Geoscience Reference
In-Depth Information
events as seen both by space-borne and in situ sensors in the Pacific Ocean Inter Tropical
Convergence Zone is presented. Second, it is revealed that the SSS from space is sys-
tematically showing lower values (negative bias) with respect to the deeper 5-10 m depth
of Argo upper salinity in this area. These effects are shown to be statistically correlated
with rain. Third, long-lived, large-area, and large-amplitude SMOS SSS anomaly patterns
in the tropical Atlantic are shown to follow local anomaly patterns in the evaporation-
precipitation (E-P) budget. Finally, some preliminary results concerning the interannual
variability of the SMOS SSS signal in the Indian and in the tropical Pacific oceans and
connections to key climate indexes will be presented and discussed.
4.1 SSS Temporal Variability Associated with Rain Events
Although satellite observations provide a better sampling of the global ocean than the
in situ observing systems, such as the Argo float array, individual SSS measurements are
obtained in rainy regions with a strong temporal variability seen on both SMOS and Argo
SSS. In Fig.
13
, we show such an example of colocated SMOS and Argo profiler mea-
surements in the Inter Tropical Convergence Zone of the Tropical Pacific, indicating a
significant surface freshening associated with a rain event. On August 11, 2010, the Argo
float WMO id#4900325 detected a freshening of 0.9 between 20- and 5.5-m depth
(Fig.
13
a). In contrast, the Argo profile derived on 22 August shows that the salinity
between 30- and 5-m depth is much more homogeneous with more saline water at 5-m
depth compared to the one recorded on 11 August.
The TRMM satellite rain rate (RR) estimates averaged over a 2 9 2 box centered on
the Argo float location indicate a significant rain rate of 1-2 mm h
-1
on 11 August that
lasted for at least a day before the Argo profile raised to the surface (Fig.
13
c). Contrarily,
negligible precipitation occurred on 22 August and during the preceding week. The first
SMOS pass collocated with the 11 August Argo profile (Fig.
13
a) was acquired also during
rainy conditions and showed a low SSS of *32.8 (0.1 saltier than the Argo SSS taken
6:30 h later, Fig.
13
c). The second SMOS pass on the 16 August occurred under non-rainy
condition (Fig.
13
c) and is 0.5 saltier. Consistent with the 22 August Argo profile
(Fig.
13
b) observations, the collocated SMOS SSS during these rain-free conditions
(Fig.
13
c) are also significantly saltier by 0.4-0.6. The large SSS variation (0.7) measured
by this Argo float at a 10-day interval and by the collocated SMOS measurements over
several SMOS passes clearly demonstrates the influence of the rain timing on the SMOS-
Argo SSS differences.
4.2 Systematically Fresher Skin SSS in Rainy Regions
The SMOS SSS map averaged over July-September 2010 is compared to optimally
interpolated in situ ISAS map averaged over the same period shown in Fig.
14
. At large
scale, SSS spatial variability sensed by SMOS is consistent with ISAS. A striking visual
feature of the SMOS SSS map compared to the ISAS map in the tropics is the freshest SSS
in the North Tropical Pacific, under the location of the ITCZ (particularly west of 120W).
When SMOS SSS are precisely colocated around Argo SSS in various regions of the
global ocean (see Boutin et al.
2012a
,
b
), a more negative bias (*-0.1 than in other
regions) and larger standard deviation are systematically observed between 5 and 15Nin
the Pacific Ocean with respect to other regions (Table
2
).
To investigate whether a systematic negative bias of *0.1 between the satellite skin
depth SSS and the *5-m depth Argo floats data could be related to rain-induced vertical
