Operational Shoreline Mapping with
High Spatial Resolution Radar
and Geographic Processing
Amina Rangoonwala, Cathleen E. Jones, Zhaohui Chi, and Elijah Ramsey III
Abstract
A comprehensive mapping technology was developed utiliz-
ing standard image processing and available
GIS
procedures
to automate shoreline identification and mapping from 2
m synthetic aperture radar (
SAR
)
HH
amplitude data. The
development used four
NASA
Uninhabited Aerial Vehicle SAR
(
UAVSAR
) data collections between summer 2009 and 2012
and a fall 2012 collection of wetlands dominantly fronted by
vegetated shorelines along the Mississippi River Delta that are
beset by severe storms, toxic releases, and relative sea-level-
rise. In comparison to shorelines interpreted from 0.3 m and 1
m orthophotography, the automated
GIS
10 m alongshore sam-
pling found
SAR
shoreline mapping accuracy to be ±2 m, well
within the lower range of reported shoreline mapping accura-
cies. The high comparability was obtained even though water
levels differed between the
SAR
and photography image pairs
and included all shorelines regardless of complexity. The
SAR
mapping technology is highly repeatable and extendable to
other
SAR
instruments with similar operational functionality.
Introduction
As is the case for most coastal deltas (Tessler
et al
., 2015), the
Mississippi River Delta (
MRD
) is losing wetland at a rapid rate,
an occurrence that reflects a general trend throughout coastal
Louisiana and arises from a loss of sediment influx, subsid-
ence-driven relative sea-level rise, and wave erosion (Wilson
and Allison, 2008). Lateral retreat of shorelines is estimated to
account for 25 percent of wetland losses in Louisiana (Wilson
and Allison, 2008) and can reach upwards of 4 m annually
on the Mississippi River Delta (
MRD
) barriers and headlands
(Flocks, 2006). Likewise, marshes occupying the interior
embayment of the
MRD
exhibit high rates of shoreline retreat
mostly attributable to wave erosion, particularly along shore-
lines exposed to longer fetch and predominant directions of
wave approach (Wilson and Allison, 2008). Shoreline stability
is known to directly influence the adjacent interior marsh
status (Knutson, 1988).
In 2010,
USGS
and
NASA
implemented a collaborative
multi-year study that relies in part on Uninhabited Aerial
Vehicle Synthetic Aperture Radar (
UAVSAR
) Polarimetric
SAR
(
PolSAR
) data to monitor changes in marshes in the north-cen-
tral Gulf of Mexico (
GOM
) (Ramsey
et al
., 2011). Here, we de-
scribe a component of that study that focused on developing
a shoreline mapping method using high resolution synthetic
aperture radar (
SAR
) that can track shoreline change of
MRD
wetlands during a time period from 2009 to 2012 that covers
both the 2010 Deepwater Horizon (
DWH
) oil spill and 2012
Hurricane Isaac (Rangoonwala
et al
., 2016).
A shoreline is defined as the boundary line between land
and usually a large body of water at one instant in time (Liu
et al
., 2011; Lipakis
et al
., 2008; Kotsovinos and Georgoulas,
2008) is known as the land-water interface (
LWI
) shoreline. Ae-
rial photographic
LWI
shoreline mapping begun in the 1960s
remains one of the most often used data sources for shoreline
position mapping (Lipakis
et al
., 2008; Robertson
et al
., 2004).
Starting in the 1980s and extending into the 1990s, optical
satellites produced 10 m spatial resolution products that were
adequate for monitoring rapid shoreline change but generally
insufficient for tracking shoreline migration (La Monica
et al
.,
2008). With 1 m to 4 m spatial resolutions availability from
Ikonos, QuickBird, WorldView, and GeoEye remote sensing
systems, satellite optical shoreline mapping became more
appropriate for assessing shoreline morphological trends (Di
et al
., 2003; La Monica
et al
., 2008; Liu
et al
., 2011).
Synthetic Aperture Radar is more flexible than passive op-
tical systems by increasing operability to both day and night
and in most weather conditions. This makes it particularly
well suited for shoreline mapping in coastal regions with
persistent cloud cover and for supporting emergency response
activities (e.g., Lipakis
et al
., 2008; Di
et al
., 2003; Liu
et al
.,
2011; Ouyang
et al
., 2010; Robinson, 2011; Wang and Allen,
2008). As for optical mapping, current
SAR
systems such as
the Advanced Land Observation Satellite (ALOS), Radarsat-2,
and TerraSAR-X with spatial resolutions up to 3 m are ade-
quate for shoreline mapping (Robinson, 2011).
LWI
mapping whether automated, semi-automated, or man-
ual (e.g., photointerpretation) relies primarily on a discernible
contrast between the water and land (Ouyang
et al
., 2010; Wang
and Allen, 2008; Landelli and Pranzini, 2008; La Monica
et al
.,
2008; Kotsovinos and Georgoulas, 2008; Lipakis
et al
., 2008;
Flocks, 2006; Liu
et al
., 2011). In order to be fully consistent
in mapping shoreline change, the
LWI
must be transformed to
a tide-coordinated shoreline (
TCS
) (Robertson
et al
., 2004). The
most direct method to transform a
LWI
shoreline to a
TCS
is to
collect the remote sensing image coincident to a tidal datum
surface, such as the Mean High Water Line (Lipakis
et al
., 2008).
Once that time-coordinated collection is accomplished, the ac-
curacy of the
TCS
location then depends on how well spatially
and spectrally the remote sensing system can define the
LWI
.
Offering similar day and night operability as
SAR
, Light De-
tection and Ranging (lidar) does not require time-coordinated
Amina Rangoonwala and Elijah Ramsey III are with the U.S.
Geological Survey, Wetland and Aquatic Research Center, 700
Cajundome Blvd., Lafayette, LA 70506 (
).
Cathleen E. Jones is with the Jet Propulsion Laboratory, Cali-
fornia Institute of Technology, 4800 Oak Grove Dr., Pasadena,
CA 91109.
Zhaohui Chi is with the University of Louisiana-Lafayette
CESU, 635 Cajundome Blvd, Lafayette, LA 70506.
Photogrammetric Engineering & Remote Sensing
Vol. 83, No. 3, March 2017, pp. 237–246.
0099-1112/17/237–246
© 2017 American Society for Photogrammetry
and Remote Sensing
doi: 10.14358/PERS.83.3.237
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
March 2017
237
