Detection of Oil Seepages in Oceans by Remote Sensing

Subhobroto Mazumder1 and Kalyan Kumar Saha2

1
Forward Base, Karaikal
2
Remote Sensing & Geomatics Div, KDMIPE Oil and Natural Gas Corporation Limited

Summary
The detection of oil seepage in oceans is of foremost importance from the exploration point of view since they are
the primary manifestations of any sort of oil accumulation beneath the ocean bottom and offer clues as to where oil deposits
may be located in ocean basin. In this paper, a general review of all the spectral regions in which remote sensing for the
detection of oil in oceans are carried out along with their advantages and disadvantages had been discussed. Of the different
passive sensors used for detection, the thermal infra red sensors can be designated as good sensors but for their poor
discriminatory capacity between oil and oil like objects. It functions on the basis of variation of the thermal inertia of the oil
with respect to the surrounding water which causes it to be visible to the thermal sensors. Among the active sensors, the laser
fluoro -sensors working on the ultraviolet wavelength are useful due to its unique capability of identifying oil in all backgrounds
like water, ice and snow. Most of the laser fluorosensors used for oil seepage detection uses a laser operating in the range
between 0.3 to 0.355 µm. The oil absorbs the ultraviolet light radiated by the fluorosensors and emits radiation in the visible
range of 0.4 to 0.65 µm with a sharp peak at 0.48 µm. This technique also helps to discriminate in between light refined
crude and heavy crude by comparing the fluorescent response of the oil with traces of various known oils. Another active
sensor that is fit especially for oil seepages is the radar which is not only powerful but also a cost-effective platform for
offshore seepage detection. This apart, it has a great detecting capacity and can work in the night and in all weather conditions
with wide swath coverage. Radar images map seepages which appear as flat patches of the surface due to the dampening
effect of the oil on the capillary wavelets, compared to the background waves. The information from the above spectral
regions used separately or together can serve as an effective tool for offshore oil exploration.

Introduction is currently flowing, driven by buoyancy from a sub-surface

origin’’ (Fig 1) (NPA, 2004). 80% of offshore oil exploration
The focus of oil exploration is moving inex-orably starts by searching for seeps. The U.S. National Academy
into ultradeep water, as exploitation of the accessi-ble parts of Sciences (NAS) completed studies of the sources, fates,
of the world’s continental shelves reaches a satura-tion point. and effects of crude oil in the marine environment. One
But as the industry moves into increasingly deeper waters component of these studies is natural oil seeps. In 2002 the
the costs of exploration also increases manifold. Satellite estimates of known crude-oil deposits in offshore North
remote sensing, after some 25 years of patient development,
America, particularly in the Gulf of Mexico and offshore
has now reached maturity as a respectable technology and
southern California, resulted in an estimate of global rate
can offer the industry an effective, low-cost, high coverage
service which is adaptable to any exploration environment. of crude-oil seepage of 160,000 metric tonnes/ annum for
Many enlightened major companies now use satellite remote North American waters.
sensing as a routine part of their initial exploration phase,
to unravel complex structure and geology ahead of
expensive fieldwork or seismic programmes. Satellite
Remote Sensing has the ability to image surface oil seeps
that originate by slow leakage from oil- and gas-filled traps.
Multitemporal satellite data over such seeps provide the
locations for follow-up surface sampling from which key
geochemical in-formation on the reservoired oil can be
obtained ahead of the drill.

Seeps and oil fields

A good definition of a seep is ‘’the surface
expression of a migration pathway, along which petroleum Fig 1: Mechanism of oil seepage from a reservoir below ocean surface

(1172)
 6th International Conference & Exposition on Petroleum Geophysics “Kolkata 2006”

Work published by BP and others in the early 90’s

(NPA, 2004) demonstrated that over 75% of the world’s
petroliferous basins contain surface seeps.

