1997 LAND USE/LAND COVER DATA
Notes on the 1997 Land Use/Land Cover Analysis
Based on notes from Michelle Hutchinson, EarthData International
EarthData International performed the 1997 Land Use/Land Cover (LU/LC) analysis. The contractor compared 1992 LU/LC data files, developed from 1992 aerial photography, against the 1997 aerial photography. The contractor updated the 1992 data files using interpretation of the 1997 photography. The data files use the Anderson Et Al Land Use Classification Scheme. The contractor reports several anomalies and data shifts that should be taken into consideration in the use of this data.
Tidal Differences
The 1997 photos appear to have been flown at a much higher tide than
the 1992 photography. Therefore, much of the Estuarine Emergent,
which is classified 600, is underwater. Because the project was
designed to not show changes that are caused only by tidal
differences, there are areas along the coast which are classified
600, which are now underwater, but are not classified as open water
(540). The emergent (600) is still there, but is not visible because
it was underwater at he time of the photography. To ensure
consistently accurate change measurements, over time, the state would
have to be able to ensure that aerial photography is always taken
under the exact same tidal conditions. This, of course, is not always possible.
In some areas, where it was obvious that the current flow had changed land cover, and that it wasn't only a tidal difference, changes are shown.
Corrections to 1992 Data
There were areas of Farmsteads and Farm Related Buildings (240) and
some Mixed Forest (430) that were not delineated when the original
1992 photos were interpreted. Those areas were added to the 1997 data
for consistency.
In some cases, the contractor has corrected mistakes in polygon labeling from he 1992 data. Areas around tree lines that were poorly delineated, which could have been caused by delineating the shadows seen in the 1992 CIR photography, were corrected to fit the correct treelines as viewed in the 1997 BW photography.
Wetland Changes
There are many areas of the state that were Palustrine Forested (600)
in the 1992 data, which were probably wetland pine, that had been cut
down prior to the 1997 photography. Upland trees surrounding these
areas, which were also cut down, have been classified as clear-cut
(440). The Palustrine areas, however, have been left as 600 and not
changed to 440. According to the Cowardin Classification System and
National Wetlands Inventory conventions, such areas are still
considered to be wetlands, even without their former cover of wetland
pine trees. The shrubs and/or emergent growth that is left, as well
as the hydric nature of the soil, argues for the continued
classification of these areas as wetlands. When the trees are allowed
to grow back, they will still be wetland trees.
In New Castle County, there are some Wetland Emergent (600) polygons that existed in 1992, but were barely visible due to the conditions of the 1992 photography, and therefore were not delineated. The contractor delineated these areas as emergent wetlands in the 1997 data after reviewing the soil survey and comparing the 1992 and 1997 photography.
Image Shifts
In some cases, the quarter quad images were not properly aligned
between 1992 and 1997. The contractor had to shift the 1992 data to
match the 1997 photography for analysis, then return the data to its
original position to avoid creating sliver polygons. The quads
affected were Hickman NE, SE, SW, and NW; Seaford West NE, SE, SW,
and NW; and Bethany Beach SE.
