Marine debris collects within the North Pacific Subtropical Convergence Zone

Abstract

Floating marine debris, particularly derelict fishing gear, is a hazard to fish, marine mammals, turtles, sea birds, coral reefs, and even human activities. To ameliorate the economic and environmental impact of marine debris, we need to efficiently locate and retrieve dangerous debris at sea. Guided by satellite-derived information, we made four flights north of Hawaii in March and April 2005. During these aerial surveys, we observed over 1800 individual pieces of debris, including 122 derelict fishing nets. The largest debris concentrations were found just north of the North Pacific Transition Zone Chlorophyll Front (TZCF) within the North Pacific Subtropical Convergence Zone (STCZ). Debris densities were significantly correlated with sea-surface temperature (SST), chlorophyll-a concentration (Chla), and the gradient of Chla. A Debris Estimated Likelihood Index (DELI) was developed to predict where high concentrations of debris would be most likely in the North Pacific during spring and early summer.

Full text

Marine debris collects within the North Pacific
Subtropical Convergence Zone
William G. Pichel
a
, James H. Churnside
b,*
, Timothy S. Veenstra
c
, David G. Foley
d,e
,
Karen S. Friedman
a
, Russell E. Brainard
f
, Jeremy B. Nicoll
g
, Quanan Zheng
h
,
Pablo Clemente-Colo
´n
a
a
National Oceanic and Atmospheric Administration (NOAA), National Environmental Satellite, Data, and Information Service,
Room 102 WWB, 5200 Auth Road, Camp Springs, MD 20746, USA
b
NOAA, Earth System Research Laboratory, 325 Broadway, Boulder, CO 80305, USA
c
Airborne Technologies Inc., 4338 N. Gunflint Trail, Wasilla, AK 99654, USA
d
Joint Institute for Marine and Atmospheric Research, University of Hawaii, USA
e
NOAA Southwest Fisheries Science Center, Environmental Research Division, 1352 Lighthouse Ave., Pacific Grove, CA 93950, USA
f
NOAA Pacific Islands Fisheries Science Center, 2570 Dole Street, Honolulu, HI 96822, USA
g
Alaska Satellite Facility, Geophysical Institute, University of Alaska Fairbanks, P.O. Box 757320, Fairbanks, AK 99775, USA
h
Department of Atmospheric and Oceanic Science, University of Maryland, College Park, MD 20742, USA
Abstract
Floating marine debris, particularly derelict fishing gear, is a hazard to fish, marine mammals, turtles, sea birds, coral reefs, and even
human activities. To ameliorate the economic and environmental impact of marine debris, we need to efficiently locate and retrieve dan-
gerous debris at sea. Guided by satellite-derived information, we made four flights north of Hawaii in March and April 2005. During
these aerial surveys, we observed over 1800 individual pieces of debris, including 122 derelict fishing nets. The largest debris concentra-
tions were found just north of the North Pacific Transition Zone Chlorophyll Front (TZCF) within the North Pacific Subtropical Con-
vergence Zone (STCZ). Debris densities were significantly correlated with sea-surface temperature (SST), chlorophyll-aconcentration
(Chla), and the gradient of Chla. A Debris Estimated Likelihood Index (DELI) was developed to predict where high concentrations
of debris would be most likely in the North Pacific during spring and early summer.
Ó2007 Elsevier Ltd. All rights reserved.
Keywords: Marine debris; Derelict fishing gear; North Pacific Subtropical Convergence Zone
1. Introduction
The increased use of long-lasting synthetic materials
since the 1950s has led to increases in the negative effects
of marine debris. Plastic particles are ingested by sea birds
and small marine animals instead of food (Carr, 1987).
Fish, birds, sea turtles, and marine mammals are entangled
and killed in derelict nets (also referred to as ‘‘ghostnets’’)
(Laist, 1987). Ships and submersibles can be trapped by the
fouling of nets and lines in propellers or damaged by colli-
sion with bulky floating debris. Nets physically damage
coral reef substrates while they continue to entangle and
kill reef animals (Donohue et al., 2001). Between 1996
and 2006, NOAA recovered a total of 511 metric tons of
fishing gear from the reefs of the Northwest Hawaiian
Island Marine National Monument (NWHI-MNM), one
of the largest marine conservation areas in the world. A
recent report estimates an annual accumulation rate of
approximately 52 metric tons of derelict fishing gear
(Dameron et al., 2006). The economic loss of commercial
0025-326X/$ - see front matter Ó2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.marpolbul.2007.04.010
*
Corresponding author. Tel.: +1 303 497 6744.
