AL
patial
data
ze and
tretch
]uired
2000),
; were
loped.
isplay
splay,
NASA
on the
y and
ogical
(OAA
1.5m
n and
f net
1996-
were
'aches
hore,
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004
/
Our study area (Assateague Island) is located within the
Assateague Island National Seashore in Virginia (Fig. 1)
between 37.883747? S to 38.020204? S latitude, and
75.389548° W to 75.220216? W longitude. The 37-mile-long
Seashore lies along the central Delaware Peninsula, stretching
along the Atlantic coast and subject to severe gales and waves.
Assateague Island is exceptionally dynamic, experiencing
average erosion rates as high as 10 feet per year in some areas
(http://soundwaves.usge.gov/2002/1 1/research.html). Important
features of Assateague are its fragile coastal elements,
characterized by sand dunes, maritime forests, inlets, lagoons,
back-barrier marshes and vegetation. The island is one in a
chain of barrier islands along the U.S. Atlantic seaboard that are
built as wave action piles up sand from the ocean floor (Allen et
al. 2000), so its study is useful for us to more fully understand
the dynamics of coastline change in the mid-Atlantic. Like
other barrier islands, Assateague is constantly changing shape
and geographical position (Dolan et al. 1997, 1992).
3. DATA ACQUISITION
We downloaded LIDAR data from the NOAA Coastal Services
Center | (http://www.csc.noaa.goc/crs/tem/index.htm) for our
study. The data sets acquired on October 11, 1996, September
16-18, 1997, February and December 1998, as well as
September and November 2000 cover the entire study data,
while the date set acquired on Oct. 11, 1996, Sept. 16-18, 1997
only covered the south end of Assateague Island, and the data
acquired on April 3, 1998 only covered the eastern shoreline.
Because of coastal conditions and environment as well as the
LIDAR data volume, the study area has been divided into six
sections.
The downloaded LIDAR data was resampled into grid DEMs
using ArcView inverse distance weighting (IDW) methods with
a planimetric (cell) resolution of 1.5 by 1.5 m. All the DEMs
were geo-referenced to the WGS84 spheroid and North
American Vertical Datum (NA VD) of 1988, respectively.
4. ANALYSIS OF TOPOGRAPHIC AND
MORPHOLOGIC CHANGES
4.1 Methods
To most effectively analyze the spatial patterns of topographic
and morphological change (erosion, deposition, or no change)
along the coastline, we partitioned the shoreline into six
sections. In each section, three study sites (also referred to as
Areas Of Interest (AOIs)) were created. Ancillary data, such as
the spatial surface profiles of the DEMs, slope and relief data of
the DEMs, panchromatic images, and USGS color infrared
(CIR) digital orthophoto quads (DOQs), are used to assist in the
identification and creation of each AOI. Heavily vegetated
areas, man-made structures such as houses and piers, and wave
activity were excluded because these factors would impact the
reliability of the change analysis (White et al. 2003). Finally,
three representative AOIs in each section were selected for
topographical and morphological change analysis using the
successive DEM data pairs in the periods of 1996-2000. The
selected AOIs represent a particular segment of coastline,
where the dune line and dry beach are obviously distinguished,
and where the processes of erosion and deposition may be
easily studied spatially. Each AOI was chosen so as to cover
almost exactly the same location and portion of coastline for
each yearly analysis. Because the data was not perfectly
consistent, dune transects and profiles were created to assist in
comparing the accuracy of the DEMs between each yearly
survey.
The basic method for topographic change identification using
DEMS is differencing the Z coordinates of the second year to
the first year values on each grid cell for each DEM pair. The
volume change at each cell location can then be computed. A
positive, negative, or zero volumetric value (m?) at a cell
represents the amount of deposition, erosion, or no change. The
morphological change of topography over the entire study area
can be obtained through summing the positive and negative
volumetric values (m?) in each cell. The volumes of deposition
for each Section can be calculated by summing the all positive
volumetric values of cells. Similarly, the volumes of erosion for
each Section can be calculated by summing the all negative
values of the cells. The net change is calculated by differing the
total deposition to the total erosion. Considering that each
Section does not cover exactly the same size of beach area in
each yearly LIDAR data, the net volumetric change per meter
square (m'/m?) is adopted for comparing the volumetric
changes at various time intervals. Using the proposed methods
above, the topographic change between selected years, and the
total volumetric change of the beach and sand dunes of the 6
sections are observed.
Elevation Range (m)
— 6.00-7.00
5.18 - 6.00
E 4.27-5.18
3.35 - 4.27
2.74 - 3.35
1.83 - 2.74
0.30 - 1.83
-0.03 - 0.30
-1.00 - -0.03
120 m
Figure 2. Section 6 segment of coastline consisting of the
primary portion of the dune line and dry beach.
4.2 Spatial Pattern of Topographical Change
The topographic differences of the study area between 1996
and 2000 are visualized via Triangulated Irregular Network
(TIN) data structure. We found that the widths of the dune,
berm, foreshore, and near shore for each section vary. For
example, the width of dune in Section 5 is wider than one in
Section 6. The berm in Section 4 narrows from north to south
and finally disappearing in Section 3 as the dune transitions
directly into the foreshore, forming a ridge. The ridge of
Section 2 is narrower than that of the other Sections, and its
height is lower ne of the other Section. The topographic
elevation in the southern Assateague Island (Section 1) has
changed greatly between 1996 and 2000. This change is
irregular over the entire study area.
Analyzing the DEM data pair between 1996 and 2000, we
found the shoreline topographic change varies largely from
south (Section 1) to north (Section 6) of the study area (see Fig.
