Identification_Information:

Identification_Information:

Citation:

Citation_Information:

Originator:

Kurt Riitters, U.S. Forest Service, Southern Research Station,

Research Triangle Park, NC USA.

Publication_Date: 2000

Title: Global Forest Fragmentation, Area Density, and Connectivity

Geospatial_Data_Presentation_Form: Map

Publication_Information:

Publication_Place: Research Triangle Park, NC

Publisher: U.S. Forest Service

Description:

Abstract:

This data set contains 18 grid maps that were used in a global

assessment of forest fragmentation (Riitters et al., 2000).

The maps portray three indices of forest spatial pattern -

forest area density, forest connectivity, and forest fragmentation -

at 1 sq. km. (100 ha) resolution. The first two indices were derived

from existing land-cover maps and were combined to produce the forest

fragmentation index that was reported in the assessment. The data set

contains one global map and five continental maps for each of the three

indices. A pixel value represents one of the indices for the surrounding

81 sq. km. (8100 ha).

The first two indices were created by applying spatial algorithms to 1 sq.

km. resolution maps of land cover obtained from the USGS EROS Data Center

Distributed Active Archive Center (DAAC) as part of the Global Land Cover

Characteristics database (GLCC)(Loveland et al. 1991, 1999). Forest area

density is the proportion of pixels in the surrounding 8100 ha that are

forest. Forest connectivity is the proportion of adjacent pixel pairs in

the surrounding 8100 ha that are both forest, given that at least one of

a pair is forest. The forest fragmentation index was created by

overlaying the first two index maps and classifying each forest pixel

according to the observed combination of forest area density and forest

connectivity for that pixel (Riitters et al., 2000).

The maps are in Lambert Azimuthal projection for five continents (15

maps), and in Interrupted Goode homolosine projection for the globe (3

maps). The map projections used are the same as the GLCC used for the

continental and global land-cover maps. The maps are in a generic

binary format (again, same as the land-cover maps) suitable for import

into geographic information systems or image processing systems.

The data set describes three aspects of forest fragmentation at one scale.

The forest area density index describes the absolute amount of forest in a

defined region, and the forest connectivity index describes the tendency

for forest pixels in the region to be adjacent (as opposed to not

adjacent). The forest fragmentation index distinguishes among types of

fragmentation (such as edges on the interior versus the exterior of a

forest patch) and it also reflects differences in the absolute amount of

forest present. No distinction was drawn between "natural" and "human-

caused" fragmentation.

Purpose:

These data were originally created as part of a global analysis of forest

fragmentation and other land-cover patterns based on digital land-cover

maps derived from remote sensing and produced by the Global Land Cover

Characteristics project. The maps portray continental patterns at

relatively coarse scale and are considered a first step towards

quantifying forest fragmentation and its potential impacts on biodiversity

at landscape-scale levels of biological organization.

Riitters et al. (2000) reported the forest fragmentation index after

stratification by individual forest types, and did not directly report

either of the other two indices. The maps in this data set enable other

investigators to modify the fragmentation index and (or) to summarize the

index values according to other stratification variables. Information

about the stratification is contained in the Supplemental_Information

section of this metadata document.

No responsibility is assumed by the U.S. Forest Service in the use of

these data.

Supplemental_Information:

The primary data source for the underlying GLCC land-cover maps is AVHRR

(advanced very high resolution radiometer) satellite imagery from the

early 1990's. Additional information about the GLCC is available at:

http://edcdaac.usgs.gov/glcc/glcc.html

Referring to Riitters et al. (2000) the continental maps of the forest

fragmentation index in this database correspond to Figures 9, 10, 11,

12, and 13, and the global map corresponds to Figure 5. The maps of

the forest area density index and the forest connectivity index do

not appear in the manuscript.

In the online manuscript (Riitters et al., 2000), the following color

scheme was used to render the fragmentation index.

Index

value Meaning Color RGB values

1 edge orange 255 110 0

2 undetermined light green 129 255 129

3 perforated yellow 255 255 0

4 interior dark green 34 255 34

5 patch dark blue 56 56 255

6 transitional light blue 162 162 255

7 unlabeled land area light gray 230 230 230

The following narrative describes an additional process step used by

Riitters et al. (2000) for stratification of the continental maps

of the forest fragmentation index. This process step was NOT used

for the maps in the database described by this metadata; this

information is only here in case someone wants to recreate or modify

the stratification analysis reported in the online manuscript:

The continental maps of the fragmentation index were masked by

using the corresponding OGE land-cover map, such that only cells

identified as forest on the OGE maps retained a non-missing index value.

