SENSORS AND OTHER TECHNICAL

1
SENSORS AND OTHER TECHNICAL DEVELOPMENTS ON ARGO
FLOATS Stephen C. Riser University of
Washington Seattle, Washington USA riser_at_ocean.was
hington.edu
2
There are now more than 2000 operational Argo
floats in the world ocean.
3
Presently at least 3 types of Argo floats
Argos antenna
SeaBird CTD unit
SOLO (cutaway)
APEX
Cowling (bladder inside)
4
Profiling floatsbasic operation
Argo parameter targets 4-year missions ?t 10
days 3? spacing 2000 m data available in
real-time.
5
Float sensors and communications.

• Basic sensors T, S, p
• Other sensors new communications methods
• SST, SSS, O2 , wind speed, rainfall, Iridium,
  
• biogeochemical parameters

6
Performance specifications of floats in ARGO.

• 4 year mean float lifetime (? 150 cycles)
• 2000 m profiles, at least sometimes
• CTD performance over the float lifetime
• T accurate to 0.005 ?C
• p accurate to 5 decibars
• S accurate to 0.01 (PSS-78)

7
4 year mean float lifetime.
UW float 423 has completed 164 profiles since
12/16/2000 and continues to perform well (battery
voltage is still ? 13 V).
9/10/05
Conclusion some floats will work for 4 years or
more, but the mean float lifetime at the present
time is shorter.
?
12/16/00
8
2000 m profiles.
?t 20.76
?
data from 71 levels
9
The SeaBird CTD unit pumps seawater through a
fluid circuit temperature and conductivity are
measured inside this loop, pressure is measured
outside. A biocide prevents fouling (growth)
inside the loop.
10
CTD stability.
UW float 063 was recovered after 3 years and
recalibrated. The results showed that salinity,
temperature, and pressure had all remained stable
to within the manufacturers specifications.
11
Argo CTD sensor stability.
One measure recalibration of recovered floats
from Oka (2005) Journal of Oceanography, 61,
775-782
12
Argo float CTD sensor stability.
Another measure delayed-mode analysis
From 132 UW floats, all in excess of 2
years since deployment, delayed-mode analysis
indicates 125 require no delayed mode salinity
correction (95) (ie, the systematic difference
between the float and the local S climatology
was ? 0.01 PSU) 1 shows S drift of ? ? 0.01
PSU/year 4 show S drift of ? 0.01 to ? 0.02
PSU/year 1 shows S drift of ? ? 0.03
PSU/year 1 shows S drift in excess of ? 0.03
PSU/year
analysis carried out by Dr. Annie Wong, UW
13
Argo S data can be compared to data from
WOCE sections collected in the 1980-1990 period.
14
?S
South
North
Eq
?S on ??
Number of Argo profiles compared
?S S (Argo) ? S (WOCE)
Riser and Ren, 2005
15
?S
?S on ?? 26.6
Riser and Ren, 2005
16
Other sensors..

• SST
• There are plans to modify the CTD sensor
• on floats in order to measure SST this will
  be
• done by adding a 2nd thermistor outside of
  the
• CTD tube that will collect samples at 1 m
  intervals
• between a depth of 20 meters and the sea
  surface.
• Precision ? 0.001 ?C accuracy perhaps 0.05
  ?C.
• Cost ? 200 per float. First tests early in
  2006.
• (ii) SSS
• There are a number of scientists interested
  in making
• high quality measurements of SSS. This
  appears
• to be difficult and perhaps costly and will
  not
• be attempted soon.

