Seatrack Web: A Numerical Tool to Protect the Baltic Sea Marine Protected Areas

A. G. Kostianoy1, C. Ambjörn2 & D. M. Soloviev3

1 - P.P. Shirshov Institute of Oceanology, Russian Academy of Sciences, 36 Nakhimovsky Pr., Moscow, 117997, Russia 2 - Swedish Meteorological and Hydrological Institute, Norrköping, Sweden

3 - Marine Hydrophysical Institute, National Academy of Sciences of Ukraine, Sevastopol, Ukraine

Abstract-Shipping activities in the Baltic Sea, including oil

transport and oil handled in harbors, have a number of negative impacts on the marine environment, marine protected areas (MPAs) and coastal zone. One of the main tasks in the ecological monitoring of the MPAs in the European seas is an operational satellite and aerial detection of oil spillages, determination of their characteristics, establishment of the pollution sources and forecast of probable trajectories of the oil spill transport. The interactive numerical model Seatrack Web SMHI is a powerful operational tool that can be used for a forecast of the oil spill drift in the vicinity of MPAs and for assessment of ecological risks related to potential oil pollution of every MPAs in the Baltic Sea. Three examples of oil spill drift modelling and of calculation a probability of the oil drift for specific points along main ship routes in the Gulf of Finland and southward of Gotland for July and August 2007 are shown.

I. INTRODUCTION

As highlighted by Oceana in its report “The Other Side of Oil Slicks”, chronic hydrocarbon contamination from washing out tanks and dumping bilge water and other oily waste represents a danger at least three times higher than that posed by the oil slicks resulting from oil tanker accidents [1, 2]. For example, in the North Sea the volume of illegal hydrocarbon dumping is estimated at 15,000–60,000 tons per year, added to which are another 10-20,000 tons of authorized dumping. Oil and gas platforms account for 75% of the oil pollution in the North Sea via seepage and the intentional release of oil-based drilling muds [3]. In the Mediterranean Sea it has been estimated at 400,000–1,000,000 tons a year. Of this about 50% comes from routine ship operations and the remaining 50% comes from land-based sources via surface runoff [3]. In the Baltic Sea this volume is estimated at another 1,750–5,000 tons a year [1, 2]. But, according to Finnish Environment Institute (http://www.ymparisto.fi, 2004), the total annual number of oil spills into the Baltic Sea may reach 10,000 and the total amount of oil running into the sea can be as much as 10,000 tons which is considerably more than the amount of oil pouring into the sea in accidents.

In recent years a number of new oil terminals have been built in the Baltic Sea area, resulting in increased transport of oil by ships and, consequently, an increased risk of accidents (Fig.1). Transportation is a reason of 45% of the oil in the sea. In the Baltic Sea about 2,000 large ships and tankers are at sea every day. Shipping activities in the Baltic Sea, including oil transport and oil handled in harbors, have a number of

negative impacts on the marine environment, marine protected areas (MPAs) and coastal zone (Fig. 2). Oil discharges from ships cause the contamination of seawater, shores, and beaches, which may persist for a long time or several years and represent a threat to marine ecosystems in MPAs and marine resources. One of the main tasks in the ecological monitoring of the MPAs in the European seas is an operational satellite and aerial detection of oil spillages, determination of their characteristics, establishment of the pollution sources and forecast of probable trajectories of the oil spill transport. Satellite imagery can help greatly identifying probable spills over very large areas and then guiding aerial surveys for precise observation of specific locations [5]. The Synthetic Aperture Radar (SAR) instrument (on board ENVISAT, ERS-2 and RADARSAT satellites), which can collect data independently of weather and light conditions, is an excellent tool to monitor and detect oil on water surfaces (Fig.3). Visible and infrared wavelength remote sensing from MODIS-Terra and –Aqua, and AVHRR NOAA radiometers can help in monitoring of suspended matter distribution and transport, as well as of bloom events in summertime (Fig.4 and 5).

10° 15° 20° 25° 30°

10° 15° 20° 25° 30°

55°

60°

65°

55°

60°

65°

S

W

E

D

E

N

F I

N L

A

N D

ESTONIA

LATVIA

LITHUANIA

POLANDGERMANY

DE

NM

AR

KN

O R

W A

Y

RUSSIA

RUSSIA BELARUS

Figure 1. Map of oil spills detected in the Baltic Sea in 1989-2002 [4].

Category 1 (/) - <1 m3, category 2 (*) - 1 m3 < 10 m3, category 3 (,) - 10 m3 < 100 m3, category 4 (&) - >100 m3.

