NOTICE -- The LBA-ECO Project website is no longer being supported.  This archive is a snapshot, as it existed in 2013, of the LBA-ECO website, maintained by NASA Goddard Space Flight Center, and now archived at the ORNL DAAC.  Links to external websites may be inactive. Final data products from the LBA project can be found at the ORNL DAAC.
banner
banner banner banner banner banner banner
banner banner banner banner banner banner banner
home aboutlibrarynews archivecontacts banner

spacer
banner
Investigations
Overview
Abstracts & Profiles
Publications
Research Sites
Meetings
Synthesis Groups
LBA-HYDROMET
LBA-Air-ECO
Logistics
Overview
Field Support
Travel
Visa
Shipping
Data
  Overview
Find LBA Data
Investigator Checklist
Process & Policy
Documentation & Archive
Training & Education
  Overview
Activities Summary
T&E Goals
Student Opportunities
  Folha Amazônica
 
spacer

Investigation:

LC-07 (Melack / Novo / Forsberg)

LBA Dataset ID:

LC07_SAR_WETLAND_MASK

Originator(s):

1. HESS, L.L.
2. MELACK, J.M.
3. NOVO, E.M.L.M.
      4. BARBOSA, C.C.F.
5. GASTIL, M.

Point(s) of Contact:

ORNL DAAC User Services Office Oak Ridge National Laboratory Oak Ridge, Tennessee 37 (ornldaac@ornl.gov)

Dataset Abstract:

The high and low water acquisitions of JERS SAR coverage of almost the whole Amazon Basin offers us a rich dataset; we have applied INPE\'s segmentation algorithms to conduct the polygon-based classification of wetlands using the JERS data. Further, we have developed a method for classifying and mapping floodplain habitats by linking multi-date SAR imagery with river stage records, incorporating information on both vegetation structure and inundation periodicity, and applied the method to sites in the central Amazon. The wetlands area, defined by the wetlands mask, was further classified into wetland habitats. A polygon-based segmentation and clustering was used to delineate wetland extent with an accuracy of 95%. A pixel-based classifier was used to map wetland vegetation and flooding state based on backscattering coefficients of two-season class combinations.

Beginning Date:

1995-09-01

Ending Date:

1996-06-30

Metadata Last Updated on:

2012-04-12

Data Status:

Archived

Access Constraints:

PUBLIC

Data Center URL:

http://daac.ornl.gov/

Distribution Contact(s):

ORNL DAAC User Services Office Oak Ridge National Laboratory Oak Ridge, Tennessee 37 (ornldaac@ornl.gov)

Access Instructions:

PUBLIC

Data Access:

IMPORTANT: The LBA-ECO Project website is no longer being supported. Links to external websites may be inactive. Final data products from the LBA project can be found at the ORNL DAAC. Please follow the fair use guidelines found in the dataset documentation when using or citing LBA data.
Datafile(s):

LBA-ECO LC-07 JERS-1 SAR Wetlands Masks and Land Cover, Amazon Basin: 1995-1996 :  http://daac.ornl.gov/cgi-bin/dsviewer.pl?ds_id=1079

Documentation/Other Supporting Documents:

LBA-ECO LC-07 JERS-1 SAR Wetlands Masks and Land Cover, Amazon Basin: 1995-1996 :  http://daac.ornl.gov/LBA/guides/LC07_SAR_Wetlands_Mask.html

Citation Information - Other Details:

Hess, L.L., J.M. Melack, E.M.L.M. Novo, C.C.F. Barbosa, and M. Gastil. 2012. LBA-ECO LC-07 JERS-1 SAR Wetlands Masks and Land Cover, Amazon Basin: 1995-1996. Data set. Available on-line [http://daac.ornl.gov] from Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge, Tennessee, U.S.A. http://dx.doi.org/10.3334/ORNLDAAC/1079

Keywords - Theme:

