Detection capabilities and measurement accuracies of State-of-the-Art oil
pollution monitoring sensors (1998)
Benny (Ben) M. Sorensen
BMS International
Skodsborgvej 222, 2850 Naerum, Denmark
ABSTRACT
Operational airborne oil pollution surveillance started in Sweden in the late seventies
using a SLAR and an IR/UV scanner. Since then, many countries in Europe and North America
have invested in such systems for routine surveillance.. A few countries have expanded the
sensor package to include a microwave radiometer and a single country is also using a
laserfluorosensor. The advantages and shortcomings of these sensors have been known for
more than 20 years. This paper discribes the status of their detection capabilities and
measurement accuracies based on a worldwide survey.
1. INTRODUCTION
World experts on oil pollution surveillance agree that a sensor package, as a minimum,
must include a Side-Looking Airborne Radar (SLAR) and an infrared/ultraviolet (IR/UV)
scanner and some cameras. The SLAR is used to detect oil on the sea surface and the
scanner to identify oil and map it. To some extent, the scanner can also be used to make
qualitative volume distribution of oil on the sea surface. A few countries have chosen to
include a microwave radiometer (MWR) for quantitative thickness/volume determination of
"thicker" layers (> 0.1 mm). Only a single country is using a
laserfluorosensor for quantitative thickness/volume determination of "thin"
layers (<0.01 mm) and for characterization of oil.
The capabilities and deficiencies of these sensors for oil pollution surveillance have
been known for many years. The capability of the SLAR to detect oil on the sea surface has
been known since the mid-seventies and so has the capability of the IR/UV scanner. Already
in 1972 did J. Hollinger demonstrate that a microwave radiometer could quantify
thicknesses and volumes of oil on the sea surface. The capabilities of the nitrogen and
excimer lasers in relation to oil have been known since the late seventies. Thus after
more than 20 years of operational oil pollution surveillance, it is appropriate to
establish how accurate the sensors are for the detection of oil and for determination of
thicknesses and volumes of oil on the sea surface.
2. THE HISTORY
Since the early eighties, many experiments and exercises have been conducted over known
types and quantities of oil to determine the measurement accuracies using airborne sensors
and manual observations. The countries around the North Sea and Sweden have signed the
Bonn Agreement which is a convention for protection of the North Sea against oil
pollution. For many years, the member states have carried out experiments where as many
surveillance aircraft as possible fly over various types of known quantities of oil to
compare the observations. It is not unusual that the highest and lowest values of observed
oil from these aircraft vary by a factor 10 or more. Experiments have also been carried
out to determine the boundary conditions for observing oil on the sea surface. This has
shown that in near range, i.e. at 2-4 km distance from the aircraft, 30 liters of oil can
be detected with SLAR at Beaufort 3-4, while 60 liters can be detected at Beaufort 5 at a
somewhat larger range.
3. THE SCENARIO
Discussions continue at various fora, such as the EU, IMO, Bonn Agreement countries and
Helsinki Convention signatories to determine which sensors and techniques, and what
accuracies are required to produce sufficient evidence to court in cases concerning
alleged discharge of oil at sea. In a juridical system, this is important information for
both the culprits and the authorities.
Some of the most important questions to be addressed are:
(a) What are the boundary conditions for detection of oil on the sea surface from
aircraft, e.g. sea state vs range, inner and outer range limits at any condition, radar
signal depression angle, etc.?
(b) Can mineral oil be unambiguously identified from aircraft, and if so, what are the
means accepted by courts of law for such identification?
(c) Is the oil quantification accuracy an important issue, and if so, what is the
acceptable accuracy?
(d) How do all the general oil types vary in terms of volume after discharge into the sea,
considering the many variable parameters, such as sea state, meteorological conditions,
light conditions, sea water temperature, etc?
(e) In areas where an oil spill of any volume is illegal, what is the minimum requirement
as evidence to court to indict an alleged spiller, e.g. how is it proven that a vessel is
discharging oil rather than just passing through the oil?
BMS International has investigated the status quo with regard to points (a)-(d). We
therefore conducted a worldwide survey among scientists, hardware developers, surveillance
program operators and sensor manufacturers, all of whom are internationally known for
their research and development in relation to airborne oil pollution surveillance and
monitoring.