Most seeps represent tiny but detectable volumes

of oil and gas which are not significantly depleting the
reservoir. Exceptions would be in some recent onshore fold
and thrust belts where accumulations have either been
breached or redistributed to tertiary traps and where the link
between surface seeps and the leaking traps is more complex.
Such geology, however, is rarely encountered in offshore
basins so that problem does not arise. Confirmation of the
presence of seeps, therefore especially in offshore basins, is
positive and in the vast majority of cases is not indicative of Fig 2a. Surfacing oil pancakes (gas bubbles burst and lost to atmosphere
breached or depleted traps.

The knowledge that surface seepage has a direct

link to subsurface oil and gas accumulations is not new. In-
deed, seepage was the stimulus for the exploration drilling
by the pioneers of our industry as long ago as the 1860s in
Pennsylvania and in Azerbaijan. The subse-quent decades
spanning the late nineteenth century and the first half of the
twentieth century heralded the first discoveries. Surface
seepage was associated with Sumatra (1885), Texas (1901),
Oklahoma (1905), Per-sia (1908), Mexico (1910), Venezuela
(1922), Iraq (1927), east Texas (1930), Bahrain (1932), and
Kuwait (1938) (NPA, 2004).

Nevertheless, seeps are not ubiquitous and their

geographic distribution, as mapped by satellite, has helped
to define the present-day black oil play fairway and may Fig 2b. Coalescing oil pancakes forming large oil seepage slick (aerial
view)
also provide part of the explanation for the disappointing
results of the dry wells. secondly since in this region the difference in the spectral
response of oil and water is the maximum here. Also, it is
Offshore oil seep found that in the presence of solar radiation, many oils show
the phenomenon of Fluorescence in the ultraviolet part of
In the offshore, seeping oil and gas are often easier the spectra.
to detect due to the fact that oil is normally transported from
the sea-bed vent to the surface as oil-coated gas bubbles. At However, this type of detection systems has their
the surface, the gas bubble bursts and the oil remains on the own disadvantages. One is that, this is dependent entirely
surface as a thin oil film (Fig 2a ). In calm sea conditions, on the sun as its source of radiation and hence can work
these can often be viewed as beautiful, iridescent concentric only during daytime and that also in a clear weather. Another
shapes, typically 0.5 to 1 metre in diameter, known as ‘oil disadvantage is that, there is a great deal of selective
pancakes’. As seepage continues over time, these coalesce scattering by the atmosphere as the ultraviolet rays travel
to form larger slicks that are detectable from aircraft (Fig back from the target to the sensor and this in turn causes a
2b) and space] very low contrast in the signal. However, the effect of
scattering is found to be much less for images that are
Detection of oil in ocean in various spectrums acquired at an altitude of 1000m or less and hence a low
flying aircraft is used in this case rather than any space based
Detection in the ultraviolet region platform.

Ultraviolet images can be used to map films of oil To overcome the shortcomings of the passive
(as thin as 0.15 µm) firstly due to their high reflectivity and ultraviolet system, active ultraviolet systems are being used

(1173)
where laser fluorosensors illuminate the sea surface
immediately below the platform with a narrow beam of
ultraviolet energy and induce fluorescence in the oil. Most
of the laser fluorosensors used for oil slick detection uses a
laser operating in the range between 0.3 to 0.355 µm. The
oil absorbs the ultraviolet light radiated by the fluorosensors
and emits radiation in the visible range of 0.4 to 0.65 µm
with a sharp peak at 0.48 µm. This technique also helps to
discriminate in between light refined crude and heavy crude
by comparing the fluorescent response of the oil with traces
of various known oils. Also the power of the fluorescence
return permits the measurement of the oil thickness in the
range between 0.15 µm to 10 mm. This technique is used
for the detection of oil spill worldwide where the sensors
are mounted on a low flying aircraft but it has a drawback
that it can be used only in clear weather. Another related
instrument is the Fraunhofer Line Discriminator which is a
passive fluorosensor using solar irradiance instead of laser.
However its low signal to noise ratio doesn’t make it much
effective.