978-1-4244-2268-5/08/$25.00 ©2008 IEEE

Figure 2. The Baltic Sea Protected Areas

Such a complex operational monitoring of the Southeastern Baltic Sea was performed in June 2004 – November 2005 on the base of daily satellite remote sensing (AVHRR NOAA, MODIS, TOPEX/Poseidon, Jason-1, ENVISAT ASAR and RADARSAT SAR imagery) of sea surface temperature (SST), sea level, chlorophyll concentration, mesoscale dynamics, wind and waves, and oil spills [6-9]. As a result a complex information on oil pollution of the sea, SST, distribution of suspended matter, chlorophyll concentration, sea currents and meteorological parameters has been received. In total 274 oil

spills were detected in 230 ASAR ENVISAT images and 17 SAR RADARSAT images received during 18 months. The interactive numerical model Seatrack Web operated by the Swedish Meteorological and Hydrological Institute (SMHI) [10] was used for a forecast of the drift of (1) all large oil spills detected by ASAR ENVISAT in the southeastern Baltic Sea and (2) virtual (simulated) oil spills from the D-6 oil platform [8-9]. This experience was applied for assessment of ecological risks related to potential oil pollution of several MPAs in the Baltic Sea.

(a)

(b)

Figure 3. Oil spills detected in the Baltic Sea Marine Protected Areas on 2 November 2004 (a) and 18 October 2005 (b) on ASAR Envisat imagery [8-9].

Figure 4. Suspended matter distribution in the northeastern Baltic Sea derived

from MODIS-Aqua on 2 April 2004.

Figure 5. Bloom in the Baltic Sea derived from MODIS-Terra on 13 July 2005.

II. SEATRACK WEB MODEL

The interactive numerical model Seatrack Web SMHI is a powerful operational tool that can be used for a forecast of the oil spill drift in the vicinity of MPAs and for assessment of ecological risks related to potential oil pollution of every

MPAs in the Baltic Sea. This version of a numerical model on the Internet platform has been developed at the Swedish Meteorological and Hydrological Institute in close co-operation with Danish authorities. The system uses two different operational weather models ECMWF and HIRLAM and a circulation model HIROMB (driven by the two weather models respectively), which calculates the current field. The model allows to forecast the oil drift for five days ahead or to make a hind cast (backward calculation) for 30 days in the whole Baltic Sea [10]. An oil spreading calculation is added to the currents, as well as oil evaporation, emulsification, sinking, stranding and dispersion. This powerful system today is in operational use in Sweden, Denmark, Finland, Poland, Estonia, Latvia, Lithuania and Russia [10].

Figs. 6-8 show three examples of oil spill drift modelling and a probability of the drift for specific points along main ship routes in the Gulf of Finland and southward of Gotland for July and August 2007. Fig.6a shows a drift of a virtual oil spill of 10 m3 during 48 hours, which was released on 23 July 2007 at a specific point (red square) of the ship route passing through the Gulf of Finland. The same numerical experiment was done daily from 1 July to 31 August 2007. Thus, basing on the compilation of 62 maps of oil spill drifts, we could construct Fig.6b and calculate a probability of oil spill drift. Fig. 6b shows that for this time period there is no impact of possible oil spills drift on the surrounding BSPAs along the coasts of Finland and Estonia, that are marked by blue and rose colors. This is explained by low wind speed and weak currents observed during July and August 2007. These weather conditions differ from those observed on 26 July – 15 August 2006, when a virtual oil spill could drift 33.5 n.m. during two days with a velocity up to 50 cm/s (Fig. 6c). Thus, potential releases of oil spills from the ships may represent a threat to seven protected areas located along the coasts of Finland and Estonia, as well as to the islands and coasts of these Baltic countries.

Fig.7 shows a significant impact of the selected point in the ship route southward of Gotland on the BSPA located eastward. In this case the area of potential pollution is very large. Fig.7c shows that only 37.3% of the BSPA surface will be free of pollution, but the other area can be polluted with different (calculated) probability. For instance, about 42.8% of the area (blue color) will be polluted with a probability of 5%, 9.4% - with a probability of 10%, etc.

Fig. 8a shows a simultaneous release of oil from a long part of the ship route located southward of Gotland (see real examples in Fig.3). Fig.8b shows a significant impact of oil pollution produced by this virtual oil release. Both BSPAs were subjected to oil pollution but at different degree that it was possible to estimate quantitatively with the help of the Seatrack Web model (Figs. 8c and 8d). For instance, about 60% of the first BSPA and 95% of the second one will be polluted with the indicated probability. In this case even the coastal zone of Gotland was potentially threatened by oil pollution with clearly calculated probability.