This was done in order to post-stratify the fragmentation index map by

OGE forest classes for assessment of forest class differences. For the

continental masks the following OGE classes were considered to be forest:

5,28,29,33,34,48 (Africa); 6,28,29,33,48,78,89 (Australia-Pacific);

4,5,21,22,23,24,25,27,29,33,34,60,61,62,90 (Europe-Asia);

3,5,20,21,22,23,24,25,26,27,29,32,33,34,60,62 (North America), and;

6,27,28,29,32,33,34,54,78 (South America).

Note that the OGE maps contain about 10 percent less forest than the IGBP

maps since there is not a 1:1 correspondence between OGE forest and IGBP

forest. The ARC/INFO COMBINE command was used to combine index values

from the forest fragmentation index grid, and the OGE forest type grid,

and the value attribute table (vat) was examined to recover the required

statistics.

Time_Period_of_Content:

Time_Period_Information:

Range_of_Dates/Times:

Beginning_Date: 1992

Ending_Date: 1993

Currentness_Reference: Dates of satellite imagery.

Status:

Progress: Complete

Maintenance_and_Update_Frequency: Irregular

Spatial_Domain:

Bounding_Coordinates:

West_Bounding_Coordinate: -180

East_Bounding_Coordinate: 180

North_Bounding_Coordinate: 90

South_Bounding_Coordinate: -90

Keywords:

Theme:

Theme_Keyword_Thesaurus: None

Theme_Keyword: pattern

Theme_Keyword: land cover

Theme_Keyword: forest

Theme_Keyword: fragmentation

Theme_Keyword: landscape

Theme_Keyword: connectivity

Theme_Keyword: density

Place:

Place_Keyword_Thesaurus: None

Place_Keyword: Earth

Place_Keyword: Africa

Place_Keyword: Asia

Place_Keyword: Australia

Place_Keyword: Europe

Place_Keyword: North America

Place_Keyword: Pacific

Place_Keyword: South America

Access_Constraints: None

Use_Constraints:

None. Citation of Riitters et al. (2000) and acknowledgement of the U.S.

Forest Service would be appreciated in products derived from these

data.

Point_of_Contact:

Contact_Information:

Contact_Person_Primary:

Contact_Person: Kurt Riitters

Contact_Organization:

U.S. Forest Service, Southern Research Station

Contact_Address:

Address_Type: mailing address

Address:

3041 Cornwallis Road

City: Research Triangle Park

State_or_Province: NC

Postal_Code: 27709

Contact_Voice_Telephone: 919-549-4015

Contact_Electronic_Mail_Address: kriitters@fs.fed.us

Data_Set_Credit:

The basic data from which these maps were derived were provided by

the EROS Data Center, Global Land Cover Characteristics Project.

Cooperators included the US Geological Survey, the University of

Nebraska - Lincoln, and the European Commission's Joint Research

Centre. The effort is part of the National Aeronautics and Space

Administration (NASA) Earth Observing System Pathfinder Program.

Funding for the project was provided by the USGS, NASA, the U.S.

Environmental Protection Agency (EPA), National Oceanic and Atmospheric

Administration (NOAA), U.S. Forest Service (USFS), and the United

Nations Environment Programme.

Native_Data_Set_Environment:

The maps of the forest area density index and the forest connectivity

index were prepared as generic binary image files, output from C programs

running under SunOS 5.6 and (or) RedHat Linux 5.0. These were converted to

grid format, ESRI ARC/INFO version 7.2.1 running under SunOS 5.6, and

combined to prepare the maps of the forest fragmentation index. All maps

are available as generic binary image files (band sequential format)

exported from ARC/INFO and compressed using gzip.

Data_Quality_Information:

Attribute_Accuracy:

Attribute_Accuracy_Report:

There is no information available about the thematic accuracy of the

indices at the scale at which they were measured. Based on superficial

comparisons to similar maps prepared from 30-meter resolution land-cover

maps for the United States, at least some of the North America map

does a reasonable job of portraying regional patterns. However, forest

fragmentation is a scale-dependent measurement and exact correspondence

is not expected with finer- or coarser-scale analyses.

The statistical accuracy can be gauged somewhat. Two of the

indices are calculated from proportions that are measured in a 9x9-pixel

analysis window - the proportion of pixels that are forest (for the forest

area density index), and the proportion of adjacent pixel pairs that are

both forest, given that one is forest (for the forest connectivity index).

