5. COMPETENCE OF SAR IMA
SPILL DETECTI
The problem of pollution of the oceanic wat oil materials is presently considered as one
As first step for solution of the oil spill re using airborne and spaceborne remote se
surveillance of the sea surface. Such system all-weather observation, be independent fro
conditions and cloud cover, define the volume of oil spill and work in real scale of
aperture radar SAR is independent of illumination conditions and allow to acquire
and night under cloud cover that is an ad remote sensing sensors.
SAR is an active remote sensing tool in wh satellite transmits microwave signals tow
surface, SAR signal after interaction with the to the antenna [22]. The interaction between
microwaves is very sensitive to variatio roughness. Rough surfaces scatter large amo
to the antenna and have bright signatures wh reflect the energy away from the anten
signatures. It is well-known, that crude substances form films of various thicknesses
Oil films locally damp sea surface roughn signatures, so-called slicks, on the SAR im
look on the SAR images as dark patch surrounding surface.
During the lifetime of oil spill in the sea it number of processes, which dramatically
chemical properties. Called by term weatheri are as follows: spreading, drift, evapo
emulsification, bacterial degradation and ph With time the physical-chemical propertie
changed due to effect of these processes. Th important role in oil spill detection by using
but relative importance of each process is n It is also reported that the thick part of oil
only 10 of spill area, while remaining oil c an area as blue shine [22].
The detectability of oil slicksspills in SA depends on the wind speed at the sea surfac
speeds, typically between 0 and 2-3 ms, th dark on SAR images. In this case the wind-g
not already developed and oil films look background, - detection in this case is impo
between 3 and 6 ms is ideal for oil slick surface roughness is developed and oil sli
patches on a bright background. However, reaches 10-12 ms, detectability is im
obstructed due to the redistribution of oil surface waves and wind-induced mixing i
layer [23]. as a result slick disappears from SAR imagery.
Other detection problem is discrimination and natural organic oil slicks [8]. Natural
very small thickness resulting from life-cyc other marine organisms can form surface
surface and, in turn, produce similar dark sig images [7]. Experiments conducted with mu
polarization SIR-CX aboard the space shu
F AGERY FOR OIL
TION
aters by crude oil and ne of the most sharp.
l response problem is sensing systems for
tem must provide the from of illumination
e position, type and of time [24]. Synthetic
of weather and sun uire SAR images day
advantage over other which an antenna on a
toward to the ocean the sea surface returns
en the sea surface and tions in sea surface
mount of energy back while smooth surfaces
tenna and have dark de oil and other oil
ses on the sea surface. ghness and give dark
R images [9];[8], they tches among brighter
it will be expose to a ly influence physical-
ering, these processes aporation, dispersion,
photo oxidation [12]. rties of oil spills are
. These processes play ing space borne SAR,
s not well understood. il film usually covers
il covers up to 90 of SAR images strongly
rface. Under low wind , the sea surface looks
d-generated waves are ooks dark on a dark
possible. Wind speed ick detection, the sea
slicks appear as dark er, when wind speed
impossible again or il spillsslicks by the
g in the upper ocean m the sea surface and
n between man-made ral biogenic films of a
cycle of plankton and ce slicks on the sea
signatures on the SAR multi-frequencymulti-
shuttle [16]; [7], as expected could provide more relia
discrimination.
6. METHO
In this study, both visual and techniques were applied. The me
the following sections.
Step 1: Oil spill areas identificati
We applied several image enhance and others, this step allowed us to i
are potential oil spills, followed identified 14 areas, where each ha
step helped us to focus our followi the identified areas without the
scenes.
Step 2: Image Filtering
Several filtering methods are used t areas representing possible oil spill
Lee 7x7, Gamma 7x7, Frost 7x used in different oil spill discrimina
[19]. It is also observed that these information boundaries of high and
testing many of these filters, we cho to all sub-images obtained from th
Step 3: Visual analysis and oil spi
In order to analyze and interpret spread of oil spill, it is very
understand some phenomena and fe spill process.
Oil spill in th immediately start to spread, dispers
processes that are collectively calle progress, duration and result of
depend on the amount, composition oil. Illegal oil slicks from tankers a
size and they are therefore easier d comparison with hazardous acciden
As well, we need to consider the en as mixing, currents, temperature an
for the weathering of spilled oil [1 impact on the spill during the first f
very important to detect the first crucial to have a continuous, hou
Large amounts of oil will result to short time after dumping. On the
volatile and dissoluble in seawater spills, in comparison to those w
spilled in the same amounts [22]
Also in our interpretation we consider the look-alike
phenomena that appear under sea-air boundary layer
conditions, which may be classified as oil spill in radar
images. These may include, natural slicks threshold
winds, wind shadows behind islands, winds slicks calm
zones, surface currents, internal waves, reflections of
Figure 2: Oil Spill Map from July 24
KompSAR- 5 reliable information for oil slick
ETHODOLOGY
and digital image processing methodology used is shown in
ation
ncement such as linear contract to identify the areas where there
d this interpretation we could has at least one oil spill. This
wing image processing only on e need to process the whole
d to discriminate high contrast ills in the images. In general
7x7 and pass 3x3 filters are ination studies [1], [17],
se filters minimize the loss of nd low contrasted areas. After
choose to apply Gamma7x7 the first interpretation step.
spill digitization
ret correctly the existence and y important to consider and
d features happing during the oil the marine environment will
ersed or degenerated by natural alled weathering over time. The
of the weathering processes tion and properties of the spilled
s are relatively thin and small in r degenerated by weathering in
dental oil spill. e environmental conditions such
and sunlight are also important [15]. As weathering processes
rst few hours after dumping, it is rst and new oil spill, then it is
ours or daily detection system. to large spills, usually within a
the other hand oils with large ter content will result to smaller
with less such content, when
This contribution has been peer-reviewed. doi:10.5194isprsarchives-XLI-B8-1115-2016
1118
Figure 3: location map of oil spill region in the study area
Figure 5: reprojection of AIS Information on the
Figure 6: Analysis and comparison of three successive data contained oil
spill the bottom topography in shallow waters, fresh water slicks,
plums of municipal sewage, wave shadows behind land or structures [19].
In each oil spill area identified, there are several oil spills, we could map and digitize each one, then this followed by
calculation of the area size, and create oil spill map Figure 2,
We used several remote sensing “image processing” software in
particularly Geomatica 15 and ENVI 5.1, as well we found Google Earth is a user friendly tool to visualize the results and
analyze it through the acquisition period.
7. ANALYSIS AND RESULTS
Kategori