4. THE SURVEY
Out of 34 organizations, companies and individuals identified as important contributors to
the field of airborne oil pollution monitoring during the last 20 years, 32 were located.
They received the questionnaire with the following queries:
- Which type of sensor or other equipment have you developed, manufactured or tested?
- Do you, or does your organization, have any quantitative results from the laboratory or
flight tests in terms of oil area coverage, oil layer thickness or oil volume on the sea
surface or below the sea surface using the sensor(s) you have developed, manufactured or
tested? Please include results from both laboratory and flight tests.
- If you are or have been involved in sensor development, can you refer to any reports,
papers or other publications which elaborate on accuracies with regard to oil coverage,
oil thickness and/or volume on the sea surface or below the sea surface?
- If you are a sensor manufacturer, are you prepared to guarantee accuracies achievable
with your equipment in terms of oil coverage, oil thickness and/or oil volume on the sea
surface or below the sea surface? Please, specify.
- Have you developed or can you refer to any other ways a means to determine oil area
coverage, thickness or volume from aircraft including sensors not mentioned above, visual
observations through colour codes, etc? If yes, please specify.
- Is remote sensing technology accepted in your country as Evidence to Court in cases
against polluters accused of discharging oil into your territorial waters? If so, what are
the minimum requirements in terms of sensors and their capabilities accepted by the court?
24 replies came from 5 developers, 4 manufacturers and 15 users/experimenters. The
developers and manufacturers have also conducted tests or participated in tests of their
own equipment.
None of the replies gave details of the detection capabilities of the radars. So far,
scanners are only useful for qualitative (relative) thickness measurements of oil. The
only two sensors used operationally to quantify oil on the sea surface are the MWRs and
the LFS.
Three commercial brands of MWR are used for routine surveillance. They are all
manufactured and used in Europe. The oldest MWR is a single frequency system which does
not function well much of the time and the manufacturer has not been prepared to make any
statements about accuracies. Another dual-frequency radiometer has shown accuracies from
an experiment as follows: 78% within the first hour after discharge of oil. A few minutes
later, the sensor recorded only 40% of the oil and yet another hour later only 23% of the
oil. Finally, after 5 hours, the recorded volume was again 41% of the discharged oil. The
gradual decrease of measured volume may partly be explained by evaporation and possibly
also by spreading, so that only a part of the oil was inside the field of view of the
radiometer - in particular because the flight pass with the 23% recording was flown at 500
feet lower altitude than the previous passes. The difference between the first two
measurements where the recorded volume increases almost by a factor of two is more likely
due to system error.
So far, the three-frequency microwave radiometer is only used by one surveillance
group. The manufacturer guarantees that it performs with an accuracy of 50%. This
statement is based on flight experiments where a known volume of mineral oil has been
discharged into the sea. The results are not associated with any comments about type of
oil, variation in flight altitude, weather conditions or time and frequency of the
overflights after the discharge - all factors which may influence the result. Therefore
the guarantee must be accepted with some reservations.
Only one laserfluorosensor is in operational use in the world today. Also here, the
manufacturer issues a guarantee of a 50% accuracy. As for the MWR, this accuracy is given
without any additional information about weather conditions, flight altitude etc.
Finally, among Bonn Agreement operators, there is a consensus that it is possible to
determine oil thicknesses on the sea surface through a colour code in the thickness range
0.05 - approx. 1-3 micrometers. However, this is purely a visual daylight observation
which depends on look angle, cloud cover and a number of other parameters.
5. CONCLUSION
After 20 years of routine oil pollution surveillance, the scientific community,
manufacturers and operators have made very little progress in establishing the boundary
conditions for oil detection and for quantitative determination of oil on the sea surface.
It is concluded that the state-of-the-art in airborne pollution surveillance does not
allow for any dependable statements about the detection capabilities and quantitative
measurements from the sensors currently in use.
For the research community the conclusion is a message to continue with much more
experimental work on the existing sensors as their capabilities are too poorly understood
for operational airborne pollution surveillance.
Comments are welcomed directly
to Ben Sorensen: bms.int@it.dk