Detection in the visible region

Fig 3: A. Natural oil slicks in the Gulf of Mexico. A west to east echo
The visible region does not provide too much of a sounder shows oil and gas rising from two vents of sea floor. B.
contrast between water and oil and it needs thorough Oil drops reach the surface and sunglint is enhanced in the area
of floating oil. C. Photograph from space shuttle shows sunglints
processing to yield better results. An idea of this can be had from slicks
from the closeness of the DN values of oil and water in the
bands 1 to 3 of LANDSAT TM sensor rate at which heat will pass through the material. Water
usually has a low thermal conductivity. The thermal
TM Band Spectral Range DN of oil DN of water conductivity depends on the sea state and the surface wind
but it is generally found that the thermal conductivity of oil
4 0.76 to 0. 90 µm 13 8
is greater than water. Thermal Inertia (P) is the property of
5 1.55 to 1.75 µm 7-50 5
a material to resist temperature changes. When thermal
6 2.08 to 2.35 µm 22 3
inertia is high, materials show a resistance towards change
(Almond , 1999) in temperature resulting in low temperature stability.
Materials with a low thermal inertia reach a high surface
The effect of wind causes surface waves in ocean temperature at day time and at night time cool to a relatively
which causes the surface of oil to show a high reflectance low temperature, vice versa for materials with high thermal
than the surrounding water in moderate to high elevation, a inertia (Sabins, 1978).
phenomenon called Sunglint (Fig 3). This makes it possible
to some extent to detect the oil seepages in ocean water. The sea water is found to have a higher thermal
inertia than oil and this causes the oil film to show a larger
Detection in the Thermal Infrared Region variation of diurnal temperature than the surrounding water.
This means that it will be visible to the thermal sensors
The oil is found to show a number of thermal during afternoon and due to its high surface temperature, it
properties that distinguish it from water. These include will act as a black body absorbing heat and becoming
thermal capacity, thermal conductivity and thermal inertia. warmer than the surrounding sea water. Again during the
Thermal Capacity ( c ) is the ability of a material to store night the oil body looses heat faster than the surrounding
heat. In case of oil spills the thermal capacity of sea water water and this causes it to be cooler than the surrounding
is found to be twice that of the overlying oil (Asanuma, water mass. However thin slicks are not thermally
1986). The Thermal Conductivity (K) is the measure of the distinguishable since the temperature pattern of oil is similar

(1174)
 6th International Conference & Exposition on Petroleum Geophysics “Kolkata 2006”

