Earth observation data is an ideal medium for the analysis
of large scale routing, particularly in areas poorly mapped
by traditional means where the latest ground cover information
is required. Satellite data is acquired over large areas,
therefore keeping down equivalent costs per square kilometre.
There are two distinct approaches for pipeline routing,
the first is designed for routing through largescale remote
terrain, the second for routing through populated regions.
Remote Terrain
Satellite imagery provides a low-cost means for the selection
procedure in pipeline routing. Imagery brings out a wealth
of topographic detail, providing an essential first step
to narrowing down the selection of possible routes for further
investigation, in addition to the identification of route
sections that might require additional treatment and protection
of materials.
Populated Regions
The routing of an oil pipeline through inhabited land requires
a detailed survey of land parcel divisions in order to compensate
landowners for land affected by the final routing. The location
of drainage canals, settlements and buildings also control
routing decisions.
Other sources of the required data would be either existing
mapping or by undertaking a traditional ground survey. Suitable
mapping is frequently either non-existent, classified, lacking
the required information, or out-of-date. Some existing,
relatively recent survey data is sometimes available, but
even more recently, drainage channels and canals can have
been constructed.
Existing Method
The data compiled prior to the acquisition
of satellite imagery is obtained from a survey using traditional
ground survey instruments and procedures, namely, triangulation
with theodolites and EDMs. This procedure for dense agricultural
land is time consuming as access is often hampered due
to numerous canals and drainage channels crossing the area.
Several hours are required to survey a few land parcels.
Errors in measuring or recording a measurement frequently
result in errors of further measurements. The ground team
may also make assumptions or estimate results in areas
of
difficult access.
The survey measurements are then manually entered into a
CAD package on a computer, ready for further analysis.
The most suitable satellite imagery is the
data with the highest spatial resolution available. Launched
in late 1999, the IKONOS-2 satellite now provides 1m panchromatic
(black and white) data and 4m multispectral (colour) data,
paving the way for highly detailed analyses. Another image
source available for many regions of the world is Russian
KVR-1000 imagery with a 2m ground resolution.
Extracting Information
The high resolution data is acquired
in a digital format. The data is then entered into an image
processing / GIS computer software package. The imagery
is then referenced to a map projection grid using 5-10, reliable,
well defined ground control points with a good 2 dimensional
spread over the study area.
The existing survey data was overlain on the imagery. With
the two datasets together, it was possible to:
Identify and correct systematic errors
Locate other errors
and changes that required ground verification
Digitise
further required land division boundaries, which were
verified on the ground for additional changes since
the image acquisition date.
Ground based survey was then used for final updates of the
most recent changes.
Ground survey procedures were found
to frequently be unreliable. Numerous changes occur which
are often traditionally only identified by fully repeating
a survey. Satellite imagery provides a rapid means of visually
identifying and monitoring changes.
Substantial traditional survey costs can be almost completely
replaced with an alternative image based procedure costing
around $20-30K (including imagery, IP/GIS software and digitising
on computer).
Constraints on traditional procedure include:
Access
Cost of revision
Survey time
Constraints on satellite imagery frequently include:
Interpretation
of imagery (requires field verification)
Other sites e.g.
building type, special land assignment (e.g. graveyards)
require field verification along line
of route.
The routing of a pipeline or other transportation link
through densely agricultural land requires a detailed
survey of land
parcel divisions in order to compensate landowners
for land affected by the final routing. Overlaying an
existing
survey
on high-resolution satellite imagery enables location
of errors and changes that require ground verification.
Further
required land division boundaries could then be digitised
from the imagery and verified on the ground.
The
routing of a pipeline through areas of intensive agriculture
may require detailed survey of land parcel divisions to
minimise compensation commitments. The pipeline route in
this example
connects two oil production areas. Surveys acquired at
different times and to different specifications are overlain
in green
and orange respectively, on 2m imagery.
The extract shown right reveals the inaccuracies in the
cadastral information previously recorded when compared against
the
land parcels that actually exist, as shown by the satellite
imagery. This confirms the value of information derived from
up-to-date satellite sources over large areas.
