A Map of the Vegetation of South America Based on Satellite Imagery

Summary:

This dataset is a 1 km resolution land cover map of South America produced from 1-15 km NOAA AVHRR data (from the period 1988-1990). Thirty-nine land cover classes are distinguished, including deforestation.

1. Data Set Overview

Data Set Identification:
A Map of the Vegetation of South America Based on Satellite Imagery
Data Set Introduction:
This dataset is a 1 km resolution land cover map of South America produced from 1-15 km data available through the Advanced Very High Resolution Radiometer (AVHRR) weather satellites of the National Oceanic and Atmospheric Administration (NOAA) from the period 1987-1991. Thirty-nine land cover classes are distinguished, including deforestation.
Objective/Purpose:
The purpose of this map was to develop an objective appraisal of landuse change over the South American conitinent
Summary of Parameters:
Land Cover Classes
There are 39 landcover classes distinguished in this dataset and these were consolidated into 13 broader groups (see DATA DESCRIPTION, Data Characteristics, Variable Description/Definition for a complete listing).
Discussion:
Not Available
Related Data Sets:
Not Applicable

2. Investigator(s)

Investigator(s) Name and Title:
Stone, T.A., M.A., Schlesinger, P., M.A., Woodwell, G.M., Ph.D., and Houghton, R.A. Ph.D.
Title of Investigation:
A Map of the Vegetation of South America Based on Satellite Imagery
Contact Information:
Thomas A. Stone, Peter Schlesinger, George M. Woodwell, and Richard A. Houghton
The Woods Hole Research Center
PO Box 296, 13 Church Street
Woods Hole, MA 02543
Phone: (508)540-9900
Fax: (508)540-9700
Email: tstone@whrc.org

3. Theory of Measurements

Not Available

4. Equipment

Sensor/Instrument Description:
Not Available
Collection Environment:
Not Available
Source/Platform:
Not Available
Source/Platform Mission Objectives:
Not Available
Key Variables:
Not Available
Principles of Operation:
Not Available
Sensor/Instrument Measurement Geometry:
Not Available
Manufacturer of Sensor/Instrument:
Not Available
Calibration:
Specifications:
Not Available
Tolerance:
Not Available
Frequency of Calibration:
Not Available
Other Calibration Information:
Not Available

5. Data Acquisition Methods

The data were acquired from The Woods Hole Research Center, Woods Hole, MA, see World Wide Web address: http://www.whrc.org

6. Observations

Data Notes:
Not Available
Field Notes:
Not Available

7. Data Description

Spatial Characteristics:
Spatial Coverage:
Location
Min X -81.3572464
Max X -34.8444290
Min Y -56.0928993
Max Y 12.4712601
In decimal degrees of Longitude and Latitude
Spatial Coverage Map:
Not Available
Spatial Resolution:
1 km x 1 km
Projection:
Geographic
Grid Description:
7624 rows by 5172 columns
Temporal Characteristics:
Temporal Coverage:
The source satellite data spanned the period from 1987-1991.
Temporal Coverage Map:
Not Applicable
Temporal Resolution:
Not Applicable
Data Characteristics:
Parameter/Variable:
Land Cover Classes
Variable Description/Definition:
Land Cover Classes

The 39 landcover classes distinguished in this dataset were consolidated into 13 broader groups listed below. They include:

Group 1. Intact Tropical Moist Forests. Includes Semi-Deciduous Tropical Moist Forests, Tropical Moist Forest with Bamboo, and Tropical Gallery Forests. These regions have relatively high vegetation vigor throughout the year. All are considered to be closed canopy forests. This group is a sub-set of Group 3.

Group 2. Degraded Tropical Moist Forest and Secondary Forest in the TMF region. Includes large blocks of cleared forest surrounded by intact tropical moist forest that were both spectrally and visually distinct. As more clearing occurs and as older clearings are abandoned to secondary regrowth the distinction becomes less well-defined. We assumed that the observed visual patterns of cleared forest were relatively recent and probably have occurred within the last 10 years. This group is a sub-set of Group 4.

Group 3. Intact Closed Forest. Includes all of Group 1 as well as Tropical Seasonal or Deciduous forest, Montane Forest and Cool Deciduous Forest and Deciduous Temperate Forest. Deciduous forests have a relatively high vegetation vigor during a part of the year. We have made this category because other classifications of forests of South America have included closed forest as a distinct category.

