Geoscience Reference
In-Depth Information
Table 8.2
Percent (%) TIA and surface area distribution for 10 Lexington-Fayette County, KY
14-digit subwatersheds. Please see text for abbreviations
14-digit HUC
Subwatershed
%TIA (ALB)
Area (ha)
05100205-270-040
Steeles Run
1.0
11,806
05100205-270-030
South Elkhorn Creek - I
2.1
4,553
05100205-270-080
Town Branch - I
5.8
2,818
05100205-280-200
Cane Run
10.0
1,822
05100205-120-010
East Hickman Creek
10.8
2,236
05100205-270-010
South Elkhorn Creek - II
12.9
2,637
05100205-280-010
North Elkhorn Creek
17.7
3,492
05100205-120-040
West Hickman Creek
22.3
6,517
05100205-270-070
Wolf Run
28.2
5,792
05100205-270-060
Town Branch - II
44.5
5,458
TOTAL
14.0
47,132
hydrologic unit code (HUC) subwatersheds corresponding to urban/suburban/rural
areas falling primarily within the corporate city boundaries of Lexington-Fayette
County, Kentucky were selected (Table
8.2
;Fig.
8.1
) (Kentucky Division of
Geographic Information
2009d
,
e
).
The post-reprojection results using the NN and BI methods were examined in
three different ways in comparison to original USGS published data (ALB): (a)
spatial distribution of imperviousness value; (b) overall frequency distributions by
imperviousness numerical class; and (c) overall effects on %TIA.
As compared with the ALB data, contrast is lost by using BI due to smoothing
effects, and even though general “crispness” is preserved, some geometric dis-
tortion of linear imperviousness features are introduced by using NN (Fig.
8.2
).
The BI method produced frequency distortions in the high pixel count classes,
whereas the NN method preserved the relationship with the original pixel distribu-
tions (Fig.
8.3
). The histogram shows that discrepancies in pixel value frequencies
were of minor consequence overall (Fig.
8.4
). As a result of this, and also pos-
sibly due to bias cancellation contributed to this agreement, when percent TIA
was calculated for each of the ten 14-digit subwatersheds, both NN and BI pro-
duced similar results, in close agreement with the estimates from the original dataset
(Fig.
8.5
).
Therefore, in order to achieve the objective of making the existing USGS
Seamless Data for imperviousness more readily integrated into existing and planned
geospatial datasets while maintaining accuracy for environmental resource poli-
cymakers at the local and state levels as well as researchers, the decision was
made to reproject the original USGS Albers Equal Area to Kentucky Single Zone
using the NN method. The previous analysis shows how important this sim-
ple choice in software options/algorithms should not be taken lightly because of
the implications of making a different choice would potentially have on estimat-
ing imperviousness in particular and land planning and management activities in
general.
