Semyon A. Grodsky and James A. Carton, University of Maryland, College Park, MD

1
Interannual variation of salinity in the
equatorial Atlantic observed by the PIRATA
moorings
Semyon A. Grodsky and James A. Carton, University
of Maryland, College Park, MD

The PIRATA (PIlot Research Array moored in
the Tropical Atlantic) project is an
international program which maintains a network
of surface or near-surface measurements including
temperature, salinity, and air-sea fluxes
Servain et al., 1998. This data (now up to 8
years long) should capture the interannual
variability of the tropical Atlantic.
Although at a weaker magnitude than the Pacific
Ocean, the Tropical Atlantic Ocean also seats a
strong climatic Variability (TAV). This
variability is decomposable in two modes the
zonal primarily equatorial mode and a
trans-equatorial, or meridional mode, both
associated with the variations of SST,
atmospheric pressure and position in latitude of
the InterTropical Convergence Zone (ITCZ). The
TAV time scales range from interannual to
decadal. During the PIRATA observations
(1997-present, Fig.1) two major interannual
equatorial warm events were observed in 1998-1999
and in 2003 (Fig.2b). With the very first PIRATA
data Servain et al. 2003 have demonstrated how
the thermal structure of the two modes of TAV is
resolved by the PIRATA at interannual periods.
Less is known about the salinity contribution to
the TAV. The two modes are strongly
distinguishable by their seasonality and impact
on the meridional migration of rainfall. Warm
phase of the zonal mode (Atlantic Nino) peaks in
JJA and shifts rainfall in the east southward
from the continent into the northern Gulf of
Guinea Hisard, 1980. The meridional SST
gradient has significant impact on the southern
extension of the ITCZ rainfall band during MAM
Moura and Shukla, 1981. The ITCZ crosses the
equator in the west and brings an abundant
precipitation to the NE Brazil when the
interhemispheric SST gradient is southward. It is
expected that both modes of the TAV impacts the
equatorial salinity by modifying the net surface
freshwater flux.
Leading EOFs salinity variability along the
equator based on the PIRATA moorings data
Projection of the leading EOFs time series
JJA regression pattern displays anomaly
convergence of the equatorial winds and some
weakening of the easterly wind in the west in
response to the Gulf of Guinea warming/
freshening. In MAM the northerly
cross-equatorial wind anomaly is indicative of
the southward shift of the ITCZ in response to
colder North Tropical Atlantic. The southward
shift of the ITCZ increases equatorial rainfall
in the west that decreases salinity.
This work presents the interannual variation of
salinity in the equatorial Atlantic observed by
the PIRATA array. Along the equator we find the
surface intensified interannual salinity
variations (e.g. Fig.2a) that are decomposed into
two EOF modes weighted towards the east and the
west (Fig.3). These two modes of salinity
variability are related with the zonal and
meridional modes of SST variability, respectively
(Figs.4, 5, 6)
PIRATA salinity data coverage
Figure 3. The first and second EOFs of salinity
anomaly along the equator (a) longitude-depth
patterns, (b) time series. /o markers label each
July/April. Positive phase of the leading EOF
(negative salinity anomaly in the Gulf of Guinea)
takes place in 1998 and 2003 in phase with the
Atlantic Nino warming (compare to Fig.2b).
Figure 6. Time regression of (a) the first EOF
time series (PC1) on the JJA wind anomaly
(arrows), (b) the second EOF time series (PC2)
on the MAM wind anomaly. Wind velocity is
combination of the monthly mean SSM/I and
QuikSCAT satellite data. Areas with time
correlation exceeding 0.5 in magnitude are shaded
gray.
Projection of the leading EOFs time series

• Resume
• Salinity measurements from the equatorial PIRATA
  moorings span 8 years (1998-2005) that provides
  data for assessing the interannual salinity
  variation in the equatorial Atlantic. In this
  still too limited records we find an impact of
  the two modes of the Tropical Atlantic
  Variability.
• During Atlantic Nino years near surface salinity
  in the Gulf Guinea decreases by around 0.1 psu in
  response to 0.5oC SST warming and 1-3 mm/day
  stronger rainfall along the northern Gulf of
  Guinea.
• North Tropical Atlantic Cold SST anomaly of
  around -0.5oC shifts ITCZ southward and produces
  an excess rainfall of around 1-3 mm/day in the
  western equatorial Atlantic in MAM that results
  in lowering of salinity by 0.1 psu there.

Figure 1. (left) PIRATA buoy location and (right)
availability of the sea surface salinity data
Salinity and SST anomalies in the Gulf of Guinea

• References
• Hisard, P., El-Nino Response of the Eastern
  Tropical Atlantic, Oceanologica Acta, 3 (1),
  69-78, 1980.
• Moura A.D., and J. Shukla, On The Dynamics Of
  Droughts In Northeast Brazil - Observations,
  Theory and Numerical Experiments with a
  General-Circulation Model, J. Atmos. Sci., 38
  (12) 2653-2675, 1981.
• Servain, J., A. J. Busalacchi, M. J. McPhaden, A.
  D. Moura, G. Reverdin, M. Vianna, and S. E.
  Zebiak, A Pilot Research Moored Array in the
  Tropical Atlantic (PIRATA), Bull. Amer. Meteorol.
  Soc., 79, 2019-2031, 1998.
• Servain, J., G. Clauzet, and I.C. Wainer, Modes
  of tropical Atlantic climate variability observed
  by PIRATA, Geoph. Res. Lett. 30 (5), Art. No.
  8003, 2003.

Figure 5. Time regression of (a) the first EOF
time series (PC1) on the JJA rainfall anomaly,
(b) the second EOF time series (PC2) on the MAM
rainfall anomaly. Precipitation is the Tropical
Rainfall Measuring Mission (TRMM) satellite data.
Areas with time correlation exceeding 0.5 in
magnitude are contoured.
Figure 4. Time regression of (a) the first EOF
time series (PC1) on the JJA SST anomaly, (b)
the second EOF time series (PC2) on the MAM SST
anomaly. Areas with time correlation exceeding
0.5 in magnitude are contoured.
Figure 2. (a) Monthly salinity anomaly in the
Gulf of Guinea (missing data are shaded gray).
(b) JJA Atlantic Nino index (SST anomaly averaged
20W-10E, 4S-2N). Warm events correspond to
freshening. Reynolds and Smith SST is used.
Monthly anomaly is defined as deviation from the
seasonal cycle.