The window contains 81 pixels, which provides more than the minimum sample

size needed by a statistical rule of thumb that 50 observations are needed

to reliably estimate a proportion. The actual number of observations in a

window was sometimes less than 50 because water pixels were considered to

be missing values. There are at most 144 adjacent pixel pairs in the

window that could potentially be used for the connectivity index; however,

only pairs of pixels were counted that included at least one forest pixel,

and pairs of pixels involving water were considered to be missing values,

so the actual number was often less than 144, and sometimes less than 50.

The thematic accuracy of the fragmentation index map was subjectively

evaluated by visual comparisons with the land-cover maps (condensed to

forest and non-forest classes). Forest pixels in the ´patch' category

appear as islands on a non-forest background. Forest pixels in the

´edge', ´perforated', ´undetermined' and ´interior' categories

collectively make up a forest background, upon which islands of the

non-forest appear. Pixels in the ´edge' and ´perforated' categories

appear on the outer and inner fringes, respectively, of large patches

of forest.

The indices are just three of the many that could have been used to

quantify "forest fragmentation;" textbooks discuss the relative strengths

and weaknesses of various indices.

The fragmentation indices are best interpreted as relative values in

comparison to other pixels on the same map. Comparisons with other maps

prepared from different data sources or at different scale are tenuous.

Logical_Consistency_Report:

In the maps of the forest area density index and the forest connectivity

index, all cell values are in the range [0,255]. In the maps of the forest

fragmentation index, all cell values are in the range [0,18].

Completeness_Report:

Every cell identified as "land" in an original land-cover map potentially

has values for each of the three indices. Some of the cells in the forest

area density index map have missing values because there were less than 21

non-water cells in the window. Some of the cells in the forest connectivity

index map have missing values because there were no forest cells in the

window. Some of the cells in the forest fragmentation index map have

missing values because the cell in the underlying land-cover map was not

forest.

Positional_Accuracy:

Horizontal_Positional_Accuracy:

Horizontal_Positional_Accuracy_Report:

The positional accuracy of any of the derived maps is the same as the

positional accuracy of the land-cover map from which it was derived.

Lineage:

Source_Information:

Source_Citation:

Citation_Information:

Originator: U.S. Geological Survey

Publication_Date: 1998

Title:

GLCC Global Land Cover Characteristics Database, Version 1.2.

Publication_Information:

Publication_Place: Sioux Falls, SD

Publisher: EROS Data Center

Other_Citation_Details:

The procedures used in development of this map are described in:

Loveland, T.R., Merchant, J.W., Ohlen, D.O., Brown, J.F. 1991.

Development of a Land-Cover Characteristics Database for the

Conterminous U.S., Photogrammetric Engineering and Remote Sensing

57:1453-1463.

Additional information is contained in: Loveland, T.R. and A.S.

Belward. 1997. The IGBP-DIS Global 1 km Land Cover Data Set,

DISCover First Results. International Journal of Remote Sensing

18:3289-3295.

Additional information is contained in: Loveland, T.R., B.C.

Reed, J.F. Brown, D.O. Ohlen, Z. Zhu, L. Yang, and J.W.

Merchant. 1999. Development of a Global Land Cover Characteristics

Database and IGBP DISCover from 1-km AVHRR Data. International

Journal of Remote Sensing, (In press).

The GLCC provides several map legends; the IGBP (International

Geosphere Biosphere Programme) legend was used to create the

maps in this data set. Also, the GLCC provides several map

projections; the Lambert Azimuthal projection was obtained for

the continental maps and the Goode Interrupted Homolosine

projection was obtained for the global maps used in this study.

Also, the GLCC provides continental maps for Europe-Asia with

projections optimized for Europe and for Asia; the Asia-optimized

projection was used in this study.

The land-cover maps used in this analysis are referred to by

the following source citation abbreviations:

Source

Citation

Abbreviation GLCC Land Cover Map Identification

AF_IGBP Africa, IGBP legend, Lambert projection

AP_IGBP Australia-Pacific, IGBP legend, Lambert projection

EA_IGBP Europe-Asia, IGBP legend, Lambert projection

NA_IGBP North America, IGBP legend, Lambert projection

SA_IGBP South America, IGBP legend, Lambert projection

GL_IGBP Global, IGBP legend, Goode projection

Type_of_Source_Media: online

Source_Time_Period_of_Content:

Time_Period_Information:

Range_of_Dates/Times:

Beginning_Date: 1992

Ending_Date: 1993

Source_Currentness_Reference: Dates of satellite imagery.

Source_Citation_Abbreviation:

AF_IGBP AP_IGBP EA_IGBP NA_IGBP SA_IGBP GL_IGBP

Source_Contribution: Land-cover maps

Source_Information:

Source_Citation:

Citation_Information:

Originator: Riitters, K.H.