17
Dissolved oxygen measurements from profiling
floats.
UW float 894, with SBE O2 sensor, has been
operating for more than 3 years
18
The first SBE O2 sensor worked well for nearly 3
years, showing only small instrument drift over
that time. During the summer of 2005 the sensor
showed a very large drift and recently has
apparently stopped working.
Float 894 8/02 7/05 99 profiles
Present and future plans 6 SBE deployed, 22
more ready for deployment in autumn 2005.
19
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
22 UW Argo floats with SBE O2 sensors will be
deployed along 33? S during November 2005. This
will be the first large-scale deployment of these
sensors.
20
Comparing SBE and Optode O2 sensors.
CTD sensors SBE O2 sensor Optode sensor
The endcap of UW float 0035 is equipped with both
SBE and Optode O2 sensors.
note requires APF9 controller
21
Comparison of the data from the SBE and Optode O2
sensors on UW float 0035
22
5.0
10.0
?5.0
?10.0
?O2 Optode ? SBE (?mol/kg)

There is a systematic difference between the 2
sensors in the upper ocean of 2 - 4. This can
probably be corrected.
23
Optical sensors are being used to infer
chlorophyll and light scattering. Scattering is
proportional to particulate organic
carbon. These measurements can be used to
examine the carbon cycle in the ocean.
24
The acoustic rain gauge (ARG)
Rainfall and wind speed can be measured
acoustically using the spectrum of ambient
acoustic noise in the ocean. The results at
right are from a TAO mooring in the equatorial
Pacific (Nystuen, 1999).
At moderate wind speeds there is good agreement
between wind measured from an anemometer and
winds inferred acoustically.
a float can make similar measurements while
submerged
25
Profiling float end cap with ARG hydrophone
added that can be used to measure wind speed and
rainfall, as on UW float 0006
2h
?
h
?
Year-long trajectory of float 0006 in the Bay of
Bengal
26
(mm/hr)
A float equipped with an ARG has been operating
successfully in the Bay of Bengal for one year
note the strong seasonal variation in rainfall
and wind as measured by the float. The float is
drifting at a depth of 700 m .
27
Faster communications IRIDIUM

• 2-way communication with floats (mission
  changeable after deployment many downloadable
  commands profile depth,
• interval between profiles, sampling interval,
  etc.)
• Fast communication compared to ARGOS (? 180
  byte/sec
• compared to
• Short times on the surface ( ? 6 minutes for a
  55 KB, high resolution deep profile, compared to
  9 hours for a low resolution 0.8 KB profile with
  ARGOS)
• Use of new sensors and high-resolution
  experiments become possible, in conjunction with
  the APF9 controller
• Cost per profile comparable to ARGOS cost per
  byte

28
Iridium floats.
Standard Iridium and GPS patch antenna
configuration
Mai-tai patch antenna configuration
deployment in Jan. 2006
6 deployed
29
Standard patch configuration in tow tank dye
test Mai-tai configuration in tow tank dye test
30
1000 m parking depth
?
data from 1000 levels
CTD data from UW Iridium float 5037, showing high
resolution T and S data (?p 2 dbar) over a 2000
m profile (1000 T/S/p samples) surface time ? 6
minutes
31
Uses of Iridium.
? Applications requiring 2-way communication
? More complex sensors that require shore-based
processing
? Drift-phase sampling
It is possible to collect and store T and p
samples as the float drifts, then transmit these
data while the float is on the surface. We are
now collecting T and p data along the drift of
Iridium-equipped floats at intervals of ?t 1
hour. This requires a negligible amount of
energy. See Gille (2002) for an example of the
potential use of such data.
32
Time series of p and T taken at a depth of 1000 m
at hourly intervals along the path of UW float
5041 during the period 8/1/05 9/4/05
The spectrum of the drift-phase T time
series Note the spectral peaks at M2 and K1
tidal frequencies
33
Future developments.

• Biogeochemical floats that measure T, S, and O2,
  as well
• as chlorophyll and particulate carbon, plus NO3
  and pH

It should be possible to estimate ?CO2 in this
manner from individual floats along their
paths. The first NO3 float is now under
construction at UW, in cooperation with Dr. Ken
Johnson of MBARI.

• Acoustically-tracked profiling floats that can
• operate under seasonal ice store profiles
• during winter and transmit during ice-free
• seasons incorporate ice avoidance algorithms

Much of the work is completed project proposed
to NSF in the US but not yet funded.
34
SUMMARY There are a number of new technical
developments taking place with profiling floats,
including new sensors and faster communications.
Some of these new features are nearly ready will
be commercially available within the next year,
while other features might not be available for
several years (if at all). When planning for
the observing system of the future, the utility
of these developments should be taken into
consideration.