E-mail address: james.h.churnside@noaa.gov (J.H. Churnside).
www.elsevier.com/locate/marpolbul
Marine Pollution Bulletin 54 (2007) 1207–1211

fish caught in ghostnets; the time and expense of rescue
operations for entangled or damaged vessels; the environ-
mental loss of endangered species and rare coral; and the
cost of cutting nets by hand from reefs can be measured
in millions of dollars annually in the US alone. To mitigate
some of these effects, we pursued the development of a pro-
cedure to economically locate and cost effectively remove
derelict nets at sea, prior to much of the environmental
damage.
Ocean circulation models, satellite remote sensing data,
and aircraft observations were used to detect derelict nets
and other debris in the open ocean. Ocean circulation mod-
els indicated surface convergence in the vicinity of the
STCZ north of the Hawaiian Islands. Satellite remote sens-
ing data, including wind stress, SST, and Chla, indicated
specific areas within the STCZ where higher concentrations
of debris might be expected. Aerial surveys, based on these
data, provided the distribution of the density of marine
debris of several types.
Ocean circulation and wind-drift models suggest that
debris in the North Pacific would tend to concentrate along
a southwest-to-northeast line north of the Hawaiian
Islands (Kubota, 1994) that coincides with the STCZ
(Fig. 1). In the central North Pacific, the STCZ is located
between 23°N and 37°N, seasonally migrating between
these extremes (Roden, 1975), and marks the boundary
between the productive waters of the North Pacific Transi-
tion Zone and the oligotrophic waters to the south. In gen-
eral, the STCZ is located along the axis of the North Pacific
high pressure ridge between the mid-latitude westerlies and
the easterly trade winds, and convergence is caused by
Ekman transport (Roden, 1975).
The TZCF migrates throughout the year, with its north-
ernmost position in summer and southernmost in winter.
Interannual climate signatures, such as El Nin
˜o Southern
Oscillation, influence the TZCF. For example, during El
Nin
˜o years, the southernmost position of the TZCF is gen-
erally further to the south than in normal years. Fig. 2 is a
time series of the TZCF position as tracked by the 0.2
mg/m
3
Chlaisopleth and the 18 °C isotherm, both of which
usually coincide with the TZCF (Bograd et al., 2004).
Chlorophyll measurements are from the Sea-viewing Wide
Field-of-view Sensor (SeaWiFS) on the OrbView-2 satellite
and SST measurements are from the Advanced Very High
Resolution Radiometer (AVHRR) on the NOAA series of
polar-orbiting operational environmental satellites. Both
were derived from a time series of 8 day averages between
160°W and 180°W. The background shows the wind stress
curl calculated from satellite scatterometers (the Advanced
Microwave Instrument on the ERS-2 spacecraft and Sea-
winds on the QuikSCAT spacecraft). The wintertime posi-
tion of the TZCF coincides with the region of maximum
negative wind stress curl (i.e., maximum convergence from
wind-driven Ekman drift) (Bograd et al., 2004).
Thus, we expected that marine debris would be concen-
trated during winter when the surface convergence is stron-
gest. Since the same winds that create the strong
convergence make aerial observations of the ocean surface
difficult, the best time to search for debris should be in the
spring, when calmer weather occurs in the region of maxi-
mum wintertime convergence, and before the debris has
Fig. 1. 14-Day SST GOES composite for the period ending March 31,
2005 with the four GhostNet flight tracks superimposed. (Flight 1, March
18 – black; Flight 2, March 27 – red violet; Flight 3, March 29 – green;
Flight 4, April 3 – blue).
Fig. 2. Time series of Chlaand SST fronts, defined by the 0.2 mg m
3
isopleth (red line) and the 18 °C isotherm (white line), respectively. The
background is wind stress curl (negative indicates convergence).