Originator: Wickham, J.D.

Originator: O'Neill, R.V.

Originator: Jones, K.B.

Originator: Smith, E.R

Publication_Date: 2000

Title: Global Scale Patterns of Forest Fragmentation

Larger_Work_Citation:

Citation_Information:

Originator: Ecological Society of America

Publication_Date: 2000

Title: Conservation Ecology (online)

Online_Linkage: http://www.consecol.org/Journal/

Type_of_Source_Media: journal article

Source_Time_Period_of_Content:

Time_Period_Information:

Single_Date/Time:

Calendar_Date: 2000

Source_Currentness_Reference: Publication date

Source_Citation_Abbreviation: GSPFF

Source_Contribution:

Derivation and description of fragmentation index.

Source_Information:

Source_Citation:

Citation_Information:

Originator: U.S. Forest Service

Publication_Date: 2000

Title: Global Maps of Forest Pattern Indices

Publication_Information:

Publication_Place: Research Triangle Park, NC

Publisher: Southern Research Station

Other_Citation_Details:

The creation of these maps is described in the

first process step in this metadata.

Type_of_Source_Media: internal files

Source_Time_Period_of_Content:

Time_Period_Information:

Single_Date/Time:

Calendar_Date: 2000

Source_Currentness_Reference: Publication date

Source_Citation_Abbreviation:

AFFDEN9 APFDEN9 EAFDEN9 NAFDEN9 SAFDEN9 GLFDEN9

AFFRAG9 APFRAG9 EAFRAG9 NAFRAG9 SAFRAG9 GLFRAG9

AFCL_9 APCL_9 EACL_9 NACL_9 SACL_9 GLCL_9

Source_Contribution:

Source files for maps included in this database.

Process_Step:

Process_Description:

Derivation of forest area density, forest connectivity, and

forest fragmentation indices from land-cover maps:

Land-cover data were obtained from EROS Data Center in band

sequential (bsq) format. For each map, a header was created using

metadata from GLCC and the format was converted to an in-house

format using an in-house software tool named BSQ2TVA.C.

Each land-cover map was subdivided into four overlapping rectangles

using an in-house software tool named SPLITTER.C. The rectangles

overlapped to avoid artifacts near image boundaries during subsequent

spatial filtering operations.

The in-house spatial filtering program SPATCONV.C was used to process

each of the four rectangles, to measure the proportion of forest and the

connectivity of forest, within 9x9-pixel windows centered on each land

pixel. Adopting the terminology in Riitters et al. (2000), let Pf be

the proportion of forest, and Pff be the connectivity of forest, within

9x9-pixel windows on the land-cover map. Pf is the forest area density

index. Pff is the forest connectivity index.

The procedure is briefly described as follows. The IGBP legends for the

continental and global land-cover maps were condensed to three classes,

forest (IGBP codes 1, 2, 3, 4, 5), water (IGBP codes 15, 17), and other

(the remaining IGBP codes). Note that the forest class does not

include woody savannas and savannas (IGBP codes 8, 9) that are often

considered to be forest by international standards. The water

category was considered to be missing data.

Next, a proportion-of-forest value (Pf) was determined for each pixel by

calculating the proportion of forest pixels in the 9x9-pixel window

centered on the subject pixel. Water pixels were excluded from the

calculation.

Also, a connectivity-of-forest value (Pff) was determined for each

pixel by calculating the proportion of adjacent pixel pairs

that are both forest, given that at least one of a pair was forest,

for pixel pairs within a 9x9-pixel window centered on the subject pixel.

"Adjacent pixels" were defined as being in cardinal directions only, and

water pixels were excluded from the calculation. Pff (roughly)

estimates the probability that, given a pixel of forest, its neighbor

was also forest. Forest connectivity is higher for larger values of

Pff.

The intermediate indices (Pf and Pff) are continuous variables that

range from zero to one. The calculated values were discretized to the

range [1,255] and stored at one-kilometer spatial resolution. The

transformation used to discretize the values was:

D = (C * 254) + 1

where

D = discretized value in range [1,255]

C = calculated value of Pf or Pff in range [0,1]

Note that precision is lost by the application of the formula; the

purpose of the transformation was to permit data storage as byte

values instead of as floating point numbers.

Note that an integer value of "1" corresponds to an index value of 0.0

and in this way the integer value "0" was reserved for missing

values. Pf and Pff were set to missing if more than three-quarters of

the cells in an analysis window were missing (water or background).

