Change in the Length of the Middle Section of the
Chandeleur Islands Oil Berm, November 17, 2010,
through September 6, 2011
By N.G. Plant and K.K. Guy
Open-File Report 2013–1075
U.S. Department of the Interior
U.S. Geological Survey
ii
U.S. Department of the Interior
Ken Salazar, Secretary
U.S. Geological Survey
Suzette M. Kimball, Acting Director
U.S. Geological Survey, Reston, Virginia 2013
For product and ordering information:
World Wide Web: http://www.usgs.gov/pubprod
Telephone: 1-888-ASK-USGS
For more information on the USGS—the Federal source for science about the Earth,
its natural and living resources, natural hazards, and the environment:
World Wide Web: http://www.usgs.gov
Telephone: 1-888-ASK-USGS
Suggested citation:
Plant, N.G., and Guy, K.K., 2013, Change in the length of the middle section of the Chandeleur Islands oil berm,
November 17, 2010, through September 6, 2011: U.S. Geological Survey Open-File Report 2013-1075, 8 p.
Any use of trade, product, or rm names is for descriptive purposes only and does not imply
endorsement by the U.S. Government.
Although this report is in the public domain, permission must be secured from the individual
copyright owners to reproduce any copyrighted material contained within this report.
iii
Contents
Introduction ..........................................................................................................................................................1
Methods
...............................................................................................................................................................3
Results .................................................................................................................................................................6
Acknowledgments ...............................................................................................................................................6
Figures
1. Chandeleur and Breton Islands (part of the Breton Island National Wildlife Refuge), the Mississippi River
Delta, the site of the Deepwater Horizon oil spill, and the location of the full extent of the Chandeleur
Islands berm
.............................................................................................................................................. 1
2.
The Chandeleur Islands berm divided into northern, middle, and southern sections ................................ 2
3.
The completed middle section of the Chandeleur Islands berm ............................................................... 3
4.
Example of a panchromatic image ............................................................................................................ 4
5.
Example of isolines generated on the basis of pixel value ........................................................................ 4
6.
Example of berm-length measurement ..................................................................................................... 5
7.
Time series of berm-length measurements from each data source. ......................................................... 6
Tables
1. Satellite multispectral and panchromatic image resolutions. ..................................................................... 4
2.
Berm-length measurements. ..................................................................................................................... 7
1
Change in the Length of the Middle Section of the
Chandeleur Islands Oil Berm, November 17, 2010,
through September 6, 2011
By N.G. Plant and K.K. Guy
Introduction
On April 20, 2010, an explosion on the Deepwater Horizon oil rig drilling at the Macondo
Prospect site in the Gulf of Mexico resulted in a marine oil spill that continued to ow through
July 15, 2010. One of the affected areas was the Breton National Wildlife Refuge, which consists
of a chain of low-lying islands, including Breton Island and the Chandeleur Islands, and their
surrounding waters. The island chain is located approximately 115–150 kilometers (km) north-
Figure 1.
Chandeleur and Breton Islands (part of the Breton Island National Wildlife Refuge), the Mississippi
River Delta, the site of the Deepwater Horizon oil spill, and the location of the full extent of the Chandeleur
Islands berm. The background image is U.S. Geological Survey Landsat 5 taken February 18, 2010, prior to
the start of berm construction.
Breton
Island
Chandeleur
Islands
Deepwater
Horizon
Mississippi
River
Delta
Berm
GULF
OF
MEXICO
0 10 20 30 MILES
0 10 20 30 KILOMETERS
Base from U.S. Geological Survey
Landsat 5, Feb 18, 2010
WGS 1984 UTM Zone 16N
89°W
30°N
29°N
ALABAMA
MISSISSIPPI
LOUISIANA
FLORIDA
Map Area
2
Base from U.S. Geological Survey
Landsat 5, Feb 18, 2010
WGS 1984 UTM Zone 16N
Northern
Section
Middle
Section
Southern
Section
As-built
Berm
C h a n d e l e u r I s l a n d s
88°48'W88°50'W
30°0'N
29°58'N
29°56'N
29°54'N
MILES
KILOMETERS
0 1 2
0 1 2
MISSISSIPPI ALABAMA
LOUISIANA
Map Area
Figure 2. The Chandeleur Islands berm divided into northern, middle, and southern sections. The as-
built berm footprint is shown in orange. The background image is U.S. Geological Survey Landsat 5 taken
February 18, 2010, prior to the start of berm construction.
northwest of the spill site (g. 1). A
sand berm was constructed seaward of, and on, the island chain.