to that of water as the underlying water heats the oil here. slicks on radar, local surface wind speeds need to be ideally
The thermal properties of oil may also change due to the between ~2m/s and ~5m/s, although thicker slicks can be
mixing of oil with water or due to the weathering of the oil observed at wind speed up to ~10m/s. The mm-cm
film. wavelength capillary and gravity waves develop in this wind
speed range, and area “available” for dampening by oil. If
This method has been used for the detection of oil the sea surface is too smooth , at wind speeds below ~2 m/
spills only but it has its own disadvantages. Often cool water s, the sea surface is becalmed (i.e capillary waves have not
currents of oceans present a similar image as oil spills. This developed) and seepage-slicks can be obscured or
problem can be minimized using information from the “camouflaged” by dense natural slicks. At wind speeds from
ultraviolet spectral ranges along with that of the thermal ~5 m/s, the local wind shear rises to a point where the
information. Clouds also act as a barrier to thermal sensing, capillary and gravity waves are destroyed, and slicks are
as they are good absorbers of the thermal radiation. dispersed mechanically and are not observed by radar. If
the higher wind speed persists, a swell develops in which
Oil Spill detection in the microwave region body waves are aligned in the wind direction (Srivastava,
2005). Slicks can develop on these body waves only if the
Since 1992, a powerful and cost-effective platform local wind shear is low; this is the case for long traveled
for offshore seepage detection has emerged. This is satellite- (mature) swell generated outside the region of observation.
borne active microwave radar or SAR (Synthetic Aperture Atleast two wind-complaint (dual images) radar scenes are
Radar). This is due to their ability to image surface oil seeps required for radar interpretation.
remotely with wide swath coverage (typically 100 x 100km
scenes for ERS and 165 x 165kms for Radarsat Wide 1) L Slick interpretation
and at low cost. Moreover, satellite data is free skies and is
being continuously acquired, thus providing multi-temporal Slicks are interpreted on the SAR imagery as dark
satellite data over any area of the globe. Such repeat seeps patches. The pollution slicks are thicker than the seepage
provide the location for follow-up surface sampling from slicks. The seepage slicks are ultra thin films. They do repeat
which key geochemical information on the reservoired oil in time, but often not observed and they often form in
can be obtained ahead of the drill. SAR satellites scan the clusters because seepage vents are never singular. (Figure
oceans continuously on fixed polar orbits. They have 4)
advantages over optical satellite systems, such as Landsat
TM and airborne systems in that they observe night and L Slick categorization
day and penetrate cloud cover. SAR creates images of the
sea surface detailing its morphology. Radar images map Satellite SAR systems are very effective at
slicks (flat patches of the surface) that can be related by observing slicks, both man-made and natural, but very few
analysis to petroleum seepage. slicks result from hydrocarbon seepage—very rarely more
than 5% of the total number of slicks detected (NPA, 2004).
Offshore seeps- detection by SAR technology Only these seepage slicks are of exploration interest. A
consistent and systematic analysis scheme to discriminate
L Weather dependence seepage slicks from other slicks has been developed (Figure
5a and b). Three main categories of slicks are distinguished:
In offshore Basins, oil seeps from reservoir can seepage slicks, pollution slicks, and natural film slicks
reach the sea surface, usually in the form of oil coated gas (Williams, 2002). Natural film slicks are formed from an
bubbles, and then form slicks identifiable from satellite. ultrathin layer of long-chain organic molecules—derived
This is due to the dampening effect of the oil on the capillary largely from decayed plankton—and appear on low wind-
wavelets, which produces an area of relative calm compared speed images.
to the background waves. The satellite data to be chosen
should be screened for weather, mainly for wind speed, as Some of the parameters used to categorize and
this parameter has a direct implication on slicks (Srivastava, discriminate seepage slicks (Williams, 2002):
2005). Wind shear controls the roughness and morphology
of the sea surface and wind speed is critical to the remote Size: The minimum resolvable seepage slick is ~100–150
detection of slicks. In order to observe ultra thin seepage m long, dependent on sea state and sensor. Slicks smaller
than this will not be resolved by SAR.

(1175)
 Direction of flows (streaming direction): Seepage slicks
will conform to dominant wind and current directions (and/
or tidal effects). Fresh pollution slicks, especially from ships,
are typically straight and have a distinct featheredge where
older material is blown by wind.

Context: This is either geologic (i.e., location over plausible

migration conduits) or geographic (i.e., location within
shipping lanes or in areas of oil production).

Backscatter reduction: Oil slicks (seeps and pollution)

generally have a greater contrast and edge enhancement (i.e.,
sharpness) than natural films have, but this property is
Fig 4: Seepage Slicks repeating in time in Gulf of Mexico The coloured dependent on wind speed.
polygons outline slicks mapped on satellite radar acquired on
dates different from the image shown. The region shown has been
sampled four times over a three-year span. (NPA Group, 2004) Edge characteristics (i.e., sharpness) and relation to wave
facets, current, and wind: These details are critical to the
analysis because they depend on the slick thickness and
material.

Repetition: A repeating emission point is shown on

successive images. This is the most compelling parameter
but is restricted to the leakiest basin types (episodic seepage
is a feature of most seepage and in most offshore basins).

Ocean features: These features have to be analyzed and

accounted for in the slick analysis.

The confidence with which a slick can be

categorized on weather-compliant images is derived from
considering all the above parameters. Seepage slicks are
ranked 1, 2, or 3, based on these confidence limits (Figure
6 a, b, c).