Parameter Topic Term Source Sensor
INUNDATION LAND SURFACE LAND USE/LAND COVER JERS-1 (JAPANESE EARTH RESOURCES SATELLITE-1) SAR (SYNTHETIC APERTURE RADAR)
LAND COVER LAND SURFACE LAND USE/LAND COVER JERS-1 (JAPANESE EARTH RESOURCES SATELLITE-1) SAR (SYNTHETIC APERTURE RADAR)
LAND USE CLASSES LAND SURFACE LAND USE/LAND COVER JERS-1 (JAPANESE EARTH RESOURCES SATELLITE-1) SAR (SYNTHETIC APERTURE RADAR)
WETLANDS LAND SURFACE LAND USE/LAND COVER JERS-1 (JAPANESE EARTH RESOURCES SATELLITE-1) SAR (SYNTHETIC APERTURE RADAR)

Uncontrolled Theme Keyword(s):  AQUATIC MACROPHYTE, AQUATIC MACROPHYTES, CLASSIFICATION, FLOODED FOREST, FLOODING, FLOODPLAIN, INUNDATION, LAND COVER CLASSIFICATION, RIVER, RIVERS, VARZEA, VEGETATION, WETLAND, WETLANDS

Keywords - Place (with associated coordinates):

Region
(click to view profile)
Site
(click to view profile)
North South East West
  AMAZON BASIN 0.00000 -16.00000 -48.00000 -72.08740

Related Publication(s):

Hess, L.L., J.M. Melack, E.M.L.M. Novo, C.C.F. Barbosa, and M. Gastil. 2003. Dual-season mapping of wetland inundation and vegetation for the central Amazon basin. Remote Sensing of Environment, Vol. 87, No. 4, pp. 404-428.

Data Characteristics (Entity and Attribute Overview):

Data Characteristics:


Part 1 Brief description:</b>

<p>The high and low water acquisitions of JERS SAR coverage of almost the whole Amazon Basin offers us a rich dataset; we have applied INPE's segmentation algorithms to conduct of polygon-based classification of wetlands using the JERS data. Further, we have developed a method for classifying and mapping floodplain habitats by linking multi-date SAR imagery with river stage records, incorporating information on both vegetation structure and inundation periodicity, and applied the method to sites in the central Amazon.</p>

A separate archive, part of LC-32, will contain the habitat class map for a much larger region of which this Central map is a subset.


Part 2 Image file structure</b>

<p>The class image is a binary array of 21609 columns by 9604 rows. Pixel values are DN from 1 to 9 for the wetland habitat classes and 0 for upland.

A text file listing the DN and their class names is provided as a legend and is in the format of a Density Slice used in ENVI. The habitat classes are further

defined in two tables included here as .rtf files, listed below.</p>

<p>For the image files of type 3arcsec each pixel represents an area of 3 arc seconds. Dimensions of the image are referenced to geographic coordinates, latitude and longitude, not a map projection. The north west corner is at 72 W, 0 N. The south east corner is at 54 W, 8 S. Map information is included in the accompanying .tfw header file.</p><p>


Part 3 List of image files</b>

<p>This subdirectory contains data products for the Amazon central region defined by a box

of 1.77 km2 area from 72 to 54 degrees west longitude from 0 to 8 degrees south latitude.</p>

One data product in two seasons in two geogrphic projectins is available here:



amazon_central_wetland_habitats-hi_100m.tfw 20-Dec-2003 17:43 1k

amazon_central_wetland_habitats-hi_100m.tif 20-Dec-2003 17:43 170M

amazon_central_wetland_habitats-hi_3arcsec.tfw 20-Dec-2003 17:32 1k

amazon_central_wetland_habitats-hi_3arcsec.tif 20-Dec-2003 17:32 198M

amazon_central_wetland_habitats-lo_100m.tfw 20-Dec-2003 17:55 1k

amazon_central_wetland_habitats-lo_100m.tif 20-Dec-2003 17:55 170M

amazon_central_wetland_habitats-lo_3arcsec.tfw 20-Dec-2003 17:35 1k

amazon_central_wetland_habitats-lo_3arcsec.tif 20-Dec-2003 17:35 198M



<p>The classified image of the amazon_central region is offered as tif images in two

formats. Geocoded tif was chosen because it preserves the pixel values. The .tfw files

are the accompanying headers and should be downloaded along with their corresponding .tif

file. The 3arcsec images are in the equal-angle format where each pixel is 3 arc seconds. The equal-area projection is approximately the same scale, with 100 m pixels. There is a binary mask of non-wetland versus wetland. There are two classified images, one for low-water and one for high-water, named lo and hi.</p>

