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:

CD-06 (Richey / Victoria)

LBA Dataset ID:

CD06_LULC_MAP_JIPARANA

Originator(s):

1. BALLESTER, M.V.R.
2. VICTORIA, D. DE C.
3. COBURN, R.
4. KRUSCHE, A.V.
5. VICTORIA, R.L.
      6. RICHEY, J.E.
7. LOGSDON, M.G.
8. MAYORGA, E.
9. MATRICARDI, E.A.T.

Point(s) of Contact:

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

Dataset Abstract:

A land use/land cover map of the Ji-Parana river basin in 1999 derived from the digital classification of 8 Landsat-7 ETM+ images. Original data were acquired from the Tropical Rain Forest Initiative (TRFIC).

Beginning Date:

1999-07-30

Ending Date:

2001-10-16

Metadata Last Updated on:

2012-05-24

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 CD-06 Ji-Parana River Basin Land Use and Land Cover Map, Brazil: 1999 :  http://daac.ornl.gov/cgi-bin/dsviewer.pl?ds_id=1087

Documentation/Other Supporting Documents:

LBA-ECO CD-06 Ji-Parana River Basin Land Use and Land Cover Map, Brazil: 1999 :  http://daac.ornl.gov/LBA/guides/CD06_LULC_Map_JiParana.html

Citation Information - Other Details:

Ballester, M.V.R., D. de C. Victoria, R. Coburn, A.V. Krusche, R.L. Victoria, J.E. Richey, M.G. Logsdon, E. Mayorga, and E. Matricardi. 2012. LBA-ECO CD-06 Ji-Parana River Basin Land Use and Land Cover Map, Brazil: 1999. 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/1087

Keywords - Theme:

Parameter Topic Term Source Sensor
LAND COVER BIOSPHERE LAND USE/LAND COVER LANDSAT-7 (LAND REMOTE-SENSING SATELLITE-7) ETM+ (ENHANCED THEMATIC MAPPER PLUS)

Uncontrolled Theme Keyword(s):  JI-PARANA RIVER, LAND USE AND LAND COVER, RONDONIA

Keywords - Place (with associated coordinates):

Region
(click to view profile)
Site
(click to view profile)
North South East West
  RONDONIA -8.03330 -12.92690 -60.01500 -63.41530

Related Publication(s):

Ballester, M.V.R., D. de C. Victoria, A.V. Krusche, R. Coburn, R.L. Victoria, J.E. Richey, M.G. Logsdon, E. Mayorga, and E. Matricardi. 2003. A remote sensing/GIS-based physical template to understand the biogeochemistry of the Ji-Parana river basin (Western Amazonia). Remote Sensing of Environment 87(4):429-445.

Data Characteristics (Entity and Attribute Overview):

Data Characteristics:

A land use/cover map for 1999 for the Ji-Parana River Basin, located in the State of Rondonia, Western Amazonia, Brazil is provided here in Erdas-Imagine format, including 9 classes:

1. Pasture

2. Forest

3. Water

4. Urban

5. Savanna

6. Annual crops

7. Regrowth / riparian

8. Bare / burned soils

9. Clouds / cloud shadows




Erdas-Imagine files:

lulc_99.img

lulc_99.rrd




Sector descriptions:

Sector descriptions at the Ji-Parana river basin



Sector River_name Order Level of land-use impact*

Code

COM-1 Comemoracao 3rd low

COM-2 Comemoracao 5th medium

PB-1 Pimenta Bueno 3rd low

PB-2 Pimenta Bueno 6th medium

JIP-1 Ji-Parana 6th high

ROLIM Rolim de Moura 5th very high

URUPA Urupa 5th very high

JIP-2 Ji-Parana 6th very high

JIP-3 Ji-Parana 6th high

JARU Jaru 6th high

MACH Machadinho 5th low

JIP-4 Ji-Parana 7th low

JIP-5 Ji-Parana 7th low

PRETO Preto 6th low



*The level of land use impact was determined using the percentage of pasture coverage in each sector, classified as:

low: 0-15%

medium: 15-30%

high: 30-50%

very high: 50-75%




Land use / land cover area percentages are provided as a single comma-delimited ASCII file:


Filename: JiParana_Land_Use_Area_1999.csv



Column Column Units/format Description

Number Heading

1 Sector_code N/A Sector codes (see above or see data set companion file: Sector_descriptions.csv).

2 Pasture % Percentage of sector where land use/land cover is pasture

3 Forest % Percentage of sector where land use/land cover is forest

4 Water % Percentage of sector where land use/land cover is water

5 Urban % Percentage of sector where land use/land cover is urban

6 Savanna % Percentage of sector where land use/land cover is savanna

7 Annual_crops % Percentage of sector where land use/land cover is annual crops

8 Regrowth % Percentage of sector where land use/land cover is regrowth / riparian

9 Bare_soils % Percentage of sector where land use/land cover is bare soils / burned soils

10 Clouds % Percentage of sector obscured by clouds / cloud shadows








Sector_code Pasture Forest Water Urban Savanna Annual_crops Regrowth Bare_soils Clouds

