landscape attributes
by taking explicitly into account the finer time-scale variations. These
issues are explored with data on the landscape spatial pattern of three West
African landscapes. The spatial structure of these landscapes is measured
from AVHRR data, for a period covering two years. It is shown that seasonal
changes in landscape spatial pattern may be much greater than that caused
by long- term changes. Combining information from multiple temporal scales
of observation is essential to quantify in a meaningful way the spatial dynamics
of landscapes.
939 Remote Sensing of Mangrove Wetlands: Relating Canopy Spectra to Site-Specific
Data
Elijah W. Ramsey III and John R. Jensen
Abstract
Download
Full Article
Remote sensing was examined as a tool to describe the spectral and structural
changes within and between mangrove species and community types. To accomplish
this goal, high-resolution canopy reflectance spectra were obtained at 21
mangrove sites in southwest Florida. Results from sets of reflectances in
combination with correlation analyses suggest blue and red reflectances were
redundant as were SPOT panchromatic and green reflectances and all normalized
difference vegetation indexes derived for each set of NIR and red reflectances.
Eighty-four percent of the LAI variance was explained by using any generated
normalized difference vegetation index; however, species composition was
not correlated to any combination of reflectance bands or vegetation index.
949 Inferring Urban Land Use from Satellite Sensor Images Using Kernel-Based
Spatial Reclassification
M.J. Barnsley and S.L. Barr
Abstract
Download
Full Article
Per-pixel classification algorithms are poorly equipped to monitor urban land
use in images acquired by the current generation of high spatial resolution
satellite sensors. This is because urban areas commonly comprise a complex
spatial assemblage of spectrally distinct land-cover types. A technique is
described that attempts to derive information on urban land use in two stages.
The application of this technique, known as SPARK (SPAtial Reclassification
Kernel), is demonstrated using a SPOT-1 HRV multispectral image of southeast
London, England. Preliminary results indicate that SPARK can be used to distinguish
quite subtle differences of land use in urban areas.
959 On Using the NOAA AVHRR "Experimental Calibrated Weekly Global Vegetation
Index"
Erik A. Williams and Dennis E. Jelinski
Abstract
Download
Full Article
Several types of real and potential error in the current bi- weekly Experimental
Global Vegetation Index (EGVI) distributed by the National Geophysical Data
Center (NGDC) are described. The real errors are related to duplication of
files among years. The potential errors arise from problems in resampling that
are associated with transferring data between two types of projections. The
failure to recognize and address these problems will lead to serious analytical
error and false inferences.
961 Comparison of Nadir and Off-Nadir Multispectral Response Patterns
for Six Tallgrass Prairie Treatments in Eastern Kansas
John W. Dunham and Kevin P. Price
Abstract
Download
Full Article
Ground-level spectroradiometer measurements taken at nadir and 45degrees off-nadir
from four sensor view azimuths were used to examine the spectral reflectance
patterns of six tallgrass prairie treatments in eastern Kansas. The six treatments
included native prairie, burned, mowed, hayed, grazed, and unmanaged grasslands.
Spectral reflectance patterns and biophysical measurements were made for
each treatment. The native and unmanaged treatments were spectrally unique
from the other four grassland treatments. View zenith and azimuth had no
effect on the spectral separability among treatments. Across all treatments,
only the off-nadir measurements taken at a 0degrees view azimuth (sun behind
sensor) were significantly different from those taken at other view zenith-
azimuth angles.
969 Visual Interpretation of Vegetation Classes from Airborne Videography:
An Evaluation of Observer Proficiency with Minimal Training
Sam Drake
Abstract
Download
Full Article
Evaluates the ability of individual subjects and small groups to correctly
identify Arizona plant communities from color airborne video footage, explores
the relationship of five background variables to subjects' success, and determines
which community types are easiest and most difficult for subjects to identify.
Three hours of training increased the mean individual score from seven correct
(pretest) to 21 correct (posttest), and mean group score from 11 to 24. Posttest
results showed no significant difference in ability among individuals or
between individuals and groups. The most difficult community to identify
was creosote- tarbush desertscrub; the easiest was paloverde-saguaro desertscrub.
Findings support the feasibility of video interpretation by minimally trained
personnel.
| Top | Home |