Group 4. Degraded Closed Forests. This includes all of Group 2 and Degraded Deciduous Temperate Forest, Secondary Seasonal Forest with Agricultural Activity, Urban Regions, Degraded Tropical Seasonal Forest and Mixed Pine with secondary forest and agriculture. Most of these categories were defined based on information from either the Brazilian vegetation map (IGBE 1988) or the UNESCO map (1980).

Group 5. Intact Woodlands. Includes Seasonally Deciduous Woodlands, Xerophytic Woodlands, Montane Woodlands, Cool Deciduous Woodlands and Tropical Open Forest Mixed. Major woodland categories included areas in Brazil described as Sertao or Caatinga and Cerrado and areas in Argentina and Paraguay described as Chaco. Typically these regions have a short growing season and are moisture limited.

Group 6. Degraded Woodlands. Includes Degraded Seasonally Deciduous Woodlands, Degraded Xerophytic Woodlands (Thornforest) and Montane Degraded Woodlands. Many of these categories were defined based on information from either the Brazilian vegetation map (IGBE 1988) or the UNESCO map (1980) or by pattern recognition.

Group 7. Intact Grasslands. Includes Savanna Grasslands and Pasture, Seasonally Flooded grasslands (Pantanal), Montane Grasslands, and Tundra or Polar Grasslands. These included the Pampas of Argentina and Uruguay, the Beni of Bolivia, the Llanos of Colombia and Venezuela, and some portions of the Altiplano of Bolivia and the Puna of Peru.

Group 8. Degraded Grasslands. Includes Agriculture, Grasslands with Agricultural Activity, and Montane Degraded Grasslands. Most agriculture was defined based on the UNESCO map and on numerous LANDSAT satellite photos of the Parana, Paraguay and Rio de la Plata river system.

Group 9. Shrub and Scrublands. Includes Xerophytic Scrubland, Xerophytic Littoral Vegetation, and Cool Deciduous Scrubland. Most of these regions are in Patagonia or along the dry coasts of Peru, Chile and Venezuela. Generally these regions have short and weak growing seasons limited by rainfall or by cold.

Group 10. Desert, Bare Soil, or Inland Salt Marsh Communities.

Group 11. Open Water.

Group 12. Snow, Rock, and Ice.

Group 13.Other: Wet Vegetation (mixed water and land pixels) and Mangroves. and Unclassified.

Unit of Measurement:
Unitless
Data Source:
Unsupervised Classifications of NOAA AVHRR LAC and GVI data
Data Range:
Not Available
Sample Data Record:
Not Available

8. Data Organization

Data Granularity:

A general description of data granularity as it applies to the IMS appears in the EOSDIS Glossary.

This dataset consists of a single tarred and GNU-zipped of the filename SAM39CL2.taz comprised of 2,997,437 8-bit bytes.

Data Format:

Within the tarred and Gnu-gzipped file file are four files: a single flat binary raster image file, made up of 5172 columns by 7624 rows comprising 39,431,328 8-bit bytes; an ASCII documentation file (SAM39CL2.DOC); a graphic image of this dataset in ..JPG (JPEG) format (SAM39CL2.JPG) and an ASCII text file describing the projection parameters of the source data used to create this dataset (SAM39CL2.PRJ).

The structure of the ASCII documentation files is as follows (portions have been copied directly from the IDRISI for Windows v. 2.0 Help System, with the permission of the IDRISI Project, Clark University, Worcester, MA):