Pff was also set to missing if Pf was zero (there is no information

about forest connectivity where there is no forest). The integer values

were stored at one-kilometer spatial resolution, at the location of the

center cell of the analysis window. Thus, a pixel value in these maps

represents a forest index value within the surrounding 81 sq. km.

on the original land-cover map.

The four rectangles of Pf and the four rectangles of Pff were exported

in an in-house format and reassembled into two single images via an

in-house software tool named LUMPER.C, which removed the overlapping

parts of the rectangles. The in-house software tool TVA2BSQ.C was then

used to convert the format to bsq and the resulting files imported into

ARC/INFO by using the command IMAGEGRID.

The nomenclature of the derived maps of the forest area density (Pf)

index and the forest connectivity (Pff) index is as follows:

AFFDEN9 - Pf for Africa derived from AF_IGBP

APFDEN9 - Pf for Australia-Pacific derived from AP_IGBP

EAFDEN9 - Pf for Europe-Asia derived from EA_IGBP

NAFDEN9 - Pf for North America derived from NA_IGBP

SAFDEN9 - Pf for South America derived from SA_IGBP

GLFDEN9 - Pf for the world derived from GL_IGBP

AFFRAG9 - Pff for Africa derived from AF_IGBP

APFRAG9 - Pff for Australia-Pacific derived from AP_IGBP

EAFRAG9 - Pff for Europe-Asia derived from EA_IGBP

NAFRAG9 - Pff for North America derived from NA_IGBP

SAFRAG9 - Pff for South America derived from SA_IGBP

GLFRAG9 - Pff for the world derived from GL_IGBP

Standard GIS overlay and analysis techniques were then used to evaluate

the values of Pf and Pff for each forested pixel and assign a

fragmentation index value. The classification model identifies six

fragmentation categories: [See erratum.]

(1) edge, if Pf > 0.6 and Pf - Pff > 0

(2) undetermined, if Pf > 0.6 and Pf = Pff

(3) perforated, if Pf > 0.6 and Pf - Pff < 0

(4) interior, if Pf = 1.0

(5) patch, if Pf < 0.4

(6) transitional, if 0.4 < Pf < 0.6

The calculated values were stored at one-kilometer spatial resolution.

Thus, a pixel value in this data represents the forest fragmentation

index within the surrounding 81 sq. km. on the original land-cover map,

based on the measured values of Pf and Pff for that analysis window.

Since the focus of the analysis was on the forest fragmentation context

for individual forest pixels (as opposed to the forest fragmentation

status of all possible landscapes in a geographic region), the forest

fragmentation index grids were masked such that any pixel that was not

forest (i.e., any pixel that did not have an IGBP code of 1, 2, 3, 4, or

5 on the land-cover map) was recoded as a missing value. The global map

of the forest fragmentation index was recoded further for presentation

purposes only, such that water pixels received the value 17, and map

background pixels (i.e., the space between the orange slices) received

the value 18.

The nomenclature of the derived maps of the forest fragmentation index

is as follows:

AFCL_9 - Forest fragmentation index for Africa

APCL_9 - Forest fragmentation index for Australia-Pacific

EACL_9 - Forest fragmentation index for Europe-Asia

NACL_9 - Forest fragmentation index for North America

SACL_9 - Forest fragmentation index for South America

GLCL_9 - Forest fragmentation index for the world

The georeferencing did not change as a result of reformatting,

exporting, importing, or spatial filtering. The derived maps are

comparable on a pixel-by-pixel basis to the original land-cover maps.

Source_Used_Citation_Abbreviation: NA_IGBP

Source_Used_Citation_Abbreviation: AP_IGBP

Source_Used_Citation_Abbreviation: EA_IGBP

Source_Used_Citation_Abbreviation: NA_IGBP

Source_Used_Citation_Abbreviation: SA_IGBP

Source_Used_Citation_Abbreviation: GL_IGBP

Source_Used_Citation_Abbreviation: GSPFF

Source_Produced_Citation_Abbreviation: AFFDEN9

Source_Produced_Citation_Abbreviation: APFDEN9

Source_Produced_Citation_Abbreviation: EAFDEN9

Source_Produced_Citation_Abbreviation: NAFDEN9

Source_Produced_Citation_Abbreviation: SAFDEN9

Source_Produced_Citation_Abbreviation: GLFDEN9

Source_Produced_Citation_Abbreviation: AFFRAG9

Source_Produced_Citation_Abbreviation: APFRAG9

Source_Produced_Citation_Abbreviation: EAFRAG9

Source_Produced_Citation_Abbreviation: NAFRAG9

Source_Produced_Citation_Abbreviation: SAFRAG9

Source_Produced_Citation_Abbreviation: GLFRAG9

Source_Produced_Citation_Abbreviation: AFCL_9

Source_Produced_Citation_Abbreviation: APCL_9

Source_Produced_Citation_Abbreviation: EACL_9

Source_Produced_Citation_Abbreviation: NACL_9

Source_Produced_Citation_Abbreviation: SACL_9

Source_Produced_Citation_Abbreviation: GLCL_9

Process_Date: 2000

Process_Step:

Process_Description:

Export of ARC/INFO grid maps to generic binary format with

header files for distribution purposes.

