5
10
15
20
25
30
35
National Center for Atmospheric Research
Boulder, CO
Los Alamos National Laboratory and New Mexico Consortium
Los Alamos, NM
University of Maryland
College Park, MD
National Center for Atmospheric Research
Boulder, CO
Submission to: Journal of Climate, Special Collection on CCSM4
Date of revision: October 2011
Corresponding author address:
Ernesto Muñoz
National Center for Atmospheric Research
NESL/CGD/OS
P.O. Box 3000
Boulder, CO, 80307-3000
E-mail: emunoz@ucar.edu

40 Abstract:
In this study we analyze important aspects of the tropical Atlantic Ocean from the new simulations of the 4 th version of the NSF-DOE coupled climate model, the Community Climate
System Model (CCSM4). The data used in this study is from several different simulations, among them a set of five 20 th
-century simulations with different initial conditions, but similar
45 radiative forcing. Among the features analyzed in this study are: the structure of the Atlantic warm pools; the main modes of sea surface temperature (SST) variability in the tropical South and North Atlantic; the variability of heat storage in the Benguela region; and differences between the model simulations and the observations in the tropical Atlantic. The results indicate that some of the biases of the tropical Atlantic have been reduced in CCSM4 compared to the
50 previous version of the CCSM. Yet, there are still significant biases in the CCSM4 sea surface temperatures (SST), with a colder tropical North Atlantic (TNA) and a hotter tropical South
Atlantic (TSA), that are related to biases in the wind stress. The biases in the TSA and the TNA are also reflected in the Atlantic warm pools in April and September with a warm pool volume greater than in observations in April and smaller than in observations in September. The
55 variability of SSTs in the tropical Atlantic is well represented in CCSM4 although the leading
EOF in CCSM4 shows more homogeneous warming than the leading EOF in observations. A heat budget analysis of the Benguela region indicates that the variability of SSTs is dominated by vertical advection.
2

60 1. Introduction
The tropical Atlantic is an important region in the Earth’s coupled climate system.
Understanding the variability of the tropical Atlantic is not only important for the communities of tropical America and tropical Africa, but is also important for other remote regions and for the climate system in general. For example, Ding et al. (2011) and Losada et al. (2009) document
65 the remote impacts that the tropical Atlantic has on ENSO. Furthermore, Losada et al., (2009),
Kucharski et al., (2008) and Zhang and Delworth (2006) show that the tropical Atlantic has an impact on the Indian Ocean climate variability. A few recent reports summarize the advances in the understanding of the tropical Atlantic climate and its variability (e.g., Hurrell et al. 2006; Xie
70 and Carton, 2004; Garzoli and Servain, 2003; Visbeck et al. 2001).
To date, the tropical Atlantic has been challenging to model adequately by coupled climate models. The recent availability of the CCSM4 provides an opportunity to assess the status of the simulation of the tropical Atlantic by one of the leading coupled climate models.
Also, the Community Earth System Model (CESM) (of which the CCSM4 is a subset) will be used as part of the next phase of the Coupled Model Intercomparison Project (CMIP5); is
75 therefore important to understand the CCSM4 simulation of the tropical Atlantic Ocean. In this study the following main aspects of the tropical Atlantic Ocean are analyzed from the 4 th version of the NSF-DOE Community Climate System Model (CCSM4).
80 a. Mean Biases in the Tropical Atlantic
The annual cycle of sea surface temperature (SST) in the tropical Atlantic is tied to the annual cycle of wind stress. When the Intertropical Convergence Zone (ITCZ) is furthest from the equator in August/September, cross-equatorial flow is at its weakest seasonal state, and
3

