Data Overview:

Data were collected by the Mar Mostro during the around-the-world Volvo Ocean
Race (VOR) 2011-2012 (http://www.volvooceanrace.com/en/home.html).  The Mar
Mostro is a 21.5m sail-powered racing yacht that was manned by the Puma Ocean
Racing team (sponsored by the PUMA sports apparel company) and powered by Berg
Propulsion of Sweden.

Data were provided to the Center for Ocean-Atmospheric Prediction Studies at the
Florida State University by Robert Hopkins, Jr. who served as the performance
analyst and coach for the Mar Mostro during the 2011-2012 VOR. Co-author for the
data was Tom Addis, Navigator and crew member for the team.

Race Legs:

Leg no.	Origin/Date			Terminus/Date
--------------------------------------------------------------------
1	Alicante, Spain      06Nov'11	*Tristan de Cunha   25Nov'11
2a	Cape Town, S. Africa 11Dec'11	Male, Maldives      27Dec'11
2b	Sharjah, UAE         04Jan'12	Abu Dhabi, UAE      04Jan'12
3a	Abu Dhabi, UAE       14Jan'12	Sharjah, UAE        14Jan'12
3b	Maldives             22Jan'12	Sanya, China        04Feb'12
4	Sanya, China         19Feb'12	Auckland, NZ        17Mar'12
5	Auckland, NZ         17Mar'12	Itajai, Brazil      06Apr'12
6	Itajai, Brazil       22Apr'12	Miami, FL USA       09May'12
7	Miami, FL USA        20May'12	Lisbon, Portugal    31May'12
8	Lisbon, Portugal     10Jun'12	(**Lorient, France) 14Jun'12
(9***	Lorient, France	     01Jul’12	Galway, Ireland     03Jul’12)
(10***	Galway, Ireland	     08Jul’12	Hono Island, Sweden 12Jul’12)

* Note that a broken mast forced a deep water pickup off the island of Tristan
de Cunha during Leg 1; this was not the planned terminus (likely Cape Town)	

** Note that Leg 8 officially ended at Lorient, France but our analysis omits
a terminating portion of this leg (owing to HYCOM sub-domain limitations)

*** Note that Legs 9 and 10 were omitted from our analysis (owing to HYCOM sub-domain limitations) and leg 10 was not part of the official Volvo race.

Data/Sensor Details:

The Nortek velocity logger layed at 4.5m depth at rest (see attached schematic
from Robert).  The following excerpts from old emails from Robert provide
additional detail regarding the Nortek and the GPS:

"Mar Mostro, our Volvo 70, is equipped with a unique Doppler velocity log (DVL)
from Nortek, the first to be used on a racing yacht. The DVL provides velocities
along and across the boat's longitudinal axis. The longitudinal velocity is used
as boatspeed while the atn(Vy/Vx) provides the leeway angle in degrees, which we
add to the boat's heading to arrive at course through the water.  Heading is
provided by an Ixsea Hydrins IMU, which also provides pitch and roll data to
transform from instrument to earth coordinates. The DVL is mounted on the bottom
of the bulb which is attached to a canting keel. The keel cant angle is also an
input to the coordinate transforms."

and also:

"We have two GPS’s sampling at 10 Hz and 5 Hz. We started with the fancy 10 Hz
model and ended with the 5 Hz model for reliability reasons. I refer to the raw
position output in the NMEA stream. I think the COG and SOG values are output at
only 1 sec intervals. In any case, the GPS output of COG and SOG is damped
inside the unit before we see the values, then we add additional damping to it
simply to calm down the displayed values to something reasonable.  Meanwhile,
the Nortek internal sampling frequency is 17 Hz, but that is downsample averaged
to 4 Hz before we run the transforms on the raw beam velocities. Mind you, these
frequencies are MUCH higher than any meaningful speed or leeway changes in the
14.5 tonne boat. The transformed boatspeed value is damped to 5 – 8 sec decaying
filter depending on sea state (more damping in rougher seas, user adjustable on
board) and 8 sec for leeway.  Now, that merely describes the filtering on the
input vectors. They are differed, and then the output vector of current rate/dir
is boxcar averaged to 60 – 120 sec! An average of less than 1 minute is too
jumpy. An average of greater than 2 minutes is too sluggish. In coastal racing
where there are more abrupt current lines, we use 1 min, else 1 - 2 min for the
open ocean.  The current rate and direction columns in the data set are
calculated and logged onboard as described above."

SST is from the AirMar DT800 (depth sounder with temperature sensor) mounted
thru-hull on centerline.  At one point early on Robert describes the sensor as
being mounted "at about 75 cm below the water plane," and in a later discussion
describes it as being mounted "at about 0.5m depth."  In his earlier metadata
report Robert also specifies the following:

"The reported SST is raw, no correction, and represents the water temperature at
the depth sounder head 75 cm below the surface, but the water that passes that
sensor is probably from a higher level due to the way that water is entrained
under the forefoot of the hull. The sampling rate is 1 Hz. The accuracy is
unknown, other than to say that the manufacturers of sonic devices are careful
with their temperature readings, since they impinge directly on depth or speed."

"The AirMar sensor is specified to an accuracy of +/-0.5C, with a resolution of
0.1C."