The continental and global maps of forest area density (Pf), forest

connectivity (Pff), and the (masked) continental and global maps of the

fragmentation index were exported from ARC/INFO GRID format to a band-

sequential (extension "bsq") format by using the ARC/INFO command

GRIDIMAGE. The "header" information for each map is contained in a

separate ascii file with the extension "hdr". All of the maps have the

following header features in common:

Nbands 1 (one band of data)

Nbits 8 (one character or byte per cell)

Layout bsq (band sequential format)

Skipbytes 0 (no embedded header)

Xdim 1000 (1000 meter cell size in horizontal direction)

Ydim 1000 (1000 meter cell size in vertical direction)

The following header information varies with the specific map:

Nrows (number of rows in the map)

Ncols (number of columns in the map)

Ulxmap (X-coordinate of the location of the center of the upper-left

corner pixel, map units)

Ulymap (Y-coordinate of the location of the center of the upper-left

corner pixel, map units)

All of the Africa maps have this header information:

Nrows 9276

Ncols 8350

Ulxmap -4458000

Ulymap 4480000

All of the Australia-Pacific maps have this header information:

Nrows 8000

Ncols 9300

Ulxmap -5000000

Ulymap 4054109

All of the Europe-Asia maps have this header information:

Nrows 12000

Ncols 13000

Ulxmap -8000000

Ulymap 6500000

All of the North America maps have this header information:

Nrows 8996

Ncols 9223

Ulxmap -4487000

Ulymap 4480000

All of the South America maps have this header information:

Nrows 8000

Ncols 6000

Ulxmap -3000000

Ulymap 3100000

All of the global maps have this header information:

Nrows 17347

Ncols 40031

Ulxmap -20015000

Ulymap 8673000

The following information describes the map projections

for the continental and global grid maps:

Map_Projection:

Map_Projection_Name: Lambert Azimuthal Equal Area for Africa

Lambert_Azimuthal_Equal_Area:

Longitude_of_Projection_Center: 20

Latitude_of_Projection_Center: 5

False_Easting: 0.0

False_Northing: 0.0

Map_Projection:

Map_Projection_Name:

Lambert Azimuthal Equal Area for Australia-Pacific

Lambert_Azimuthal_Equal_Area:

Longitude_of_Projection_Center: 135

Latitude_of_Projection_Center: 15

False_Easting: 0.0

False_Northing: 0.0

Map_Projection:

Map_Projection_Name: Lambert Azimuthal Equal Area for Europe-Asia

Lambert_Azimuthal_Equal_Area:

Longitude_of_Projection_Center: 100

Latitude_of_Projection_Center: 45

False_Easting: 0.0

False_Northing: 0.0

Map_Projection:

Map_Projection_Name: Lambert Azimuthal Equal Area for North America

Lambert_Azimuthal_Equal_Area:

Longitude_of_Projection_Center: -100

Latitude_of_Projection_Center: 50

False_Easting: 0.0

False_Northing: 0.0

Map_Projection:

Map_Projection_Name: Lambert Azimuthal Equal Area for South America

Lambert_Azimuthal_Equal_Area:

Longitude_of_Projection_Center: -60

Latitude_of_Projection_Center: -15

False_Easting: 0.0

False_Northing: 0.0

Map_Projection:

Map_Projection_Name: Interrupted Goode homolosine for the world

All bsq files were compressed by using the gzip utility.

