Page 2 - hydro-maritime_2010

airborne and space borne imaging; radar and LIDAR imaging. Furthermore, beach elevation within the intertidal zone is
retrievable as well using ground survey, photogrammetry, SAR interferometry and laser ranging.
Traditionally, shoreline is three-dimensionally mapped by means of stereography on aerial photographs that, where
[2]
available, are used to detect past shoreline position . Near infrared and especially thermal infrared images can easily
divide emerged and submerged domains, however their use is limited by a coarse spatial resolution. Optical images are
characterized by higher spatial resolution, although needs to be pointed out that spectral the confusion due to turbidity,
[1]
foam and beached submerged vegetation limits the accuracy of shoreline detection. SAR images can be acquired both
all time and in all weather conditions, highlighting interesting potentialities, in spite of this signal echo from a sea
surface depends heavily on the instantaneous sea state: waves may cause echo overwhelming the return from the
[6]
[7]
emerged area . According to Baghdadi et al. , radar images yield a better definition of the shoreline with higher
[8]
incidence angle, whereas Kim et al. , report the higher accuracy in shoreline extraction the shorter the wavelength.
LIDAR data intrinsically provide information about beach topography and near-shore bathymetry, shoreline can be
[9]
extracted by intersecting LIDAR DEM and tidal datum surface . The integration of complimentary information such as
multispectral and geometric data [10]-[11] through fusion techniques can overcome the shortcomings of individual
methodologies.
However these studies allow retrieving only the instantaneous position of the shoreline that should also be considered
from a temporal point of view. This research aims to take into account both wave and tide effects, acting during the
image acquisition, on the user shoreline digitization. Each wave instantaneously determines the shoreline motion; the
sea-wave of operational interest depends on the “ordinary sea-storm”.

2. METHODOLOGY
Methodology includes several steps aiming to estimate the uncertainty in shoreline positioning by means of remote
sensing data, namely: the geological and geomorphological characterization of the study area, in order to identify the
sediment size and composition constituting the beach and determining its average slope; the diachronic analysis through
orthophotos acquired for suitable period to identify erosion phenomena (10 years, from June 2006 to August 2005 in the
present study); georeferencing, mosaicking of the images and digitization of the shoreline at a suitable scale; the
characterization of the ordinary sea-storm for each wind direction and year; the wave motion propagation from offshore
to nearshore; the estimation of the ordinary sea-storm maximum run-up; the displacement computation of the shoreline
[9]
due to tide influence. Wave motion plays the most important role in characterize the instantaneous shoreline position .
2.1 Maritime hydraulic study
In order to study the wave propagation from off to near shore the geographic aspect of the coast needs to be known. It
allows calculating the wind sector in terms of angle amplitude, wind direction, blind sectors due to islands interferences.
Once identified the wind sector, wave data recorded by the southern Mazara del Vallo buoy were acquired. The buoy is
part of the National Wave-meter Network (Rete Ondametrica Nazionale) and is located (37°31'00''N, 12°32'00''E) near
the Mazara del Vallo harbor [12] .

2.2 Statistical analysis of wave data
In order to describe the methodology, the term “ordinary sea-storm” needs to be defined. A sea-storm is a succession of
sea states during which the significant wave height, Hs(t), is usually greater than a critical threshold, hcrt, while it is lower
than the threshold only for periods shorter than Δtcrt. According to Boccotti [13] , the temporal, Δtcrt., and height, hcrt,
thresholds in Mediterranean Sea can be assumed equal to 12 hours and 1.5 m respectively. Within this work the prefix
“ordinary” refers to a sea-storm that is reached at least once a year. To simply retrieve the ordinary value, records were
filtered and sea storms were grouped, for each year, depending on 30° sectors centred on 12 incoming directions ranging
between 15° and 345° N. For each sea-storm height peak values were retrieved and the average time period was
calculated.
2.3 SWAN modelling
SWAN is a third generation model based on the energy balance spectrum, able to accurately assess wave parameters on
lakes, estuaries and coastal zones, given the bathymetry, the outline wave “clime” and wind initial conditions. The wave
celerity propagation is obtained according to the linear theory [14]-[15] . The model is based on the principle that surfaces
characteristics are retrievable from wave spectral variance, E (σ, θ), representing the wave energy distribution on angular
r
frequencies, σ, and propagation direction, θ. The model assesses the wave evolution through the action density N (| x |, t;

Proc. of SPIE Vol. 7824 78241Z-2

Downloaded from SPIE Digital Library on 05 Nov 2010 to 151.82.204.68. Terms of Use: http://spiedl.org/terms