ITEM DESCRIPTION
title A descriptive name of the file.
data type The type of numbers stored in the file. Allowable entries are byte, integer and real.
file type The format in which the Image file is stored.
columns The number of columns in the image.
rows The number of rows in the image.
ref. system The name of the geographic referencing system used with the file.
ref. units The unit of measure used in the specified reference system. Allowable entries are m, ft, mi, km, deg and radians.
unit dist The scaling factor between the given coordinates and actual measurements on the ground.
min X The minimum X coordinate (left edge) of the image.
max X The maximum X coordinate (right edge) of the image.
min Y The minimum Y coordinate (bottom edge) of the image.
max Y The maximum Y coordinate (top edge) of the image.
pos'n error A measure of the accuracy of the positions in the image.
resolution The inherent resolution of the image. In most cases, this should correspond to the result of dividing the range of reference coordinates in X by the number of columns in the image.
min value The minimum value in the image.
max value The maximum value in the image.
value units The unit of measure of the values in the image. The term classes is used for all qualitative data sets, and that whenever standard linear units are appropriate, that the same abbreviations that are used for reference units should also be used (m, ft, mi, km, deg, rad).
value error This field records the error in the data values that appear in image cells. For qualitative data, this should be recorded as a proportional error. For quantitative data, the value here should be an RMS error figure.
flag value Any value in the image that is not a data value, but rather has a special meaning. If there is no flag value, this entry should remain blank.
flag def'n Definition of the above flag value. The most common data flags are those used to indicate background cells and missing data cells.
legend cats The number of legend categories present.
lineage Description of the history by which the values were recorded/derived.
completeness The degree to which the values describe the subject matter indicated.
consistency The logical consistency of the file.

9. Data Manipulations

Formulae:
Derivation Techniques and Algorithms:
Not Available
Data Processing Sequence:
Processing Steps:

The primary reliance in this work was on the NOAA AVHRR Local Area Coverage (LAC) satellite data. These data have a resolution of 1.1 km at nadir and are available for the whole earth twice daily. They were supplemented by higher resolution satellite imagery available for certain sections of the continent, by photographs, by earlier maps, and by personal experience on the ground. Because the map produced from the work described here is digital, it has no specific map scale. The scale of a map is determined by the size of the paper map printed. If we were to print the map with each 1 km2 cell at 1 mm2, a reasonable choice, the map scale would be 1:1,000,000.

Source data included NOAA 9, 10 and 11 AVHRR satellite imagery for South America. The above-named satellites acquire digital reflectivity and emissivity data from the surface of the earth from the visible red (0.58 - 0.68 microns), near-infrared (0.725 - 1.1 microns), mid-infrared (3.5 - 3.93 microns) and thermal (10.3 - 11.3 and 11.5 - 12.5 microns) portions of the electromagnetic spectrum (Kidwell 1988). Thirty-four computer compatible tapes were used, of which the majority were from 1988, the year of primary focus for this project. Others data were from 1987, 1989, 1990, and 1991 (see REFERENCES). The AVHRR data were purchased from the US Geological Survey's (USGS) EROS Data Center (EDC) in Sioux Falls, South Dakota. The AVHRR data were converted from the original 10 bit digital number format to an 8 bit format and corrected for radiometric and atmospheric effects. All five bands of AVHRR data were rectified to a latitude-longitude grid with 1 km resolution. With the LAC data acquired, we were able to cover, cloud-free, 69.6% of South America.

Because cloud-free 1 km resolution data was not available for all of all South America, a three year (1986-1988) weekly data set of 15 km resolution Global Vegetation Index (GVI) data from NOAA (Kidwell 1988) was utilized in the place of the missing 1 km data. The use of the 15 km data accounted for 30.4 % of final version the map. Various sources of supplemental information have been used (see REFERENCES).

Processing Changes:
Not Applicable
Calculations:
Special Corrections/Adjustments:
Not Applicable
Calculated Variables:
Not Applicable
Graphs and Plots:
Not Applicable

10. Errors

Sources of Error:
Not Available
Quality Assessment:
Data Validation by Source:
Not Available
Confidence Level/Accuracy Judgement:
Not Available
Measurement Error for Parameters:
Not Available
Additional Quality Assessments:
Not Available
Data Verification by Data Center:
Not Available

11. Notes

Limitations of the Data:
Not Available
Known Problems with the Data:
Not Available
Usage Guidance:
Latitude of True Scale: 0.0
False Easting: 0.0
False Northing: 0.0
Pixel Dimension: 1000 meters
Any Other Relevant Information about the Study:
Not Available

12. Application of the Data Set

Regional, national, and sub-national-level vegetation and landuse change assessments