1208 W.G. Pichel et al. / Marine Pollution Bulletin 54 (2007) 1207–1211

had time to disperse. For these reasons, we made a series of
aerial surveys in the spring of 2005.
2. Methods
We made four flights (Fig. 1) out of Honolulu, Hawaii
between March 18, 2005 and April 3, 2005 on a Lockheed
WP-3D. A short test flight was followed by three flights of
roughly 9 h each. During the observation periods, the flight
altitude was 300 m and the speed was about 100 m s
1
. The
flight tracks were chosen to include areas of high SST and
Chlagradients and also regions of low gradients, based on
multi-day composites of satellite-derived SST and Chlafor
January through March. We also used meteorological fore-
casts to avoid high winds (>15 m s
1
), since rough seas
make visual observations difficult.
Four to six people made visual observations and
reported them over the aircraft intercom system to a recor-
der, who wrote down all observations within each 1-minute
period. Most sightings were first made by one or both
pilots. An experienced observer at a window behind the
cockpit on each side of the cabin confirmed the pilot sight-
ings and occasionally added sightings that were missed by
the pilots. Periodically, 1 or 2 additional observers near
the rear of the cabin checked the performance of the pri-
mary observers.
After the flights, the density of debris was estimated for
each 6 min flight period using the speed of the aircraft and
an effective width of the visual survey of 1000 m. This
width was estimated from geometric measurements of vis-
ibility from the aircraft windows made on the ground
and from angular measurements of some of the sightings
that were made during flight.
Satellite data at the positions of the debris-density esti-
mates were extracted from 14-day composite images of
Chlaconcentration and SST. These were obtained from
the MODIS/Aqua and the infrared imager on the NOAA
Geostationary Operational Environmental Satellite
(GOES-10), respectively. The spatial resolution of the
SST data set was 0.05 by 0.05°, while that of the Chladata
was 0.025 by 0.025°. We also calculated the magnitude of
the gradients of chlorophyll and SST to get a total of four
geophysical parameters. These gradients were calculated
using adjacent 0.05°resolution elements, and the longitudi-
nal gradients were adjusted by the cosine of the latitude.
These four geophysical parameters were binned by value
using the following bin widths: SST – 0.5 °C; SST gradient
– 0.25 °C per degree of latitude; Chla– 0.025 mg m
3
; and
Chlagradient – 0.015 mg m
3
per degree of latitude.
The total number of debris sightings corresponding to
each bin of each of the geophysical parameters was divided
by the time spent searching in water with that parameter to
obtain a measure of debris per unit effort (DPUE) for each
parameter.We fit each DPUE to a fourth-order polyno-
mial. Finally, we defined a Debris Estimated Likelihood
Index (DELI) as the product of the DPUEs for SST, Chla,
and Chlagradient, normalized by the largest value of this
product.
3. Results
The total numbers of objects observed are presented by
type in Table 1. By far, the most common form of debris
was fishing floats, typically foam ‘‘corks’’ and plastic
buoys. Less common were large net bundles, but some of
these were very large (two were greater than 10 m in diam-
eter). The category ‘‘general debris’’ includes plastic bags,
pails, and any other miscellaneous debris.
Fig. 3 presents the observed debris densities for Flights
2, 3 and 4 (Flight 1 was a test and calibration flight). Here
the 6 min totals for debris sightings have been displayed as
debris density (number of sightings per km
2
) superimposed
on a background of Chlainferred from the Moderate Res-
olution Imaging Spectroradiometer (MODIS) on the
NASA Aqua satellite for the 14-day period spanning the
time period of the flights. One can clearly see a large
increase in debris density in the regions of higher chloro-
phyll (above the TZCF with chlorophyll equal to or greater
than 0.2 mg m
3
– in general the orange and red regions in
Fig. 3). It should be noted, however, that a number of nets
were sighted south of 30°N on the western leg of Flight 3
and the eastern leg of Flight 4 (which were close in longi-
tude to each other). This may mark an area of convergence
in the past, a ‘‘fossil front’’ (Roden, 1980) with debris still
resident.
