Voyage to the Gulf of Carpentaria
[Introduction] [The
Voyage] [The Voyagers] [The
Vessel]
[Ship
to Shore] [Web Links]
Report 7 – Outwitting Cyclone Ingrid
11 March 2005
Lat |
S 15º 40' |
Long |
E 138º 5' |
Sampling operations were moved nearer to the shelter of Groote
Eylandt on Thursday as Cyclone Ingrid advanced west across the Gulf
of Carpentaria. Quantitative ecologist Wayne Rochester explains how
he 'adaptively updates' the voyage track and associated work plan.
We have just completed sampling at our first 42 sampling stations in
a region north of Mornington Island. Our cruise track for Mornington
looks like spaghetti and was updated four times in the seven days we
took to complete it. Are we nuts?
Actually, there is method in our apparent madness. So far the results
have been positive: we are on schedule with sampling and the response
of the skipper and crew has for the most part been bemusement rather
than irritation.
The spaghetti and the updates arise from our need for an optimised,
but adaptive, cruise track to accommodate the survey design, operational
requirements and hiccups.
The spaghetti arises because we require a track with a reasonably consistent
11 nautical mile steaming distance between sampling sites, rather than
one following the shortest path.
This is for operational reasons such as keeping the ship moving, sorting
the catch and maintaining a rhythm onboard. As a bonus, it also scatters
the samples for nearby sites in time.
The need to adjust the track during the survey arises because our sampling
sites are stratified by day and night in addition to trawling intensity.
Whenever we complete more or less samples than planned, day sites get
pushed into the night or vice versa. We must then recalculate the cruise
track to ensure that we visit day sites in the daytime and night sites
at night.
The cruise track is calculated with a simulated annealing optimiser.
Importantly, the optimiser can be applied to a partly completed cruise
track so changes can be made during the voyage.
The optimiser is integrated with the operational geographic information
system (GIS) that is used to design and monitor the survey. So we can
adaptively update the sampling site locations and cruise track in real
time in response to operational hiccups or new knowledge, and print
tables and maps of the updated track for distribution to the skipper,
crew and scientists onboard.
About Wayne Rochester
More voyage reports
Updated:
29/03/07