equatorial easterlies are at their strongest. These are upwelling favorable winds; therefore, the coldest eastern equatorial SSTs occur during this season (Fig. 2b). When the ITCZ is at its
85 southernmost position in boreal spring, meridional cross-equatorial flow dominates the deep tropics, and a hemispheric dipole structure in SST is present. During this period, the warmest
SSTs are present along the equator (see Bates, 2008; Okumura and Xie, 2006 for more details on the tropical Atlantic annual cycle).
Among the known biases of coupled models in the tropical Atlantic region are: a warm
90 bias in the tropical southeastern Atlantic, a barrier layer thicker than observed, and relaxed zonal winds along the equator related to weaker precipitation over the Amazon region. (Doi et al.
2012; Tozuka et al. 2011; Richter et al. 2011; Wahl et al. 2011; Richter and Xie, 2008; Breugem et al. 2008; Chang et al. 2007).
95 b. The Atlantic Warm Pools
Warm pools (WP) have been defined as those regions of the ocean with temperatures greater than 28.5°C (Tian et al. 2001; Wang and Enfield, 2003). The Atlantic WP has a component in the northwestern tropical Atlantic spanning the Gulf of Mexico and the Caribbean
Sea (i.e., the Intra-Americas Sea) that peaks during boreal summer and early fall (Wang et al.,
100 2006). The other component of the Atlantic WP peaks during boreal winter and spring when the waters of the tropical South Atlantic reach temperatures greater than 28.5°C also forming a warm pool.
Beyond its surface manifestation and extent, the Atlantic WPs in the Tropical North
Atlantic (TNA) and the Tropical South Atlantic (TSA) have vertical and horizontal profiles that
105 are important with respect to the heat content of the upper layer of the ocean. The heat content in
4

the TNA-WP is also available for tropical storms and hurricanes that travel through that region
(Wang et al. 2006). Furthermore, the earlier CCM3 modeling study of Saravanan and Chang
(2000) regarding the Caribbean sea surface temperatures (SSTs) already pointed to the Caribbean as critical in the teleconnections with the tropical Pacific. That is to say, Saravanan and Chang
110 (2000) found that the Caribbean heat sources can affect the Pacific through an upper-level Gilltype circulation. This and other impacts of the TNA-WP were documented by Wang et al.
(2008) from a model. c. Leading modes of Tropical Atlantic Variability
Observed changes in SST, which affect the meridional gradient of SST in the tropical
115 Atlantic, occur on a wide range of time scales. On multidecadal timescales, the Atlantic
Multidecadal Oscillation (Enfield et al., 2001) is mostly confined to the North Atlantic, and possibly reflects changes in the strength of the Atlantic Meridional Overturning Circulation
(Muñoz et al., 2011). At shorter, decadal timescales, the out-of-phase variations of SST in the northern and southern tropical Atlantic are self-sustained and driven by the wind-evaporation-
120 SST feedback in the trade winds (Carton et al., 1996; Chang et al., 1997). At shorter timescales remote impacts of the El Niño-Southern Oscillation (ENSO; Enfield and Mayer, 1997) and the
North Atlantic Oscillation (Czaja et al., 2002) produce different SST responses in the northern and southern tropical sectors and thus contribute to the observed lack of coherence between SST variations in the two regions.
125 The tropical Atlantic variability (TAV) has been observed to have a few main modes of variability that are predominant at different times of the year (Servain et al. 1990; Servain et al.
2003). One of the modes of variability is the so-called meridional mode or interhemispheric mode and is usually observed in the boreal spring (Servain et al. 1998; Mahajan et al. 2010).
5

The meridional mode is characterized by a meridional gradient of SST anomalies from one
130 subtropical region to its counterpart in the other hemisphere, and a pattern of surface wind anomalies from the colder subtropics to the warmer subtropics (Nobre and Shukla, 1996).
Another mode of variability is the so-called zonal mode or Atlantic Niño and is predominant in the boreal summer (Tokinaga and Xie, 2011; Carton and Huang, 1994; Zebiak, 1993; Shannon et al. 1986). Yet, the identification of Tropical Atlantic modes of variability have also benefited
135 from the use of statistical techniques such as rotated Empirical Orthogonal Functions (rEOFs).
Modes of TAV by using rEOFs have been referred to as the southern tropical Atlantic (STA) pattern, the northern tropical Atlantic (NTA) pattern, and southern subtropical Atlantic (SSA) pattern (Hu et al., 2008; Bates, 2010). Previous studies have analyzed these modes and the dynamics and thermodynamics that explain their variability (Bates, 2010; Huang and Shukla,
140 2005; Florenchie et al. 2004; Chang et al. 1997; Carton et al. 1996; Shannon et al. 1987). d. Heat budget of the Benguela region
A majority of coupled GCMs have SST biases in the tropical Pacific and Atlantic including notorious warm biases in the southeastern tropical basins (Zuidema et al., 2011). But the most severe warm SST bias (in excess of 5°C) occurs along the Benguela coast of
145 southwestern Africa (e.g. Chang et al., 2007; Richter and Xie 2008). This spurious pool of abnormally warm water simulated by a majority of climate models alters large scale meridional
SST gradient across the tropical Atlantic, and thus projects on the natural mode of the tropical
150
Atlantic variability known as the meridional or inter-hemispheric mode ( e.g. Xie and Carton,
2004) .
Periodic changes of SST in the northern and southern tropical Atlantic meridionally displace the Intertropical Convergence Zone (ITCZ) and affect rainfall over surrounding
6