The distribution of debris density by environmental
parameter is presented in Fig. 4. The densities are averaged
over 6 min sections of the flights (36 km
2
). The figure also
includes a least-squares fit of a fourth-order polynomial
to each distribution. The dashed, cyan lines represent the
prediction error estimate. They represent one standard
deviation above and below the fit, based on a covariance
matrix
C¼R0R0Tkrk2=n;ð1Þ
where R0is the inverse of the Cholesky factor Rfrom a QR
decomposition of the Vandermonde matrix of the fit (Press
et al., 1992), the superscript T represents the transpose of
that matrix, nis the number of degrees of freedom, and
krkis the norm of the residuals.
Table 1
Debris totals by category
Debris type Flight 2 Flight 3 Flight 4 Total
Buoys 24 31 13 68
General debris 29 86 158 273
Floats 201 560 631 1394
Lines 1 9 13 23
Logs 4 2 1 7
Nets 20 43 59 122
Total 279 731 875 1885
W.G. Pichel et al. / Marine Pollution Bulletin 54 (2007) 1207–1211 1209

There was a strong correlation between the density of
debris and the geophysical parameters, but this was partly
because we concentrated our search in areas where we
expected to find debris based on the satellite observations.
After removing this effect through the calculation of the
amount of debris per unit effort (DPUE), we obtained cor-
relations of 0.87 (n= 1832, p< 0.0001) with SST, 0.61
(n= 1832, p= 0.01) with SST gradient, 0.76 (n= 1630,
p= 0.0005) with Chla, and 0.90 (n= 1550, p< 0.0001) with
Chlagradient. Since we observed significant correlations
between debris density and three of the geophysical param-
eters, we developed a method to combine these parameters
into a measure of the relative likelihood of encountering
concentrations of marine debris, the Debris Estimated
Likelihood Index (DELI). This produces a map (Fig. 5)
that can be used to direct future aerial searches, and can
also be used as a starting point for an estimate of the total
amount of debris in the North Pacific. This map clearly
shows that there are very localized regions of high likeli-
hood. As derived, the DELI map is valid only for spring
and early summer. As the convergence moves north, it will
weaken and debris will begin to disperse from the early
summer positions.
We also recorded animal sightings, including birds, dol-
phins, whales, turtles, and fish (Table 2). No attempt was
made to get more detailed information than presented.
The largest category is dolphins, but this included one
group of several hundred animals. For our comparison,
Fig. 3. Debris density for GhostNet flights 2 (red violet), 3 (green), and 4 (blue) over a background of Chla.
Fig. 4. Distribution of debris density by environmental parameter, along
with the fourth-order polynomial fit (red line) and the lines representing
±1 standard deviation (cyan).
Fig. 5. Debris Estimated Likelihood Index (DELI) with relative likeli-
hood of finding debris indicated by color. Gray denotes Hawaiian Islands,
with circles at the locations of reefs too small to be resolved.
1210 W.G. Pichel et al. / Marine Pollution Bulletin 54 (2007) 1207–1211

we used 250 animals, but the actual number could be any-
where between 200 and 300. Without this group, there were
more birds than any other type. Many of these were clearly
albatross, but most were not identified. The fish that could
be identified were all sharks.
The overall correlation between animal sightings and
debris sightings was only 0.19 (n= 150, p= 0.02). How-
ever, this low correlation was strongly influenced by a
group of several hundred dolphins that were travelling in
the area. If we remove these animals from the data set,
the correlation improves to 0.75 (n= 149, p< 0.0001).
We conclude that the pelagic animals we observed were
preferentially foraging in the same convergence zones that
concentrate marine debris, thus increasing their risk (Polo-
vina et al., 2001; Seki et al., 2002).
4. Conclusions
The challenge was to develop a cost-effective method to
search the vast ocean. Such a method has been found for
the North Pacific. From ocean circulation models, we
know that the highest concentration of debris will be con-
centrated in the STCZ in the spring and early summer. We
have shown that aerial surveys can be localized within the
STCZ using satellite-derived information about SST and
Chla.
