indication of W EBCAM MONITARING A ND INVESTIGATION in situ National Institute for Land and Infrastructure Management NILIM Coastal Marine and Disaster Prevention Department Coastal Zone Systems Division ID: 223677
Download Presentation The PPT/PDF document "Movement of plastic debris on beaches" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Movement of plastic debris on beaches~indication of WEBCAM MONITARINGAND INVESTIGATION in situ~
National Institute for Land and Infrastructure Management (NILIM)Coastal, Marine and Disaster Prevention DepartmentCoastal Zone Systems DivisionTomoya KataokaHirofumi Hinata
Lunchtime Discussion at IPRC
20
February 2013Slide2
Kurochio
currentPlastic pollution
Risk caused by plastic debris
Threats to marine animals
e
x.
Ingestion of plastics, Entangled sea turtles
Pollution by to
xic chemicals in plastic debrisex. Lead that leach into beach (Nakashima et al., EST, 2012)Transportation of plastic debrisOutflow from sourcesWashing ashoreReturning to oceanWashing ashore … To Pacific ocean
Jean
Jean
Jean
Wash ashore
O
utflow
O
utflow
O
utflow
Return offshore
Wash ashore
Return offshore
To Pacific ocean
O
utflowSlide3
Transportation process in East Asian SeasMonitoring quantity of debris at multiple sites Webcam monitoring (Low cost and Few labor)Simulating drifting debris Ocean circulation model and its quantity at multiple siteBeaching process in nearshore (included returning process offshore)Movement of debris on beach Investigation in situ
Today’s First Topic
Today’s
Second
T
opic
Our
focusSlide4
(e) Ishigaki
(a)Tsushima
(b)Wajima
(c)Tobishima
）
(d) Wakkanai
(i) Shiriya
(h)Niijima
(g)Muroto
(f)Tanegashima
Nine monitoring sites
Tsushima current
：
4 sites
Tsushima current: 4 sites
Kuroshio current: 4 sites
Tsushima current: 4 sites
Kuroshio current: 4 sites
Oyashio: 1 site
【Today’s topic】
Tsushima Current: 4 sites
Webcam monitoring sitesSlide5
The webcam monitoring system
Webcam
Control box
Storage batteries
Our laboratory in NILIM
Transmission via the Internet
Time
Photographs
Solar panels
Operating time
every two hours from 7:00 to 15:00 (i.e., five
operations
every day)
Taking photographs
five photographs every operating time
Daily number of photographs
25 photographs (i.e., 5
(operations)×
5 (photos)=25)
Saving in a built
-in
Local storage
(SDHC card, 32GB)
Webcam monitoring systemSlide6
Animation of webcam image
Tobishima
December 2010Slide7
1. Generation of color referencesColor references for detecting pixels of plastic debris are generated using CIELUV color space.CIELUV color space: This space is one of three-dimensional color spaces Colors are expressed by (L*, u*
, v*).2. Detection of plastic debris pixels using the color referencesPlastic debris pixels are detected using color references and a composite method.
v
*
u
*
u
*
-
v
*
coordinate
v
*
u
*
L
*
-
v
*
coordinate
L
*
-
u
*
coordinate
L
*
L
*
1) T. Kataoka et al.(2012), MPB, 64, 1829-1836
Technique for detecting plastic debri
s pixelsSlide8
,
C
Color distribution of plastic debris “C” in the CIELUV color space.
Color references, which are generated from webcam images at Tobishima.
Three ellipses form the ellipsoid
body
Ellipsoid
body in the CIELUV color space
Color references
G
E
A
H
D
F
B
a
Periods: p1, November 21-27, 2010; p2, December 14-21, 2010; p3, February 16-22, 2011; p4, April 1-7, 2011; p5, May 18-24, 2011
Generation of color referencesSlide9
,
【Detection of plastic debris using a single photograph】
Plastic debris with various colors
detection
e.g., Driftwoods
misdetection
Detection using one photograph
misdetected
Composite method:
(
L
*
,
u
*
,
v
*
) of the pixel is frequently located in the ellipsoid body of a certain color using photographs for three days.
Plastic debris pixels
Not plastic debris
For 3 days
Plastic debris
Using one photo
Plastic debris
white
Using photos for three days
Driftwood
Plastic debris
Detection using CRs and compositeSlide10
Projective transformation
1)
1) S.
Magome et al.(2007),JO,63,761-773
(
X
,
Y
): Geographic coordinate
(
x
,
y
): Photographic coordinate
Coefficients
A
=
N
×
a
N
: Number of plastic pixels
a
: Area of a single pixel
Covered area
A
Beached plastic debris quantity
C
overed area
GPS
Webcam image
Converted image
Error of computed area: within5.0%
Calculation of debris quantity
m
0 5 10Slide11
clean-up
clean-up
Err.:
19%
, Corr.:
0.93
Err.:
12%
, Corr.:
0.81
Err.: 20
%
, Corr.:
0.94
Err.: 21
%
, Corr.:
0.73
Time series of daily covered
areas
Short-term fluctuations
depended on changes
in the weather conditions (e.g., amount of
sunlight, wind).
Removing
the short-term
fluctuations
T
ime
series of
30-day
moving averages (bold line
).
Time series of plastic debris quantitySlide12
Period:
1month (April , 2011)
Time: 7:00, 9:00, 11:00, 13:00, 15:00Plastic Debris: Three type
(
Tank (Blue), Cylinder Buoy (White), Cube Buoy (White))
Cylinder
Side
Cylinder
Top
Cube
Top
Cube
Side
Tank
Top
Tank
Side
Calculate average value and standard deviation of (L, u, v)
Calculate average value and standard deviation of (L, u, v)
What factor causes underestimation?Slide13
Blue (Tank)White (Cube)White(Cylinder)
E
W
Morning
E
W
E
W
Afternoon
Noon
Monthly variability of
debris
color (Side)Slide14
Wakkanai
2010/10/9
Tobishima
2010/11/5
Wajima
2010/12/5
Tsushima
2010/11/22
Webcam sites
Wind obs. sites
Comparison with wind data
Wind
Wind
Wind
Wind
Q
in
: Quantity of debris washed ashore
Q
out
: Quantity of debris returned offshore
Factors for variability of debris quantity
Q
in
>
Q
out
Increase
Q
in
<
Q
out
decrease
Q
in
=
Q
out
not change
Near shore current
Wave
WindSlide15
【Major conclusion for webcam monitoring】Sequential monitoring of quantity of debris at multiple sites
SucceededDebris quantity
not only vary by intensification of westerly wind but also by other factors
At
Wajima
and Tsushima
increase according to intensification westerly wind for Nov. 2010 – Mar. 2011
At Wakkanai and Tobishima Not vary despite intensification of westerly windAt Tsushima
Increase greatly despite no westerly wind
Factor of the underestimation Strong dependence on sunlight angle
Using only photographs on certain time when variability of debris color is the lowest
Measurement error can be minimized【Major conclusion for Investigation in situ】
Found seasonal variation of movement and quantity of debris
Established residual function of debris
Estimated residence time on
Wadahama beach
152 days (Three type buoys)98 days (PET bottle)
Obtained system characteristics on Wadahama beach
Summary