Geoscience Reference
In-Depth Information
The areas of developed land effectively added between 2001 and 2005 totaled
8,160 ha. A total of 6,178 regions were delineated using the same region-group
algorithm as as described for the 1992-2001 data, grouping pixels into regions of
continuous developed class increase (i.e. classes 21, 22, 23, and 24). The TIA was
calculated by using individual areas of developed class, based on the contribution
of each class to TIA (e.g. developed, open space). From preliminary visual verifi-
cation and proximity analysis it became obvious that the surfaces effectively added
were occurring in the vicinity of previously existent developed areas. The region
group approach utilized created a very large proportion of regions less than 1 ha,
accounting for 86% of the total count and for 15% of the total area added. In con-
trast, the total surface area for regions over 20 ha in surface area (
n
74) totaled
2,838 ha. While information on diffuse (non-point) change may be lost by grouping
and filtering by size, this type of generalization procedure allows for focusing envi-
ronmental and watershed management efforts on areas of detected - but compact
and extensive change.
Surprisingly, the mean %TIA was similar for all region size classes - 30-40%,
thus each region's contribution to total increase in TIA is a proxy for its contribution
to the total area, with a maximum of 35% for areas >20 ha - with 1% of total number
of regions, and a minimum of 15% for areas between 0 and 1 ha - representing 86%
of the total number of regions.
Centroids for contiguous developed areas effectively added in the two time
periods, attributed with the total area developed. Using these, kernel densities
were calculated, and mean densities (ha/ha) for each individual subwatershed were
estimated. The kernel density procedure yielded similar range of values for the
1992-2001 and 2001-2005 analysis, with a maximum of 0.06 ha/ha (Fig.
8.11
).
Not surprisingly, the subwatersheds affected by imperviousness increase were
often in and around Kentucky's major metropolitan and peri-urban areas : 1.West
Central KY (
Louisville-LaGrange-Shepherdsville-Shelbyville); 2.
East Central
KY (
Lexington-Nicholasville-Georgetown; Lawrenceburg-Winchester-Richmond);
3.
Northern KY (
Florence-Covington-Independence- Alexandria; Warsaw-Sparta-
Ghent)
; 4. South Central KY (
Somerset
) and 5. Western KY (
Bowling Green-
Owensboro
).
Just in the four years intervening between 2001 and 2005, 7% of the study
subwatersheds experienced added development densities of between 1 and slightly
over 5%. Although these values translated into about 4% of the watersheds having
an increase of %TIA of between 1 and 6%, a somewhat complex trend is revealed
across mean %TIA classes (Fig.
8.12
).
A recurrent or persistent densification of imperviousness in the selected sub-
watersheds was suggested by the relationship between kernel densities in the two
consecutive time periods (Fig.
8.13
).
=