<p>The files ending in 100m are in an equal-area projection. These pixels are 100 m x 100 m, 10000 square meters, or 1 hectacre,

in area. These images are in an Albers projection.</p>

<p>These GeoTiff images are appropriate for special image processing software such as IDL or ENVI, PCI, ArcGIS and such but are not appropriate for consumer photo viewing software.

These tif files will not open in a browser window. (That is not an error. These images are not appropriate to open in a browser window. For that, use the (below) pdf of the much-reduced resolution figure.)</p>


Part 4 Habitat class definitions</b>


The habitat classes are described in this reference:

Hess LL, Melack JM, Novo EMLM, Barbosa CCF, Gastil M. 2003. Dual-season mapping of wetland inundation and vegetation for the central Amazon basin. Remote Sensing of Environment.

Two tables excerpted from Hess (2003) give definitions of upland and wetland classes.

Those tables are in these files:

Hess_RSE_2003_Table2_Habitat_Classes.rtf

Hess_RSE_2003_Table3_Terms_Wetland_Vegetation.rtf





Color Legend

DN Color Habitat Class

1 blue Open water

2 yellow Bare or herbaceous, non-flooded

3 magenta Herbaceous, flooded

4 red Shrub, non-flooded

5 cyan Shrub, flooded

6 tan Woodland, flooded

7 green Forest, non-flooded

8 white Forest, flooded

9 na mixed

0 black Upland (masked as not floodable)



(The above table is in the text file RSE_2003_1digit_colors.dsr)

Data Application and Derivation:

Application:

This data set has been applied by our group to create a wetlands class map and by other groups (see related publication information section). The primary application of this dataset is to mask upland area from images to enable analysis on only the floodable areas. Cited in Related Publication Information are four examples of further work which used this mask. Melack, 2004, used this mask in conjunction with its derived class map to quantify methane emissions. Novo, 2006, applied this mask to MODIS images to detect seasonal changes in chlorophyll. Frappart, 2005, applied this mask over the Negro River Basin to estimate water storage. Richey, 2002, was first to apply this mask to estimate outgassing of CO2.



Derivation:

This data set is derived from the Central Amazon Wetlands Mask, an LC-07 product, which was derived from the JERS-1 GRFM mosaic of the Amazon Basin (see Rosenqvist 2000).

Quality Assessment (Data Quality Attribute Accuracy Report):

Quality Assessment:

(Excerpt from Hess, 2003) Aerial videographic surveys were carried out in

order to acquire data sets suitable for training of classification

algorithms and assessment of classification accuracy. Each validation sample comprised a 100x100-m area on a VOAM97 or VOAM99 video mosaic (Fig. 3), and the

accompanying video clips. Center points of these 1-ha

samples were selected by random sampling of flight times

along flight tracks within the study quadrat. Nine of the 365 video samples examined could not be

unambiguously labeled. Of the remaining 356 samples, 339

(95.2%) were mapped correctly on the wetlands mask. With

less than 21 misclassified samples, the map therefore is

accepted according to the specified standard of a less than 5% chance that the map is less than 90% accurate. The 17

misclassified samples were evenly divided between errors

of commission (wrongly mapped as wetlands) and errors of

omission (wrongly mapped as nonwetland). Misclassified

samples were examined to determine the source of the

error, and categorized as classification errors, registration

errors, size-based errors, and errors related to apparent

disparities in flooding between video flight and mosaic.

Most errors of commission were nonwetland clearings and

second growth adjacent to rivers or lakes. Registration

errors involved samples within 1.5 pixels of a wetland/

nonwetland boundary, and size-based errors resulted when

a land cover unit smaller than the 50-ha minimum mapping

was identifiable on the videography.

