|
|
 |
FRAGSTATS
Spatial
Pattern Analysis Program for Categorical Maps
|
|
Welcome
To FRAGSTATS!
|
| FRAGSTATS is a computer
software program designed to compute a wide variety of landscape metrics
for categorical map patterns.The original software (version 2) was released
in the public domain during 1995 in association with the publication
of a USDA Forest Service General Technical Report (McGarigal
and Marks 1995). |
 |
 |
|
|
Since then, hundreds of professionals have
enjoyed the use of FRAGSTATS. Due to its growing popularity, we have completely
revamped the program. The purpose of this web site is to facilitate dissemination
of the new software (version 3) and to facilitate communication among FRAGSTATS
users.
|
|
About
the Authors
|
|
The original version of FRAGSTATS, published
in 1995, was developed by Dr.
Kevin McGarigal and Barbara Marks of Oregon State University.
Ms. Marks was the programmer and primary technical support person for the
original release. She now works for Hewlett Packard and no longer provides
technical support.
The current version (3) was developed by
Dr. Kevin McGarigal with programming by Eduard
Ene and additional programming assistance by Chris Holmes.
Chris Holmes was responsible for the initial reprogramming. He now works
for Compaq Computers. Eduard Ene, a research assistant at the University
of Massachusetts, is now the principal programmer. Dr. McGarigal continues
to be the principal developer of FRAGSTATS and is the primary contact person
for questions and comments.
|
|
What's
in This Website
|
|
This web site is logically organized into the following pages:
- FRAGSTATS Documentation
-- This page contains links to all FRAGSTATS documentation, including
background material, user guidelines, and a description of the metrics.
- Downloads
-- This
page is where you go to download the software and documentation. The original
FRAGSTATS (version 2) and the new release (version 3), as well as all
future patch updates and releases can be downloaded from this page.
- FAQ -- This
page contains answers to frequently asked questions, including questions
and answers extracted from the FRAGSTATS list serve that warrant permanent
posting.
- Links -- This
page contains links to other web sites containing information on related
software.
|
|
FRAGSTATS Users Group
|
|
If you have questions about FRAGSTATS
or are seeking advice on a FRAGSTATS-related problem, then the FRAGSTATS
list serve is the place to go. This list serve is designed to facilitate
communication, problem-solving, and idea-sharing among FRAGSTATS users
on issues dealing with landscape metrics and the analysis of landscape
pattern. In addition, all patches and releases of new versions will be
announced via this list serve. To join this list serve, simply send an
email to the following address requesting that you be added to the list:
fragstats-request@forwild.umass.edu
|
|
|
|
The
major difference between version 2 and 3 is the addition of a graphical
user interface to facilitate the parameterization of FRAGSTATS, and the
addition of several new landscape metrics and analysis capabilities. Like
version 2, version 3 accepts input images in a variety of formats (e.g.,
Arc Grid, Ascii, ERDAS, EDRISI, Binary); however, version 3 is limited
to raster formats only. Because of this difference, version 2 still has
utility for users that require or prefer to use vector coverages in Arc/Info.
All other users, however, should find version 3 preferable and considerably
more powerful. Here are the key distinguishing features of each version:
Version 2:
Two separate versions of FRAGSTATS
2 exist; one that accepts Arc/Info polygon coverages (vector) and one that
accepts raster images in a variety of formats. The vector version is an
Arc/Info AML developed on a SUN workstation using Arc/Info version 6.1;
it will not run with earlier versions of Arc/Info. Because of limitations
in Arc/Info, the AML calls several C programs that were developed in a
Unix environment and compiled with the GNU C compiler (note, they may not
compile with other compilers). The raster version is a stand-alone C program
that accepts ASCII image files, 8 or 16 bit BINARY image files, Arc/Info
SVF files (which Arc/Info no longer supports), ERDAS image files, and IDRISI
image files. The raster version also was developed on a SUN workstation
in the Unix operating environment. It is written in C and also compiled
with the GNU C compiler. The raster version also has been compiled to run
in the DOS environment on a personal computer (PC). The directions for
running the DOS version on a PC are exactly the same as the Unix version.
Both vector and raster versions respond to command line input or allow
the user to answer a series of prompts. Both versions generate the same
array of metrics, including a variety of area metrics, patch density, size
and variability metrics, edge metrics, shape metrics, core area metrics,
diversity metrics, and contagion and interspersion metrics, although a
few additional metrics (i.e., nearest-neighbor metrics and contagion) are
computed in the raster version. The format of the output files is exactly
the same.
Version 3:
Version 3 is a stand-alone program
written in Microsoft Visual C++ for use in the Windows Operating environment.
Like version 2, it accepts raster images in a variety of formats, including
Arc Grid, ASCII, 16 or 32 bit BINARY, ERDAS, and IDRISI image files. Some
of the key features that distinguish version 3 from version 2 include:
- Graphical User Interface.--FRAGSTATS analyses are
parameterized via a graphical user interface which allows the user to
tailor each run to their specific needs.
- New Landscape Metrics.--Several new metrics have been added.
-
Nearest Neighbor Distance.--Calculation of nearest neighbor distance
are slightly different in version 3 because instead of cell edge-to-edge
distance (as in version 2), we now use cell center-to-center distance.