Source_Used_Citation_Abbreviation: AFFDEN9

Source_Used_Citation_Abbreviation: APFDEN9

Source_Used_Citation_Abbreviation: EAFDEN9

Source_Used_Citation_Abbreviation: NAFDEN9

Source_Used_Citation_Abbreviation: SAFDEN9

Source_Used_Citation_Abbreviation: GLFDEN9

Source_Used_Citation_Abbreviation: AFFRAG9

Source_Used_Citation_Abbreviation: APFRAG9

Source_Used_Citation_Abbreviation: EAFRAG9

Source_Used_Citation_Abbreviation: NAFRAG9

Source_Used_Citation_Abbreviation: SAFRAG9

Source_Used_Citation_Abbreviation: GLFRAG9

Source_Used_Citation_Abbreviation: AFCL_9

Source_Used_Citation_Abbreviation: APCL_9

Source_Used_Citation_Abbreviation: EACL_9

Source_Used_Citation_Abbreviation: NACL_9

Source_Used_Citation_Abbreviation: SACL_9

Source_Used_Citation_Abbreviation: GLCL_9

Source_Produced_Citation_Abbreviation: AFFDEN9.bsq.gz

Source_Produced_Citation_Abbreviation: APFDEN9.bsq.gz

Source_Produced_Citation_Abbreviation: EAFDEN9.bsq.gz

Source_Produced_Citation_Abbreviation: NAFDEN9.bsq.gz

Source_Produced_Citation_Abbreviation: SAFDEN9.bsq.gz

Source_Produced_Citation_Abbreviation: GLFDEN9.bsq.gz

Source_Produced_Citation_Abbreviation: AFFRAG9.bsq.gz

Source_Produced_Citation_Abbreviation: APFRAG9.bsq.gz

Source_Produced_Citation_Abbreviation: EAFRAG9.bsq.gz

Source_Produced_Citation_Abbreviation: NAFRAG9.bsq.gz

Source_Produced_Citation_Abbreviation: SAFRAG9.bsq.gz

Source_Produced_Citation_Abbreviation: GLFRAG9.bsq.gz

Source_Produced_Citation_Abbreviation: AFCL_9.bsq.gz

Source_Produced_Citation_Abbreviation: APCL_9.bsq.gz

Source_Produced_Citation_Abbreviation: EACL_9.bsq.gz

Source_Produced_Citation_Abbreviation: NACL_9.bsq.gz

Source_Produced_Citation_Abbreviation: SACL_9.bsq.gz

Source_Produced_Citation_Abbreviation: GLCL_9.bsq.gz

Process_Date: 2000

Spatial_Data_Organization_Information:

Direct_Spatial_Reference_Method: Raster

Raster_Object_Information:

Raster_Object_Type: Grid cell

Spatial_Reference_Information:

Horizontal_Coordinate_System_Definition:

Planar:

Local_Planar:

Local_Planar_Description:

The continental maps all use the Lambert Azimuthal

Equal Area Projection with different origins as

detailed in the process step above. The global

maps use the Interrupted Goode Homolosine projection

as detailed in the process step above.

Local_Planar_Georeference_Information:

The georeference information appears in the process

step above.

Planar_Coordinate_Information:

Planar_Coordinate_Encoding_Method: Row and column

Coordinate_Representation:

Abscissa_Resolution: 1000

Ordinate_Resolution: 1000

Planar_Distance_Units: meters

Entity_and_Attribute_Information:

Detailed_Description:

Entity_Type:

Entity_Type_Label:

Forest area density index or forest connectivity index

grid cell

Entity_Type_Definition:

Any of the data elements in a map of either the forest

area density index, or the forest connectivity index

Entity_Type_Definition_Source: U.S. Forest Service

Attribute:

Attribute_Label:

Forest area density index or forest connectivity index

grid cell

Attribute_Definition:

The value is a byte representation of a continuous [0,1]

index value in the area surrounding the grid cell. Larger

values indicate higher index values for either forest area

density or forest connectivity. Zero (0) values indicate

missing data. A byte value of one (1) indicates an original

index value of zero (0.0). A byte value of 255 indicates an

original index value of one (1.0). Original index values

from [0,1] were linearly stretched to the interval [1,255].

Attribute_Definition_Source: U.S. Forest Service

Attribute_Domain_Values:

Range_Domain:

Range_Domain_Minimum: 0

Range_Domain_Maximum: 255

Detailed_Description:

Entity_Type:

Entity_Type_Label: Forest fragmentation index grid cell

Entity_Type_Definition:

Any of the data elements in a map of the forest fragmentation

index.

Entity_Type_Definition_Source: U.S. Forest Service

Attribute:

Attribute_Label: Forest fragmentation grid cell value

Attribute_Definition:

The value is a coarse-scale measure of forest fragmentation in the area

surrounding the grid cell. Fragmentation is described by six classes.

Attribute_Definition_Source: U.S. Forest Service

Attribute_Domain_Values:

Enumerated_Domain:

Enumerated_Domain_Value: 0

Enumerated_Domain_Value_Definition:

Water (on the continental maps only) or unlabeled land area.