13. Future Modifications and Plans

Not Available

14. Software

Software Description:
Two softwares are required to read the files in this dataset:
the shareware tar program tar.exe
the GNU compression utility gzip.exe
Software Access:
The GNU-gzip program (gzip.exe) and shareware tar program (tar.exe) are available via Anonymous FTP from the following site: wuarchive.wustl.edu, in the directory, /systems/msdos/gnuish, files: gzip124x.zip and gnutar.zip

15. Data Access

Contact Information:
1) Source Data Contact:
Thomas A. Stone
The Woods Hole Research Center
PO Box 296, 149 Woods Hole Rd.
Woods Hole, MA 02543
Phone: (508)540-9900
Fax: (508)540-9700
Email: tstone@whrc.org
Data Center Identification:
Not Applicable
Procedures for Obtaining Data:
see Woods Hole Research Center FTP site: 1. Land cover of South America at 1 km resolution based on satellite imagery (sa39cl2.exe or sa39cl2.taz).
Data Center Status/Plans:
Not Applicable

16. Output Products and Availability

Not Applicable

17. References

The published reference for this work is:
Stone, T.A., P. Schlesinger, G.M. Woodwell, and R.A. Houghton, 1994. A Map of the Vegetation of South America Based on Satellite Imagery. Photogrammetric Engineering and Remote Sensing. 60(5):541-551.
Satellite Imagery Data Used in Ordered by Date
Satellite     Date Acquired Satellite Scene ID
NOAA 09 January 08 1987 AL09010887205540
NOAA 09 January 09 1987 AL09010987204440
NOAA 09 February 17 1988 AL09021788203510
NOAA 09 March 09 1988 AL09030988200820
NOAA 09 July 01 1988 AL09070188192520
NOAA 09 July 02 1988 AL09070288191410
NOAA 09 July 03 1988 AL09070388190310
NOAA 09 July 03 1988 AL09070388204330
NOAA 09 July 04 1988 AL09070488203230
NOAA 09 July 06 1988 AL09070688201210
NOAA 09 July 09 1988 AL09070988193900
NOAA 09 July 10 1988 AL09071088192800
NOAA 09 July 12 1988 AL09071288190550
NOAA 09 July 12 1988 AL09071288205040
NOAA 09 July 14 1988 AL09071488202400
NOAA 09 July 15 1988 AL09071588201440
NOAA 09 July 17 1988 AL09071788195230
NOAA 09 July 20 1988 AL09072088191920
NOAA 09 August 16 1988 AL09081688192820
NOAA 11 December 21 1988 AL11122188182200
NOAA 11 December 29 1988 AL11122988183550
NOAA 11 December 31 1988 AL11123188181559
NOAA 11 January 11 1989 AL11011189180550
NOAA 10 July 17 1990 AL10071790110200
NOAA 11 July 22 1990 AL11072290104721
NOAA 11 July 31 1990 AL11073190175240
NOAA 11 August 29 1990 AL11082990173640
NOAA 11 September 15 1990 AL11091590175229
NOAA 11 November 16 1990 AL11111690181139
NOAA 11 January 03 1991 AL11010391192442
NOAA 11 January 13 1991 AL11011391191343
NOAA 11 January 20 1991 AL11012091193647
NOAA 11 February 08 1991 AL11020891192554
NOAA 11 February 18 1991 AL11021891191340
Atlas and Map References Used:

Banco Central del Ecuador. 1982. Atlas del Ecuador. Paris: Les Editions J.A.

CIA, 1971. Argentina [map 1:7,550,000], 500044 10-71, Central Intelligence Agency, Washington.

CIA, 1983. Uruguay. Central Intelligence Agency, Washington, D.C.

Defense Mapping Agency, various dates. Operational Navigation Chart(s) K26, K27, L26-28, M25-29, N25-28, P26-28, Q26-28, R23-24, S21, and T-18, Scale 1:1,000,000. St. Louis, Mo.

DNPM (Dept. National da Producao Mineral) Projeto Radambrasil, (several dates and volumes), Rio De Janeiro, Brazil.

Geomundo, 1985. Atlas De Bolivia, Comando General del Ejercito Ediciones GEOMUNDO, Barcelona 1998 pp.

Huber O. and C. Alarcon, 1988. Mapa de Vegetacion De Venezuela [1:2,000,000], Ministerio Del Ambiente Y de Los Naturales Renovables, Republic de Venezuela, Caracas.