Construction began at the northern end of the Chandeleur Islands in June 2010 and ended in April
2011. The berm consisted of three distinct sections based on where the berm was placed relative to
the islands (g. 2). The northern section of the berm was built in open water on a submerged portion
of the Chandeleur Islands platform. The middle section was built approximately 70
–90 meters (m)
seaward of the Chandeleur Islands. The southern section was built on the islands’ beaches. Repeated
Landsat and SPOT satellite imagery and airborne lidar were used to observe the disintegration of
the berm over time. The methods used to analyze the remotely sensed data and the resulting, derived
data for the middle section (g. 3) are reported here.
3
88°49'W88°50'W
29°59'N
29°58'N
29°57'N
C h a n d e l e u r I s l a n d s
Completed
Middle
Section
of Berm
0 0.25 0.5 KILOMETERS
0 0.25 0.5 MILES
Base from U.S. Geological Survey
Landsat 5, December 3, 2010
WGS 1984 UTM Zone 16N
MISSISSIPPI
ALABAMA
LOUISIANA
Map Area
Figure 3.
The completed middle section of the Chandeleur Islands berm. This U.S. Geological Survey
Landsat 5 image was taken December 3, 2010, and shows the middle section of the berm, which lies
approximately 70–90 m offshore of the Chandeleur Islands.
Methods
The Chandeleur Islands berm was built approximately 50-m wide (above mean high water)
and 2-m high relative to the North American Vertical Datum of 1988 (NAVD 88) and using the 1996
Geoid model (Geoid 96). Because of the large size of the berm combined with the highly reec-
tive nature of sand, observations from satellite imagery were possible. Medium resolution (5–30 m)
Landsat and SPOT satellite imagery (table 1) provided relatively frequent observation opportunities.
Additionally, two high resolution lidar elevation datasets were used for measuring berm length. A
total of 29 observations were made from November 17, 2010, the rst date usable satellite imagery
was available after the completion of the middle section of the berm, to September 6, 2011, when
the berm no longer occupied its as-built footprint. For the purpose of these berm-length measure-
ments, only those portions of the berm that occupied its as-built footprint (as estimated from a
sequence of SPOT satellite images obtained during the construction period: September 5, 2010;
October 1, 2010; December 7, 2010; and April 3, 2011) were measured.
4
Table 1.
Satellite multispectral and panchromatic image resolutions.
Satellite Multispectral resolution Panchromatic resolution
SPOT 4 20 m 10 m
SPOT 5 10 m 5 m
Landsat 5 30 m (none)
Landsat 7 30 m 15 m
Satellite images were selected on the basis of availability, clear view of the berm, and reso-
lution (ground sampling interval or cell size). The sensors on Landsat 5, Landsat 7, SPOT 4, and
SPOT 5 differ in their spectral bands and resolution (table 1). When available, panchromatic bands
were used because of their higher resolutions. When not available, single bands from the multispec-
tral images were selected. Band 5 (1.55 to 1.75 micrometers (µm)) at 30-m resolution was used for
Figure 4.
Example of a panchromatic image.
Water, waves, island, beach, berm, and breaches
in the berm are labeled. The dry sand berm and
beaches appear as light grays to white, the water
as darker shades of gray, and the wet sand of the
berm breaches as mid-tone grays.
waves
island
water
beach
berm
berm
breaches
Figure 5. Example of isolines generated on
the basis of pixel value. The isolines are drawn
in pink on top of the U.S. Geological Survey
Landsat 7 panchromatic image from which they
were made. The berm is the wide, light streak
of pixels running from the upper left to the lower
right of the gure. The mid-tone grays in the
upper right are waves.
