The Argo Steering Team

1
Argo the challenge of continuing 10 years of
progress.

• The Argo Steering Team
• GODAE Final Symposium
• Nice, 12-15 November 2008

2
Outline

• The evolution of Argo.
• Argos effectiveness.
• The Argo-era global ocean.
• Argo and ocean surface data.
• The future of Argo.
• Ocean Data Assimilation is a key application for
  Argo.
• Please submit comments and questions for the Argo
  Roundtable.

Argo deployment training with N.Z. Minister of
Research, Science and Technology, October 2007.
Photo by A. Blackwell (NIWA)
3
A technology revolution

• In the 1990s, a global survey required many years
  of research vessel effort.
• Today high quality data can be collected anywhere
  in the world without a ship being present at the
  time. (continuous, global data)

In the WOCE global survey, 8000 CTD profiles
were collected by RVs from 1991-1997.
Argo obtains 9000 CTD profiles per month.
4
An idea to a global array in 10 years.
400,000 high quality profiles have been collected
during 2004-2008. Left profiles per 1o box,
2004-2008. 3000 floats obtain 9000 profiles per
month.

• The key factors were
• The enabling technology.
• An international partnership of science and
  agencies.
• An open data policy, with free and immediate
  access.
• GODAE is one of Argos parents.

5
Argos impact is greatest in the southern
hemisphere.
Red dots Winter 1950-2000 (WOD) Black dots
Winter 2008 (Argo)
Argo obtains more winter T,S profiles in a single
year than in all pre-Argo winters combined.
Argo 3o x 3o design
Nevertheless, Argo has not yet achieved its
designed coverage in the southern hemisphere,
where an additional 750 floats are needed. The
shortfall will impact many applications.
Floats per degree of latitude
Argo present distribution
Equal area distribution
6
How effective is Argo for large-scale variability?
Argo SST anomaly, Dec 2006
Argo SST anomaly
Niño 3.4 (Argo)
NOAA OI SST anomaly, Dec 2006
NOAA OI SST anomaly
Niño 3.4 (NOAA OI)

• Argo mapping error can be estimated in several
  ways
• Formal OI error estimates.
• Maps from subsets of Argo data.
• Altimetric height subsampling experiments.
• Comparison to independent datasets such as SST.

7
The large-scale signal and Argo sampling noise
Argo is most effective in the tropics. More
floats are needed in the southern hemisphere.
Altimetric height subsampling experiment Zonally
averaged variance of large-scale (10o x 10o x 3
months) anomalies (mean and annual cycle
removed) Black Signal variance from 15-year
AVISO dataset, 1993-2007. Blue Signal variance
from 4-year dataset, 2004-2007. Thick red Noise
upper bound, SSH minus steric height,
2004-2007. Thin red Noise lower bound, SSH minus
subsampled and re-mapped SSH, 1993-2007.
SSH
SIGNAL
NOISE
SST
Argo/ NOAA OI SST comparison Black Signal
variance from 15-year dataset, 1993-2007. Blue
Signal variance from 4-year dataset,
2004-2007. Red Noise upper bound, NOAA OI SST
minus Argo SST, 2004-2007.
SIGNAL
NOISE
8
Systematic errors?

• (Right) Willis et al. (2008) noted that the
  increase in global sea level is not seen in
  4-year records of steric sea level and ocean
  mass.
• How accurate is the global mean temperature and
  steric height from Argo?
• A highest priority for Argo is to identify and
  correct systematic errors (e.g. p0 drift) and to
  estimate their impact.

Global mean sea level variability (top), steric
component from Argo (middle), mass component from
GRACE (bot). Grey lines represent the residual of
the other two measurements. From Willis,
Chambers, and Nerem (GRL, 2008).
9
Testing the Argo dataset.
New techniques are being developed for error
detection.
Altimetric height is used to flag anomalies in
Argo steric height for expert examination (Guinehu
t et al., 2008). Here a problem is detected in
data from Float 5900984.
10
Testing the Argo dataset.
Argo and shipboard transects show similar and
consistent decadal signals in temperature along
24.5oN in the Atlantic. (Vélez-Belchí,
Hernández-Guerra and Fraile-Nuez, 2008).
RAPID - IGY
Argo - IGY
RAPID - IGY
Argo - WOCE
11
The Argo-era ocean.
The Argo dataset, 2004-2008, provides an accurate
5-year mean and annual cycle for the global
ocean. Argo can be compared to past datasets and
is a baseline for observing future
evolution. Maps of Argo-minus-WOA01 steric
height highlight the large and deep density
changes south of 30oS. Roemmich and Gilson (2008)
12
The Argo-era ocean.

• The southern hemisphere ocean is warmer and
  fresher in the Argo era than in WOA01.
• The northern hemisphere is warmer and saltier.
• Heat gain is dominated by the southern hemisphere
  (larger area).
• The surface layer stratification is increased.
• The Argo ocean is fresher in high rainfall
  regions, saltier in high evaporation regions
  (increase in the global hydrological cycle?)

Global averages of Argo-minus-WOA01 T and S
Zonal averages of T, S, and s? from Argo
(contours), and the Argo- minus-WOA01 differences
(colors). Roemmich and Gilson (2008).
13
The Argo-era ocean.

• Argo-minus-WOA01 salinity differences on density
  surfaces, excluding the upper 200m.
• Subsurface waters are fresher in the SH
  intermediate waters (all oceans) and below the
  ITCZ.
• Subsurface waters are saltier below evaporative
  regions and NH intermediate waters (Atlantic).

14
Argo and ocean surface datasets.
Argo SH
AVISO SSH
Argo SST
NOAA OI SST
Argo A-S flux
NOC A-S flux
Argo zonally-averaged annual cycle is compared to
altimetric height, SST, and air-sea flux.
15
Argo and ocean surface datasets.
Sea Surface Height
Sea Surface Temperature
Hemispheric and global annual cycles are compared
for consistency and for complementary
information. SH southern hemisphere NH
northern hemisphere GL global
SH
SH
GL
GL
NH
NH
Argo SH AVISO SSH Difference
Argo SST NOAA OI SST
Air-Sea flux
Sea Surface Salinity
NH
NH
GL
GL
SH
SH
Argo heat gain NOC A-S flux
Argo SSS WOA01 SSS
16
Argos future two paths forward.

• Improved implementation for Argos original
  objectives
• Increased float lifetime.
• Enhanced float capabilities.
• Better coverage in the southern hemisphere.
• Detection/correction of systematic errors.

Right A 6year record, with stable salinity, by
a UW float in the Indian Ocean (fig. provided by
A. Wong).
Float lifetimes continue to improve through
technical innovation and careful handling.
17
Argos future two paths forward.

• 2. Potential objectives to increase Argos
  value
• Abyssal floats to sample the full water column.
• New sensors biological, geochemical, surface
  layer,
• Regional arrays in marginal seas.
• High latitude floats under seasonal ice.
• Glider sampling in boundary currents.

Argo was designed for the ice-free oceans, but
there are now many floats in the seasonal ice
zones.
A UW float is deployed through the ice on a
cruise by R/V Aurora Australis. Photo G. Williams
18
Argos future the planning process.

• This GODAE F.S. Argo paper is intended to
  initiate a review of Argos status and a
  discussion of its future priorities. Contribute
  via the Argo Round Table.
• Argos 3rd Science Workshop, The Future of Argo
  will be held in Hangzhou in March 2009.
  http//www.argo.ucsd.edu/ASW3.html .
• Broad input and participation are invited
  (including ODA Argo users).
• A Community White Paper for OceanObs09 will be
  developed.