Hueck, K. 1972. Vegetationskarte Von Sudamerika [1:8,000,000]. Gustav Fischer Verlag, Stuttgart.

IBDF, 1982. Mapa de Alteracao Da Cobertura Vegetal Natural, Programma de Monitoramento Da Coberatura Florestal Do Brasil, Brasilia, DF.

IBDF, 1988. Mapa de Alteracao Da Cobertura Florestal, Programma de Monitoramento Da Coberatura Florestal Do Brasil, Estado do Para.

IBGE/IBDF, 1988. Mapa de Vegetacao do Brasil. Ministerio da Agricultura.

Instituto Geografica Militar (Paraguay). 1977. Atlas del Paraguay.

Instituto Geografica Militar (Chile). 1988. Atlas Geografico de Chile. Santiago.

Instituto Geografico Militar (Argentina). Atlas de la Republica Argentina.

Instituto Geografica Militar (Ecuador). 1978. Atlas Geografico de la Republica del Ecuador.

International Travel Map Productions, 1990-91, South America South 1:4,000,000, Lambert's Azimuthal Equal-Area Projection, Vancouver, Canada.

International Travel Map Productions 1986-87, South America North 1:4,000,000, Lambert's Azimuthal Equal-Area Projection, Vancouver, Canada.

International Travel Map Productions, 1988-89, South America Northeast 1:4,000,000, Lambert's Azimuthal Equal-Area Projection, Vancouver, Canada.

Ministeria de Defensa. 1989. Proyecto Especial Atlas del Peru. Lima: Instituto Geograf National.

Ministerio de Agricultura y CRIA. 1960. Atlas Agricola de Venezuela. Vegetacion Mapa No. 12. Direccion de Planificacion Agropecuaria.

Ministério Da Agricultura, 1985. Estado de Rondônia, Mapa da Alteraçao Cobertura Vegetal Natural, Dept. de Economia Florestal, Programa de Monitoramento da Cobertura Florestal do Brasil, Brasilia, D.F.

National Geographic Society. 1983. South America. 1:10,700,000. Chamberlin Trimetric Projection.

OAS. 1964. Indice Anotado de los Trabajos Aerofotograf y los Mapas Topo.

OAS. 1988. Suriname Planatlas. The National Planning Office of Suriname (SPS), Regional Development and Physical Planning Department and The Organization of American States (OAS), Executive Secretariat for Economic and Social Affairs, Department of Regional Development (DRD).

Preciada, Alfonso Perez. 1989. Atlas y geografia de Colombia. Bogota: Circulo de Lectores, S.A.

UNESCO, 1980. Vegetation Map of South America [1:5,000,000], United Nations Educational, Scientific, and Cultural Organization, Paris.

UNESCO, 1981. Vegetation Map of South America Explanatory Notes. United Nations Educational, Scientific, and Cultural Organization, Paris, 189 pp.

World Resources Institute, 1990. World Resources 1990-1991. Oxford University Press, New York, 383 pp.

Other References Used in this Metadata:

Eastman, J.R., 1997. IDRISI for Windows Version 2.0. Clark Labs for Cartographic Technology and Geographic Analysis, Clark University, Worcester, MA

Kidwell, K., 1988. NOAA Polar Orbiter Data Users Guide, NOAA/NESDIS, Nat. Clim. Data Center, Washington, D.C.

18. Glossary of Terms

Not Available

19. List of Acronyms

Acronym Definition
ASCII American Standard Code for Information Interchange
AVHRR Advanced Very High Resolution Radiometer
EDC EROS Data Center
EROS Earth Resources Observation Systems
GVI Global Vegetation Index
GNU GNU's not UNIX
JPEG Joint Photographic Experts Group
LAC Local Area Coverage
NOAA National Oceanic and Atmospheric Administration
PAL Pathfinder AVHRR Land program
RMS Root Mean Square
TMF Tropical Moist Forest
USGS United States Geologic Survey

20. Document Information

Document Revision Date:
October 26, 2004
Document Review Date:
Not Available
Document ID:
(currently leave this blank)
Citation:
(currently leave this blank)
Document Curator:
Not Available
Document URL:
http://whrc.org/LBA_data/Braz_Veg_AVHRR.htm