We believe that this makes more sense given the way in which raster grids
represent real-world edges (i.e., the true edge is as often as not in
the center of the grid cell) and overcomes some computational difficulties
with indices that use nearest neighbor distance in the denominator (hence
eliminating the possibility of having a zero in the denominator).
- Core
Area.--Core area metrics now allow for user-specified unique edge
depths for each unique edge type, instead of requiring a constant edge
depth for all edges.
-
Moving Window Analysis.--A moving window analysis option is provided
in which a moving window of specified size and shape is passed over the
landscape and returns a value for each selected metric at the class and
landscape levels to the focal cell. The result is a new grid for each
metric in which the cell values represent the computed metric values.
This is a powerful new tool, but comes at the cost of huge computer processing
demands.
- Background & Border.--Rules
regarding the specification and handling of background and border have
been modified slightly to achieve greater consistancy and logic.
- Command Line Execution.--FRAGSTATS
can be executed from a command line to allow for execution from within
another program (as an external execution)
|
|
Disclaimer
|
|
This software is in the public
domain, and the recipient may not assert any proprietary rights thereto
nor represent it to anyone as other than an Oregon State University-produced
program (version 2.0) or University of Massachusetts-produced program (version
3.0). FRAGSTATS is provided "as is" without warranty of any kind,
including, but not limited to, the implied warranties of merchantability
and fitness for a particular purpose. The user assumes all responsibility
for the accuracy and suitability of this program for a specific application.
In no event will the authors or the University be liable for any damages,
including lost profits, lost savings, or other incidental or consequential
damages arising from the use of or the inability to use this program.
As the authors of FRAGSTATS, we
are VERY concerned about the potential for misuse of this program. Like
most tools, FRAGSTATS is only as "good" as the user. FRAGSTATS
crunches out a lot of numbers about the input landscape. These numbers
can easily become "golden" in the hands of uninformed users.
Unfortunately, the "garbage in- garbage out" axiom applies here.
We have done our best in the documentation to stress the importance of
defining the landscape at a scale and in a manner that is relevant and
meaningful to the phenomenon under consideration. Moreover, we have stressed
the importance of understanding the exact meaning of each metric before
it is used. These and other important considerations in any landscape pattern
analysis are discussed in the "background material" contained
on this web site. We strongly urge you to read the entire documentation
before ever running FRAGSTATS.
We wish to remind users that we
are not in the commercial software marketing business. We are scientists
who recognized the need for a tool like FRAGSTATS to assist us in our research
on landscape ecological issues. Therefore, we do not wish to spend a great
deal of time consulting on trivial matters concerning the use of FRAGSTATS.
However, we do recognize an obligation to provide some level of information
support. Of course, we welcome and encourage your criticisms and suggestions
about the program at all times. We will welcome questions about how to
run FRAGSTATS or interpret the output only after the user has read the
entire documentation. This is only fair and will eliminate many trivial
questions. In this regard, we encourage you to join the FRAGSTATS list
serve and seek assistance from the vast community of users. They are the
real experts and the probable source of most answers. Finally, we are always
interested in learning about how others have applied FRAGSTATS in ecological
investigations and management applications. Therefore, we encourage you
to contact us and describe your application after using FRAGSTATS.
We
hope that FRAGSTATS is of great assistance in your work and we look forward
to hearing about your applications.
|
|
Acknowledgements
|
| Many
individuals provided valuable feedback during the development and many revisions
of the original software (version 2), including Steve Garman, Eric Gustafson,
Jeff Nighbert, Tom Moore, Catherine Rogers, and David Wallin. We are especially
grateful to Catherine Rogers and Eric Gustafson for their comprehensive
and detailed testing of the original program and their many useful suggestions.
Initial funding for development of the original software was provided through
the Coastal Oregon Productivity Enhancement (COPE) program; COPE is a cooperative
research and technology transfer effort among Oregon State University, USDA
Forest Service, USDI Bureau of Land Management, other state and federal
agencies, forest industry, county governments, and resource protection organizations.
Subsequent funding for the completion of the FRAGSTATS 2 documentation (McGarigal
and Marks 1995) was provided by the USDI Bureau of Land Management Cooperative
Research Unit and the USDA Forest Service, Pacific Northwest Research Station,
Corvallis, Oregon. Funding for the development of version 3 was provided
by the USDA Forest Service, Rocky Mountain Region, Denver, Colorado and
the University of Massachusetts, Amherst. We are especially grateful to
Claudia Regan for her support of the project.
|
|
How
to Cite FRAGSTATS
|
|
Unfortunately, we can't provide a proper journal or book citation for
the current version of FRAGSTATS. Currently, this website is the sole means
of "publication" of the software. We have a FRAGSTATS book in
the making, but it will be some time before it is finished. In the meantime,
I suggest the following citation:
McGarigal, K., S. A. Cushman, M. C. Neel, and E. Ene. 2002. FRAGSTATS:
Spatial Pattern Analysis Program for Categorical Maps. Computer software
program produced by the authors at the University of Massachusetts, Amherst.
Available at the following web site: www.umass.edu/landeco/research/fragstats/fragstats.html
|
|
 |