L Seep sampling

The most valuable by-products of satellite seep

detection is providing sea-surface locations for follow-up
slick sampling from which crucial geochemical information
of the reservoired oil can be obtained. Although some slicks
may be partly biodegraded, a sufficient concentration of
biomarkers will be present to enable the seep to be
Fig 5 : Slick Categorization (a) Parameters and principles; (b) Slick
characterized (by GC-MS, carbon isotopes, etc.) and for
categorization template, showing confidence and overlap of oil-source correlations to be established. The satellite
slick categories Williams, 2002 records the position of a slick to within ± ~300 m. Seepage
slicks are located at sea using GPS, and are sampled from
Location: Slicks located in known or predicted sedimentary small boats (Figure 2a), ideally in tandem with a light aircraft
basins (as opposed to areas of oceanic crust or basement to help position the boat over the active seep.
terrains) are potentially of interest.
Conclusion
Morphology: Seeps are generally elongated, with high
aspect ratios (length to width), and with a recognizable If applied correctly, remote sensing serves as an
emission point. important tool that provides the early detection of the oil

(1176)
 6th International Conference & Exposition on Petroleum Geophysics “Kolkata 2006”

Askne. J, 2002, Oil Spill Detection, Radar and Remote Sensing ,

Page F.36
Bartsch. N, Gruner. K, Keydel. W, Witte. F, 1987, Contributions
to Oil Spill Detection and Analysis with Radar and
Microwave Radiometry; Results of Archimedes – II
Campaign, IEEE Transactions of Geoscience and
Remote Sensing, 25, pp 677-689
Borzelli. G, Ciappa. A,Ulivieri. C, Anyonelli. G, Laenve. G, 1996,
A New Perspective of Oil Slick Detection from Space
by NOAA Satellites. International Journal of Remote
Sensing, 4, pp 1279-1292
Blume. H. J, Huhnerfuss. H, Alpers. W, 1983, Variation of
Microwave Brightness Temperature of Sea Surface
Covered with Mineral and Monomolecular Oil Films,
IEEE Transactions of Geoscience and Remote Sensing,
21, pp 295-299
Fig 6. (a) Rank 1: nested multiple source slicks. Image is 30km across Camagni. P, Colombo. G, Koechler. C, Pedrini. C, Omnetto. N,
(b) Example of Rank 2 (remnant) pollution slick. (Fragmented Rossi. G, 1998, Diagnostics of Oil Pollution by Laser
pollution from ship, six days after discharge) (c) Rank 3: small
Induced Fluorescence, IEEE Transactions of Geoscience
cluster slicks are ~ 120m long (NPA Group, 2004)
and Remote Sensing, 26, pp 22-25
Okoro. C, 2004, Pipeline Vandalisation and Oil Spillage Monitoring
seepages and oil spills, provides size estimates, and help using Remote Sensing : A Case for Nigeria, http://
predict the movement of the slicks and possibly the nature nigerianewsnow.com/ News/ October04/ Oil%
of the oil. Though the entire electromagnetic spectrum 20Spillage% 20Monitoring% 20using% 20Remote%
beginning from the ultraviolet can be utilized in remote 20Sensing.htm
sensing of oil, the best results are obtained in the ultraviolet, Clouts. E. A, 1989, Spectra Reflectance Properties of
thermal and microwave regions of the spectrum. Of these, Hydrocarbons: Remote Sensing Implications, Science,
the active microwave sensors come near to the description 245, pp 165-168
of an ideal remote sensing system with large ground Cross. A.M, 1992, Monitoring Marine Oil Spills using AVHRR
coverage, round the clock operations under all weather data observations of the coast of Kuwait and Saudi
conditions, cost effective and are useful for detection of Arabia during January 1991, International Journal of
Remote Sensing, 245, pp 165-168
both oil seepages and oil spills. These technology can offer
Engelhardt. F. R, 1999, Remote Sensing of Oil Spill Detection and
the oil industry an effective, low-cost technique for reducing Response, Pure and Applied Chemistry, 71, pp 103-
risk in high cost exploration environment. 111
Espedal. H. A, Wahl. T, 1999, Satellite SAR Oil Spill Detection
Acknowledgements using Wind History information, International Journal
of Remote Sensing, 20, pp 49-65
Authors are thankful to ED- Head KDMIPE and Fingas. M.F, Brown. C. E, 2000, Review of Oil Spill Remote
GM, Head Forward Base, Karaikal for their permission to Sensing presented at Spillcon 2000, Darwin, Australia,
publish this paper. Thanks are also due to Prof Dr.R.P.Singh, Environment Science Division, Environment
and Asstt Prof Dr. Onkar Dixit, Deptt. Of Civil Engineering, Technology Centre, Canada, www.gcmd.nasa.gov/
IIT Kanpur and also to Shri Kanchan Maity, Research conferences/rsce.html
Fujita. M, Masuko. H, Yoshikado. S, Okamoto. K, Inomata. H,
Fellow, Dept of Earth Sciences, University of South Carolina
Fugono. N, 1986, SIR-B Experiments in Japan; Sensor
for their help and guidance in preparing this paper. Calibration and Oil Spill Detection over Ocean, IEEE
Transactions of Geoscience and Remote Sensing, 24,
Views expressed in this paper are that of the author pp 567-574
only and may not necessarily be of ONGC. Kvenvolden .K. A. and Cooper. C. K. 2001, Assessing the Rate of
Natural Seepage of Crude Oil into the Marine
References Environment
Legg. C. A, 1991, The Arabian Gulf Oil Slick, January and February,
Almond. S, 2000, The Remote Sensing of Oil Slicks from Satellite 1991, International Journal of Remote Sensing, 12, pp
Platforms 1795-1796
Andersen. J, 1995, Oil Spill monitoring by the use of Radar and Lillesand T.M and Kiefer W. A, Remote Sensing and Image
Aircraft Interpretation, 2nd Edition, John Wiley and Sons.