COM-1 12.6 47.37 0.03 1.86 0.02 32.41 0.32 5.38 0

PB-1 8.4 90 0 0.04 0 0.13 1.15 0.27 0

COM-2 28.8 66.72 0.26 0.64 0.39 0.41 2.47 0.3 0

PB-1 32.63 58.93 0.21 0.3 0.45 0.02 2.85 0.1 4.51

JIP-1 65.84 26.17 0.05 0.67 0.47 0 6.5 0.29 0

ROLIM 68.95 27.45 0.23 1.45 0.97 0.02 0.57 0.37 0

URUPA 42.6 49.88 0.09 0.29 0.33 0.01 6.27 0.52 0

JIP-2 55.7 36.8 0.32 0.82 0.94 0.05 4.43 0.94 0

JIP-3 52.84 36.19 0.1 0.53 0.22 0.01 9.43 0.68 0

JARU 39.93 53.36 0.5 1.09 0.29 0.1 3.8 0.94 0

MACH 21.69 68.28 0.03 0.21 0.08 0.03 9.57 0.11 0

JIP-4 7.9 86.41 0.63 0.12 2.1 0.02 2.52 0.31 0

JIP-5 0.56 89.37 1.96 0.09 1.8 0 2.83 0.03 3.35

PRETO 3.49 78.26 1.25 0.14 0.01 0.01 6.63 0.04 10.18

Entire_Basin 29.6 61.5 0.38 0.18 1.39 0.13 3.64 0.57 2.57

Data Application and Derivation:

Land use characterization and modeling

Quality Assessment (Data Quality Attribute Accuracy Report):

Quality Assessment:

Fifteen 1999 1 meter resolution videography mosaics of selected areas were produced and provided by the Brazilian National Institute for Space Research (INPE). This data set was originally acquired as part of the four Validation Over flights for Amazon Mosaics (VOAM) aerial video surveys fulfilled by Hess et al. (2002). The 1999 flight used a digital camcorder system, circumscribing the Brazilian Amazon, documenting ground conditions for wetlands, forests, savannas, and human-impacted areas. Global Positioning System information encoded on the video audio track was extracted by mosaicking software that automatically generate geocoded digital mosaics from video clips. A laser altimeter recorded profiles of terrain and vegetation canopy heights. A more detailed description of these data can be found in Hess et al., 2002 [verify this reference and provide in References section]. With a resolution of 1 m, the videography was sufficient to be considered top-quality ground reference data. The coordinates of 17 points, and the associated land use/land cover, were selected from these 15 videography mosaics to constitute reference data for the classification accuracy assessment. The points were located in the most remote areas of the basin, where ground access is very difficult (Figure 1) [Insert graph here].



Accuracy assessment also included 147 control points collected in the field and later brought into the ERDAS-IMAGINE accuracy assessment module. The classification points (class values) and ground truth points (reference points) were compared and quantitatively summarized to compute a matrix and an accuracy assessment report.



Overall classification accuracy was 89.12%, with an overall Kappa Statistic of 0.8667. These values were above the minimum limits for acceptance described in the literature (85%, Guptill and Morrison, 1995). [verify this reference and provide in References section]

Process Description:

Data Acquisition Materials and Methods:

Using ERDAS-IMAGINE, land use/land cover maps were produced from a digital classification of eight Landsat scenes. The images were acquired from the Tropical Rain Forest Information Center (TRFIC) at Michigan State University, as 1G products, radiometrically and geometrically corrected 0R images. Geodetic errors of these products are approximately 30 meters and this precision was achieved employing ground control points (http://www.bsrsi.msu.edu/trfic/data_portal/Landsat7doc). Visual inspection was used to determine geo-registration consistency between neighborhood scenes. Pre-classification processing included haze reduction in input images using the method based on the Tasseled Cap transformation, which yields a component, correlated with haze, that is then removed, and the image is transformed back into RGB space. We also used a histogram match before creating a final mosaic. Color composites were created for each scene from bands 1 to 5 and 7 and visually inspected. Scene 231-68 was used as the reference image because: a) it was representative of the entire study area; b) it contained no visible clouds or smoke, and c) it had a good distribution of the different land covers known to exist in the study area. Each individual band, from all other images, was histogram-matched to the corresponding band of scene 231-68. The supervised classification consisted of 183 signatures from training sites found throughout the mosaic; training sites were chosen in a spatially even and thorough manner across the entire image and processed as encoded radiance. The final classification included 9 classes: 1. forest; 2. water; 3. pasture; 4. riparian / regrowth; 5.savanna; 6. annual crops; 7. urban; 8. bare/burned soils and 9. clouds/cloud shadows. These classes were selected because, as in previous land use/land cover classifications using Landsat TM data (Kimes et al., 1998; Roberts et al., in press) [verify this reference and provide in References section], we were not able to accurately distinguish other land use/land cover classes, such as crop types, perennial crops, and secondary growth forest ages.



The classification contained < 1% of pixels which could not be correctly classified with the training sites. Shadow pixels (mainly along the edges of forest and large rivers) were mis-classified as either water or burned soils; these pixels are, for the most part, isolated or in groups of 2 or 3. In order to correct this problem the following five steps were taken: 1) re-code the classified image into a binary mask containing only water and burned soil pixels (value of 1); 2) in this mask, group all connected pixels into separate entities; 3) filter out any groups smaller than 4 pixels in size, leaving a layer of all of the water and burned soil pixels which were not to be altered; 4) obtain a shadow mask by recoding this layer so that all of these pixels had a value of 1; 5) overlay this mask on the classified image and, using a neighborhood analysis operating in a 5x5 window, change the shadow pixels to the value of their surroundings.

References:



Guptill, S. C., & Morrison, J. L. (1995). Elements of spatial data quality. UK: Elsevier Press.





Kimes, D.S.L., Nelson, R.F., Skole, D.L., & Sala, W.A. (1998). Accuracies in mapping secondary tropical forest age from sequential satellite imagery. Remote Sensing of the Environment, 65, 112-120.





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 ]





Roberts, D. A., Numata, I., Holmes, K., Batista, G., Krug, T., Monteiro, A., Powell, B., & Chadwick, O. A. (2002). Large area mapping of landcover change in Rondonia using multitemporal spectral mixture analysis and decision tree classifiers. Journal Geophysical Research. [LBA-ECO Pub ID = 254 ]

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

NASA logo
ORNL DAAC
Get Acrobat Reader