The data collected during the aerial surveys support the
hypothesis of a substantial increase in the rate of debris
deposition on the reefs of the recently established North-
western Hawaiian Islands Marine National Monument
(NWHI-MNM) in winter and early spring when the TZCF
moves south into the islands. They also suggest that a vari-
ety of marine animals tend to be concentrated in the same
areas as marine debris.
Acknowledgements
The authors acknowledge the expertise and dedication
of the pilots, crew, and support personnel of the NOAA
Aircraft Operations Center; the assistance of those who
served as observers on the flights; the insight provided by
James Ingraham and his drift model studies of debris circu-
lation in the Pacific; Michael Van Woert for his insights
concerning the subtropical convergence zone; Capt.
Charles Moore for sharing his experience with actual deb-
ris observations at sea. The NOAA CoastWatch program
provided the GOES SST data; the NASA Goddard Space
Flight Center (GSFC) provided the MODIS satellite data;
the NOAA National Oceanographic Data Center provided
the Pathfinder SST data; GSFC and Orbimage Inc. pro-
vided SeaWiFS Chlorophyll adata; and the French Insti-
tute for the Research and Exploitation of the Sea
(IFREMER) provided the scatterometer wind data. The
views, opinions, and findings contained in this report are
those of the authors and should not be construed as an offi-
cial National Oceanic and Atmospheric Administration or
US Government position, policy, or decision.
References
Bograd, S.J., Foley, D.G., Schwing, F.B., Wilson, C., Laurs, R.M.,
Polovina, J.J., Howell, E.A., Brainard, R.E., 2004. On the seasonal
and interannual migrations of the transition zone chlorophyll front.
Geophysical Research Letters 31 (L17204), 1–5. doi:10.1029/
2004GL02063.
Carr, A., 1987. Impact of nondegradable marine debris on the ecology and
survival outlook of sea turtles. Marine Pollution Bulletin 18, 352–
356.
Dameron, O.J., Parke, M., Albins, M.A., Brainard, R., Marine debris
accumulation in the Northwestern Hawaiian Islands: an examination
of rates and processes. Marine Pollution Bulletin. doi:10.1016/
j.marpolbul.2006.11.019.
Donohue, M., Boland, R.C., Sramek, C.M., Antonelis, G.A., 2001.
Derelict fishing gear in the Northwestern Hawaiian Islands: diving
surveys and debris removal in 1999 confirm threat to coral reef
ecosystems. Marine Pollution Bulletin 42, 1301–1312.
Kubota, M., 1994. A mechanism for the accumulation of floating marine
debris north of Hawaii. Journal of Physical Oceanography 24, 1059–
1064.
Laist, D.W., 1987. Overview of the biological effects of lost and discarded
plastic debris in the marine environment. Marine Pollution Bulletin 18,
319–326.
Polovina, J.J., Howell, E., Kobayashi, D.R., Seki, M.P., 2001. The
transition zone chlorophyll front, a dynamic global feature defining
migration and forage habitat for marine resources. Progress in
Oceanography 29, 469–483.
Press, W.H., Teukolsky, S.A., Vetterling, W.T., Flannery, B.P., 1992.
Numerical Recipes in C: The Art of Scientific Computing, second ed.
Cambridge University Press, Cambridge, pp. 96–102.
Roden, Gunnar, 1975. On North Pacific temperature, salinity, sound
velocity and density fronts and their relation to the wind and energy
flux fields. Journal of Physical Oceanography 5, 557–571.
Roden, Gunnar, 1980. On the subtropical frontal zone north of Hawaii
during winter. Journal of Physical Oceanography 10, 342–362.
Seki, M.P., Polovina, J.J., Kobayashi, D.R., Bidigare, R.R., Mitchum,
G.T., 2002. An oceanographic characterization of swordfish (Xiphias
gladius) longline fishing grounds in the springtime subtropical North
Pacific. Fisheries Oceanography 11, 251–266.
Table 2
Animal sightings by type
Animal type Flight 2 Flight 3 Flight 4 Total
Birds 65 32 46 143
Turtles 1 0 1 2
Whales 1 0 19 20
Dolphins 0 0 256 256
Fish 0 4 3 7
Total 67 36 325 428
W.G. Pichel et al. / Marine Pollution Bulletin 54 (2007) 1207–1211 1211