5
polygon
not
measured
polygon
measured
as-built
footprint
selected
isolines
measure
line
Landsat 5 images. Of the four 20-m resolution multispectral SPOT 4 images used, band 1 (0.50 to
0.59 µm) was used for three dates and band 3 (0.78 to 0.89 µm) was used for one date (table 2).
Water has lower reectivity than sand in the satellite images and, therefore, has a lower pixel
intensity value. In a typical gray-scale representation where low values are dark and high values
are light, water will appear dark and sand will appear very light or white. Wet sand is less reective
than dry sand and appears in mid-tone grays (g. 4). The relatively high pixel values of dry sand
were used to delimit the berm footprint. This method is subject to bias errors caused by dif
ferences
in water levels when different images were acquired, and no corrections for these biases have been
made here. The water levels from a nearby location (Station 8761305, Shell Beach, Louisiana),
referenced to the mean sea level datum, are included in this report (table 2).
Each image was visually examined to determine the footprint of the berm. Isolines based on
pixel values were generated for each image using the Contour tool in ArcGIS® (g. 5). A contour
interval of 5 intensity units was used for Landsat 5 band 5, SPOT 4, and SPOT 5 images (g. 5),
and a contour interval of 2 intensity units was used for Landsat 7 panchromatic images. Because the
pixel-intensity values for water, dry sand, and wet sand were not consistent between images, xed
contour levels were not used to delineate the berm. Instead, the contours were overlaid on the image
and one of these contours was selected to represent the footprint of each berm segment as a polygon
in the geographic information system (GIS). This footprint was then used to measure the length of
the berm segment. Only those portions of the berm footprint that occupied the original as-built foot-
print were used to measure berm length (g. 6). Once sand was moved beyond the as-built footprint
by overwash, inundation, or breaching, it no longer contributed to the measured length of the berm.
The berm footprints obtained from two lidar elevation datasets were based on elevation
rather than reectivity. Contours were generated at 10-centimeter (cm) intervals and were compared
to the berm footprints obtained from satellite imagery. The 100-cm (NAVD 88, Geoid 96) contour
was selected to represent the subaerial portion of the berm. This level is well above the typical
Figure 6.
Example of berm-length measurement.
The red line represents the berm’s as-built footprint.
One yellow line encloses a small area that appears to
be on the berm; however, this area does not fall within
the as-built footprint. Therefore, this area is no longer
considered part of the berm and is not measured. A
second yellow line encloses a larger area and mostly
falls within the as-built footprint. The brown line
represents the resulting berm-length measurement.
6
water level, allowing retrieval of topographic lidar from each survey. Similarly to the treatment of
the satellite imagery, berm-length measurements were estimated where the 100-cm lidar-elevation
contour fell within the as-built footprint. Some clusters of small polygons appeared in the lidar berm
footprints. These clusters were measured as if they were one large polygon.
Results
The results from the satellite and lidar data analysis are presented in gure 7, which shows
a time series of berm lengths derived from each of the sensors. The measurements are listed in
table 2. The accuracy of the berm-length measurements was quantied from the differences between
sequential length measurements, excluding the large length changes observed on January 12, 2011,
via Landsat 7. The root mean square difference was 236 m.
0
500
1000
1500
2000
2500
3000
3500
Nov-2010 Jan-2011 Feb-2011 Apr-2011 June-2011 July-2011 Sept-2011
Berm length (m)
SPOT 4
SPOT 5
Landsat 5
Landsat 7
Lidar
Figure 7.
Time series of berm-length measurements from each data source.
Acknowledgments
The data collection and analysis presented in this report were conducted with support from
the U.S. Geological Survey’s Coastal and Marine Geology Program. The clarity of the report was
improved by reviews from C. Sherwood and J. Flocks. We thank B. Boynton, T. Burress, and
K. Naugle for editing, producing, and publishing the nal document.
7
Table 2. Berm-length measurements.