Enumerated_Domain_Value_Definition_Source: U.S. Forest Service

Attribute_Domain_Values:

Enumerated_Domain:

Enumerated_Domain_Value: 1

Enumerated_Domain_Value_Definition:

Edge - most of the cells in the surrounding area are forested

and this cell appears to be part of the outside edge of a forest

patch.

Enumerated_Domain_Value_Definition_Source: U.S. Forest Service

Attribute_Domain_Values:

Enumerated_Domain:

Enumerated_Domain_Value: 2

Enumerated_Domain_Value_Definition:

Undetermined - most of the cells in the surrounding area are

forested but this cell could not be classified as to the type of

fragmentation in the surrounding area.

Enumerated_Domain_Value_Definition_Source: U.S. Forest Service

Attribute_Domain_Values:

Enumerated_Domain:

Enumerated_Domain_Value: 3

Enumerated_Domain_Value_Definition:

Perforated - most of the cells in the surrounding area are forested

and this cell appears to be part of an inside edge of a forest

patch. In other words, this cell is near a non-forest inclusion

within a forest patch.

Enumerated_Domain_Value_Definition_Source: U.S. Forest Service

Attribute_Domain_Values:

Enumerated_Domain:

Enumerated_Domain_Value: 4

Enumerated_Domain_Value_Definition:

Interior - all of the cells in the surrounding area are labeled

as forest in the land-cover map.

Enumerated_Domain_Value_Definition_Source: U.S. Forest Service

Attribute_Domain_Values:

Enumerated_Domain:

Enumerated_Domain_Value: 5

Enumerated_Domain_Value_Definition:

Patch - most of the cells in the surrounding area are not forested

and this cell is part of a forest inclusion or patch of forest on

a non-forest background.

Enumerated_Domain_Value_Definition_Source: U.S. Forest Service

Attribute_Domain_Values:

Enumerated_Domain:

Enumerated_Domain_Value: 6

Enumerated_Domain_Value_Definition:

Transitional - about half of the cells in the surrounding area are

forested and this cell may appear to be part of a patch, edge, or

perforation depending on the local forest pattern.

Enumerated_Domain_Value_Definition_Source: U.S. Forest Service

Attribute_Domain_Values:

Enumerated_Domain:

Enumerated_Domain_Value: 17

Enumerated_Domain_Value_Definition: Water on the global map

Enumerated_Domain_Value_Definition_Source: U.S. Forest Service

Attribute_Domain_Values:

Enumerated_Domain:

Enumerated_Domain_Value: 18

Enumerated_Domain_Value_Definition: Background on the global map

Enumerated_Domain_Value_Definition_Source: U.S. Forest Service

Distribution_Information:

Distributor:

Contact_Information:

Contact_Organization_Primary:

Contact_Organization:

Southern Research Station

U.S. Forest Service

Contact_Address:

Address_Type: mailing address

Address:

3041 Cornwallis Road

City: Research Triangle Park

State_or_Province: NC

Postal_Code: 27709

Contact_Voice_Telephone: 919-549-4015

Contact_Instructions:

In addition to the address above, the Southern Research Station,

Research Triangle Park Laboratory telephone number is 919-549-4000.

Distribution_Liability:

Although these data have been processed successfully on a

computer system at the U.S. Forest Service, no warranty

expressed or implied is made by the U.S. Forest Service

regarding the utility of the data on any other system, nor

shall the act of distribution constitute any such warranty.

Standard_Order_Process:

Digital_Form:

Digital_Transfer_Information:

Format_Name: bsq.gz

Digital_Transfer_Option:

Online_Option:

Computer_Contact_Information:

Network_Address:

Network_Resource_Name:

Please use the links provided in Appendix 3 of the online

manuscript (Riitters et al. 2000).

Metadata_Reference_Information:

Metadata_Date: 20000924

Metadata_Contact:

Contact_Information:

Contact_Person_Primary:

Contact_Person: Kurt Riitters

Contact_Organization: U.S. Forest Service

Contact_Address:

Address_Type: mailing address

Address: 3041 Cornwallis Road

City: Research Triangle Park

State_or_Province: NC

Postal_Code: 27709

Contact_Voice_Telephone: 919-549-4015

Contact_Electronic_Mail_Address: kriitters@fs.fed.us

Metadata_Standard_Name: FGDC Content Standards for Digital

Geospatial Metadata

Metadata_Standard_Version: FGDC-STD-001-1998

Metadata_Security_Information:

Metadata_Security_Classification_System: None

Metadata_Security_Classification: Unclassified

Metadata_Security_Handling_Description: None