Acquisition of accurate water depths and identification
of areas of potential hazards in the area of a proposed offshore
seismic programme can greatly reduce risk of damage to expensive
equipment and vessels. With both time and cost benefits,
satellite data is able to provide this valuable information
with a density of effective measurements difficult to obtain
with conventional depth soundings. The imagery is used to
derive contour charts and enhanced colour image maps, which
then enable determination of the survey feasibility and strategy.
In the following case study, the customer wished to plan
a seismic survey in an area offshore Cuba. There were no
reliable depth charts for the area and the region was known
to contain numerous coral reefs. Time was limited before
the survey needed to proceed. The vessels required water
depths sufficient enough to align for each survey track and
also a turning area of several kilometres outside the survey
zone which was aligned closely to the coast.
The survey area was approximately 45 000km². The optimum
data source for this size area was Landsat TM imagery. This
optical multispectral data was used to model the exponential
decay of the transmission of solar radiation (of the blue
waveband - Band 1), to and from the sea floor. The calibration
of the decay model was achieved by using a best fit to the
existing sparse depth sounding control points. This enabled
effects such as sea floor reflectance and sea water clarity
to be accounted for. Additional control for the model was
obtained by measuring the reflectance from known shallow
water close to the shoreline and from offshore areas where
the depth was such that the reflectance contained a negligible
component from the sea floor.
The results from such a satellite survey depend greatly
on factors such as the water clarity, solar incidence angle,
sea floor composition and atmospheric conditions. Depth resolution
decreases as depth increases.
Under favourable conditions, which are usually found close
to the arid areas of the world (where there is little river
flow carrying sediments, micro-organism levels are low, solar
elevation is high and atmospheric haze is minimal), depth
penetration of around 25 metres and depth resolutions of
2-3m at 15-20m can be expected. In the area around Cuba,
a depth penetration of 20m with a resolution of 2-3m at 10-15m
deep was found to be possible.
Using other Landsat TM wavebands, it is possible to detect
the location of sea floor vegetation. The bathymetric results
from areas of vegetation are less reliable. Thus, the imagery
highlighted areas where additional caution was required.
Delivered products to the customer included: photographic
imagery enhanced separately for the land and offshore areas,
and contour depth charts with hazards and areas of caution
marked.
Evaluation
The customer felt that satellite data provided a means of
undertaking a survey that would otherwise have required the
time consuming and costly deployment of a survey vessel into
the region.
The short lead time required for the satellite data acquisition
and processing meant the seismic survey and exploration operations
could be carried out over a significantly reduced period.
Apart from the definition of the required area and product
information, the customer's organisation time was relatively
short. The time period to produce the charts was dramatically
less than the pre-survey reconnaissance that would otherwise
have been necessary. In addition, order of magnitude improvements
were made to the cost and final detail.
Satellite imagery provided means to produce a bathymetric
chart of much improved reliability compared to the existing
older charts. Locations of hazardous coral reefs were readily
identified and the picture element (pixel) size of the image
data ensured hazards, which may not have been covered by
bathymetry from a survey vessel, were identified.
The products derived above cost a total of £25000,
with the data cost of £15000 accounting for the largest
proportion of the total project cost. Of the total mapping
project duration of 6 weeks, image acquisition accounted
for 3 weeks. The cost for a bathymetic survey vessel to acquire
a similar level of information would have involved costs
an order of magnitude greater and taken many more weeks to
arrange and complete.