(a)

(b)

(c)

Figure 6. Modelling of oil spill drift in the Gulf of Finland: (a) shows oil spill drift on 23 July 2007; (b) shows probability (%) of oil spill drift calculated on the base of daily modelling at this point for real wind and currents conditions in July-August 2007; (c) shows probability (%) of oil spill drift calculated for 26 July – 15 August 2006. BSPAs are shown in blue, Important Bird Areas –

in rose colors.

(a)

(b)

(c)

Figure 7. Modelling of oil spill drift southward of Gotland: (a) shows oil spill drift on 12 July 2007; (b) shows probability (%) of oil spill drift calculated on the base of daily modelling at this point for real wind and currents conditions in July-August 2007; (c) shows the impact of this point in the ship route on

the BSPA.

(a)

(b)

(c) (d)

Fig. 8. Modelling of oil spill drift released from a long part of the ship route located southward of Gotland: (a) shows oil spill drift on 12 July 2007; (b)

shows probability (%) of oil spill drift calculated on the base of daily modelling at this line for real wind and currents conditions in July-August

2007; (c) and (d) show the impact of this part of the ship route on both BSPAs.

III. CONCLUSIONS

We can conclude that ENVISAT ASAR provides effective capabilities to monitor oil spills, in particular, in the Baltic Sea. Combined with satellite remote sensing of SST, suspended matter, sea level, chlorophyll concentration, mesoscale dynamics, wind and waves, this observational system, that includes also a Seatrack Web numerical model of SMHI, represents a powerful method for long-term monitoring of ecological state over the Baltic Sea, as well as smaller areas of particular interest, such as marine protected areas. Seatrack Web model allows also to assess quantitatively the ecological risks related to potential oil pollution of every MPAs in the Baltic Sea resulted from main ship routes, oil terminals and ports.

ACKNOWLEDGMENT

This study was supported by the INTAS “MOPED: Monitoring of Oil Pollution using Earth Observation Data: a multisensor, multiplatform approach” Project № 06-1000028-9091.

REFERENCES [1] Oceana. The other side of oil slick. The dumping of hydrocarbons from

ships into the seas and oceans of Europe. 2003, 31 p. [2] Oceana.. The EU fleet and chronic hydrocarbon contamination of the

oceans. 2004, 58 p.

[3] UNESCO. The integrated, strategic design plan for the coastal ocean observations module of the Global Ocean Observing System. GOOS Report N 125, IOC information Documents Series N 1183, 2003, 190 p.

[4] HELCOM, 2002 (www.helcom.fi). [5] A. Kostianoy, C. Ambjörn, and D. Soloviev, “Complex satellite

monitoring of the Baltic Sea marine protected areas”, Abstracts, European Symposium on Marine Protected Areasas a Tool for Fisheries, Management and Ecosystem Conservation, Murcia, Spain, 25-28 September 2007, p. 170.

[6] A.G. Kostianoy, S.A. Lebedev, K.Ts. Litovchenko, S.V. Stanichny, and O.E. Pichuzhkina, “Satellite remote sensing of oil spill pollution in the southeastern Baltic Sea”, Gayana, 2004, vol. 68, N 2, part 2, p. 327-332.

[7] A.G. Kostianoy, S.A. Lebedev, K.Ts. Litovchenko, S.V. Stanichny, and O.E. Pichuzhkina, “Oil spill monitoring in the Southeastern Baltic Sea”, Environmental Research, Engineering and Management, 2005, N 3(33), p. 73-79.

[8] A.G. Kostianoy, S.A. Lebedev, D.M. Soloviev, and O.E. Pichuzhkina, Satellite monitoring of the Southeastern Baltic Sea. Annual Report 2004. Lukoil-Kaliningradmorneft, Kaliningrad, 2005, 36 p.

[9] A.G. Kostianoy, K. Ts. Litovchenko, O. Yu. Lavrova, M.I. Mityagina, T.Yu. Bocharova, S.A. Lebedev, S.V. Stanichny, D.M. Soloviev, A.M. Sirota, and O.E. Pichuzhkina, “Operational satellite monitoring of oil spill pollution in the southeastern Baltic Sea: 18 months experience”, Environmental Research, Engineering and Management, 2006, N 4(38), p. 70-77.

[10] C. Ambjörn, “Seatrack Web, forecast of oil spills, the new version 2.0”, Abstracts, US/EU–Baltic Int. Symp. on “Integrated Ocean Observation Systems for Managing Global & Regional Ecosystems Using Marine Research, Monitoring & Technologies”, May 23-25, 2006, Klaipeda, Lithuania.