Date (MDY)
Image Time
(UT) Sensor Spectrum Isoline
Number of Line
Segments
Total Length
(m)
Water Level
(m)
Area of Isoline
Polygons (m
2
) Area of Berm (m
2
)
12/7/2010 16:33 SPOT 4 band 1 140 5 3389 -0.42 138628 126175
1/2/2011 16:32 SPOT 4 band 1 110 6 3026 -0.14 92235 82066
1/13/2011 16:20 SPOT 4 pan 165 5 2367 -0.32 63334 41709
1/23/2011 16:27 SPOT 4 pan 190 6
2484 -0.32 78718 56094
2/13/2011 16:22 SPOT 4 pan 170 10 2542 -0.67 63022 52760
2/24/2011 16:10 SPOT 4 band 1 220 10 2489 -0.22 80451 60213
3/16/2011 16:25 SPOT 4 pan 220 5 2368 -0.09 84400 71383
6/13/2011 16:09 SPOT 4 pan 50 7 1575 0.34 35668 30480
6/18/2011 16:13 SPOT 4 pan 65 5 1620 0.07 29956 27381
7/19/2011 16:15 SPOT 4 pan 175 15 1588 0.24 32739 28136
8/4/2011 16:06 SPOT 4 pan 200
11 1089 0.00 18749 17368
8/9/2011 16:10 SPOT 4 pan 205 9 886 0.06 10708 10476
8/30/2011 16:05 SPOT 4 band 3 185 5 998 -0.01 24985 13283
4/3/2011 16:36 SPOT 5 pan 200 8 2263 -0.11 46342 45864
4/29/2011 16:35 SPOT 5 pan 220 11 1868 0.19 54238 50614
9/6/2011 16:34 SPOT 5 pan --- 0 0 0.56 0 0
11/17/2010 16:15 Landsat 5 band 5 60 3 3390 0.19 142364 117665
12/3/2010 16:16 Landsat 5 band 5 60 4 3340 -0.48 126145 107662
12/19/2010 16:16 Landsat 5 band 5 65 4 3318 -0.19 121676 108867
2/21/2011 16:16 Landsat 5 band 5 75 8 2333 -0.17 70199 47633
3/25/2011 16:16 Landsat 5 band 5 65 5 2066 -0.01 82788 67412
5/28/2011 16:15 Landsat 5 band 5 80 5 1729 0.25 64399 52123
7/31/2011 16:15 Landsat 5 band 5 70 8 1048
0.23 20510 19110
8/16/2011 16:15 Landsat 5 band 5 110 9 1156 0.08 18342 17367
8
Image Time Number of Line Total Length Water Level Area of Isoline
Date (MDY) (UT) Sensor Spectrum Isoline Segments (m) (m) Polygons (m
2
) Area of Berm (m
2
)
1/12/2011 16:19 Landsat 7 pan (band 8) 42 5 2237 -0.39 68479 60817
3/1/2011 16:19 Landsat 7 pan (band 8) 52 5 2668 -0.20 111199 92659
8/24/2011 16:19 Landsat 7 pan (band 8) 78 4 1089 0.35 22476 14944
2/12/2011 --- Lidar elevation cm 100 6 2247 --- 41525 34540
5/31/2011 --- Lidar elevation cm 100 8 1330 --- 12244 10776
Date (MDY) = Date in month/day/year format.
Image Time (UT) = Universal Time that image was acquired in hours and minutes (HH:MM).
Sensor = Image source type.
Spectrum = The satellite image band used or, for lidar, the elevation used, in centimeters.
Isoline = Satellite image pixel value or lidar elevation of contour line used to delineate berm.
Number of Line Segments = Number of line segments in berm measurement.
Total Length (m) = Total length of berm, in meters.
Water Level (m) = Water level Shell Beach, Louisiana, tide station, at the time of image collection, in meters using Mean Sea Level as the datum.
Area of Isoline Polygons (m
2
) = Area of isoline polygons falling, at least in part, within the as-built footprint, in square meters.
Area of Berm (m
2
) = Area of the portion of the isoline polygons that fall within the as-built footprint, in square meters.
Table 2. Berm-length measurements.—Continued