Process Description:

Data Acquisition Materials and Methods:

The wetland mask was created from the JERS-1 mosaic of the Amazon Basin produced by the Global Rain Forest Mapping Project. (See reference.) Additional acquisition materials included TM images, topographic maps at 1:100,000 from the Instituto Brasileiro de Geografia e Estatistica (IBGE), and digital videography. The digital videography and how it was used in creation of the wetlands mask is described in Hess, 2002.



Wetland extent was mapped for the central Amazon region, using mosaicked L-band synthetic aperture radar (SAR) imagery acquired by the Japanese Earth Resources Satellite-1. For the wetland portion of the 18 x 8 degree study area, dual-season radar mosaics were used to map inundation extent and vegetation under both low-water and high-water conditions at 100 m resolution, producing the first high-resolution wetlands map for the region. Thematic accuracy of the mapping was assessed using high-resolution digital videography acquired during two aerial surveys of the Brazilian Amazon.



A polygon-based segmentation and clustering was used to delineate wetland extent with an accuracy of 95%. A pixel-based classifier was used to map wetland vegetation and flooding state based on backscattering coefficients of two-season class combinations. Producer\'s accuracy for flooded and nonflooded forest classes ranged from 78% to 91%, with lower accuracy (63 - 65%) for flooded herbaceous vegetation. Seventeen percent of the study quadrat was occupied by wetlands, which were 96% inundated at high water and 26% inundated at low water. Flooded forest constituted nearly 70% of the entire wetland area at high water, but there are large regional variation in the proportions of wetland habitats. The SAR-based mapping provides a basis for improved estimates of the contribution of wetlands to biogeochemical and hydrological processes in the Amazon basin, a key question in the Large-Scale Biosphere-

Atmosphere Experiment in Amazonia.

References:

Melack, J.M., L.L. Hess, M. Gastil, B.R. Forsberg, S.K. Hamilton, I.B.T. Lima, and E.M.L.M. Novo. 2004. Regionalization of methane emissions in the Amazon Basin with microwave remote sensing. Global Change Biology 10(5):530-544. [LBA-ECO Pub ID = 399 ]



Richey, J.E., J.M. Melack, A.K. Aufdenkampe, V.M. Ballester, and L.L. Hess. 2002. Outgassing from Amazonian rivers and wetlands as a large tropical source of atmospheric CO2. Nature 416(6881):617-620. [LBA-ECO Pub ID = 233]





Novo, E.M.L.M., C.C.D. Barbosa, R.M. de Freitas, Y.E. Shimabukuro, J.M. Melack, and W. Pereira. 2006. Seasonal changes in chlorophyll distributions in Amazon floodplain lakes derived from MODIS images. Limnology 7(3):153-161. [LBA-ECO Pub ID = 758 ]



Hess, L.L., E.M.L.M. Novo, D.M. Slaymaker, J. Holt, C. Steffen, D.M. Valeriano, L.A.K. Mertes, T. Krug, J.M. Melack, M. Gastil, C. Holmes, and C. Hayward. 2002. Geocoded digital videography for validation of land cover mapping in the Amazon basin. International Journal of Remote Sensing 23(7):1527-1555. [LBA-ECO Pub ID = 217 ]



Rosenqvist, A., Shimada, M., Chapman, B., & Freeman, A. 2000. The Global Rain Forest Mapping project -- a review. International Journal of Remote Sensing, 21, 1375�1387.



Frederic Frappart, Frederique Seyler, Jean-Michel Martinez, Juan G. Leon and Anny Cazenave. 2005. Floodplain water storage in the Negro River basin estimated from microwave remote sensing of inundation area and water levels, Remote Sensing of Environment, 99, 387-399.

Skip navigation linksHOME | ABOUT | LIBRARY | NEWS ARCHIVE | CONTACTS | INVESTIGATIONS | LOGISTICS | DATA |TRAINING & EDUCATION

NASA logo
ORNL DAAC
Get Acrobat Reader