(1177)
Lundin. E and Torstensson. K. 2002, Essay in Earth System Sabins F. F, 1978, Remote Sensing: principles and Interpretation,
Science Techniques Freeman, pp 334-392
Mansor. S.B, Poy T.T, 2004, Remote Sensing and GIS Application Srivastava. K. S and Arulmani. M, 2005, Radars Click slicks to
In Oil Spills Risk Assessment evaluate Hydrocarbons, Geospatial today, vol 3, Issue
NPA Group, 2004, Key Features of OBS Technology, Crockham 5, pp 55-57.
Park, Edenbridge, Kent, TN8 6SR, England Shanko Barry, 2003, Off Shore Oil Detection
Palmer .D, Borstad G.A, Boxall S.R, 1994, Airborne Multispectral Williams. A and Huntley .A, 1998, Oil From Space - ‘Detecting
Remote Sensing of the January 1994 Shetlands Oil Spill, the Sleeping Giants of the Deep- Water Caspian by
www.borstad.com. Satellite’- Petrex-98
Schanda. E, 1996, Physical fundamentals of Remote Sensing, Williams. A, 2000, The role of Satellite exploration in the search
Springer-Verlag, pp 29-30 and pp 124-129. for new petroleum reserves in South Asia, SPE-PAPG
Schiewe. J, 2003, Remote Sensing And Aerial Surveillance, An Annual Conference, Islamabad, Nov. 9-10, 2000
Undergraduate Report Submitted in Partial Fulfillment Williams, A., and G. Lawrence, 2002, The role of satellite seep
of the Requirements for the Degree of Honours Bachelor detection in exploring the South Atlantic’s ultradeep
of Science in Forestry, Faculty of Forestry and the water, AAPG Studies in Geology No. 48 and SEG
Forest Environment, Lakehead University. Geophysical References Series No. 11, p. 327–344.

(1178)