Relative Wilderness of the Eastern United States

Relative Wildness of the eastern United States. A combination of
data representing the six attributes of the land that contribute to
its wildness. These attributes are solitude, remoteness, uncontrolled
process, natural composition, unaltered structure, and pollution. For
a full treatment of the wildness concept see "Indicators of Wildness:
Using Attributes of the Land to Assess the Context of Wilderness" in
McCool, Stephen F.; Cole, David N.; Borrie, William T.; O'Loughlin,
Jennifer, comps. 2000. Wilderness science in a time of change
conference--Volume 2; Wilderness in the context of larger ecosystems;
1999 May 23-27; Missoula MT. Proceeding RMRS-P-15-VOL-2. Ogden., UT:
U.S. Department of Agriculture, Forest Service, Rocky Mountain
Research Station.

This data for the eastern United States is produced as part of a TWS
effort to define the concept of wildness in a manner that allows its
mapping. This effort is undertaken in support of our strategic goal of
working toward the creation of a national network of wildlands. The
wildness concept offers a new way in which to look at landscapes. It
demonstrates that wildness is a continuum, that it is relative to the
landscape being examined, and that it is present in all
landscapes--large and small, urban and undeveloped. The data not only
allows for the creation of hard copy and digital maps of relative
wildness in a landscape, it can be used as input to other processes in
which wildness may have meaning.

The wildness concept, as developed by The Wilderness Society, embodies
subtleties it is important to understand so that wildness mapping data
is not inappropriately used or interpreted. While review of the
conceptual paper referenced elsewhere in this metadata provides
information necessary for proper understanding, that paper does not
address some of the subtleties discovered during our development of
the mapping process subsequent to that paper's writing. Following is
an attempt to highlight some of the more important of these
subtleties.

Our wildness index is a unitless relative rank--ordinal data in
technical language. It does not show wildness in any absolute
sense. There is no zero wildness, and a rank of 30 is not twice as
wild as a rank of 15. From this fact follow several others.

First, one cannot compare amounts of different wildness values in a
landscape. A comparison of the number of grid cells with, say,
wildness value 30 to the number of grid cells with the value 10 is
inappropriate. The creation of the wildness attribute data
(e.g. remoteness, solitude, etc.) usually involves assigning a range
of source data values to five attribute ranks on an equal-area
basis. This is done because there is often no more meaningful way in
which to specify class breakpoints. Though combining the six attribute
data sets often gives a highly variable number of cells in the
different wildness classes, this only clouds the fact that the
underlying data values are assigned so as to uniformly distribute
them.

Second, the wildness rank of any piece of land is dependent on the
context in which that piece of land is considered. In the context of
the contiguous United States, much of the Appalachian Mountains are
only moderately wild: the vast open spaces of the West being far
wilder by comparison. In the same context even the largest of city
parks receives the lowest wildness ranking. In the context of the
eastern United States, however, the Appalachians are some of the
wildest lands, and in the context of a metropolitan area a city park
would also receive a high wildness ranking. Thus, no single wildness
data set is suitable to all locals and scales. Indeed, a primary value
of our wildness mapping is that it allows us to view the land at
different scales and to notice the way in which wildness patterns
shift. Such multi-scale analysis adds a richness to our understanding
of the landscape that a single-scale view does not.

Third, when viewing wildness data developed for a single study area,
looking for lands with high wildness index values does not give a full
picture of the important wildlands in the area. Equally worth
consideration are what one might call "wildness peaks"--areas of
locally high wildness rank. These areas of relatively higher wildness
than their surroundings can be important elements of a network of
wildlands. Consideration of such peaks also extends the utility of
wildness data developed at a single scale to analyses of smaller
areas. It is these peaks that would most likely have the highest
wildness index values if a complete wildness mapping project were
undertaken for that smaller area.

[Summary provided by The Wilderness Society]