continents in the northeastern Brazil and African Sahel (see e.g. Xie and Carton, 2004 and references therein). Observation-based analyses of SST variability in the tropical Atlantic indicate that the standard deviation of anomalous
1
SST is strongest in areas adjacent to the
155 western coast of Africa (Doi et al., 2010) and reaches a maximum in the Angola-Benguela frontal zone (referred to as the Benguela region in this paper; see e.g. Florenchie et al., 2003).
The enhanced variability of SST in the Benguela region suggests that this particular region contributes significantly to the meridional gradient of tropical Atlantic SST and thus plays a role in tropical Atlantic variability. But due to complex chain of the air-sea-land feedbacks the
160 accuracy of coupled simulations of SST in the tropical Atlantic and particularly in the Benguela region still remains a challenge.
Observations and model simulations indicate that SST in the Benguela region is affected by local and remote impacts. Florenchie et al. (2003) have suggested a link between the
Benguela warm events and weakening of the zonal equatorial winds 1 to 2 months in advance,
165 which remotely impact the Benguela region via Kelvin waves propagating eastward along the
Equator and further south along the coast. On longer time scales Chang et al. (2007) and Richter and Xie (2008) have shown that abnormally weak equatorial easterly wind is responsible in part for the time mean warm bias of the Benguela SST. In addition to remote mechanisms, the impact of local meridional winds and upwelling on the Benguela SST has been demonstrated by Large
170 and Danabasoglu (2006). Impact of local upwelling is also emphasized by Grodsky et al. (2011) who argue that adequate representation of magnitude of the southerly Benguela low-level wind jet (Nicholson, 2010) is crucial for maintaining the zonal sea level gradient in the coastal ocean, and thus cold water transport by the coastal jet of Benguela Current. Independent of its origin
1 Anomalous signal is calculated by subtracting the monthly seasonal cycle.
7

any warm SST bias in the Benguela region grows and expands via the positive feedbacks from
175 marine stratocumulus clouds (e.g. Mechoso et al., 1995).
Relative impacts of local versus distant winds on the Benguela SST have been addressed in a number of recent reports (Richter et al. 2010; Rouault, 2010). In particular, Richter et al.
(2010) have demonstrated based on observations and model simulations that impact of local upwelling on Benguela SST is comparable to the remote impact of the Equatorial winds. As a
180 part of the CCSM4 evaluation in this paper we perform the heat budget analysis of the upper ocean layer in the Benguela region in order to quantify relative contributions of the air-sea heat fluxes versus heat advection terms and evaluate their relationship with changes in local and
185 remote winds. e. Organization of the manuscript
Section 2 lists the data used in this study. Section 3 introduces the general improvements and biases in the tropical Atlantic of the CCSM4 surface fields compared to observations and those of CCSM3 (the previous CCSM version). In Section 4 an analysis of the structure of the
Atlantic Warm Pools (in the north and south tropical Atlantic) is presented based on a suite of ensemble simulations. In Section 5 the main modes of sea surface temperature (SST) variability
190 in the tropical Atlantic are compared against those from observations. In Section 6 the variability of the tropical South Atlantic, in specific the Benguela region, is further analyzed based on the heat storage. Summaries and discussion are presented as the final section of the manuscript.
195
3. Model and Observational Data a. Model data
8

In this study we compare the data from CCSM4 simulations against observations in the tropical Atlantic. The main CCSM4 data set analyzed is an ensemble of five 20 th Century (20C)
200 CCSM4 simulations (see Gent et al. (2011)). These CCSM4 ensemble members were run in the same manner with the exception of their initial state. Each of the CCSM4 members was initialized from the 1850 control simulation, with the five initializations chosen to represent different states of the Atlantic meridional overturning circulation (see Gent et al., 2011, this issue). The period used for the analyses in this study span from 1950 to 2005 of the CCSM4 20C
205 simulations, i.e., the last few decades of data. Complete descriptions of these simulations are provided by Gent et al. (2011).
The CCSM4 20C ensemble simulations were compared to observations, to a CCSM3
20C ensemble mean, and to a coupled ocean-sea ice experiment forced by the CORE v2 Inter-
Annual Forcing data (Large and Yeager, 2009). This ocean simulation (POP-CORE) was
210 conducted using the 1-degree horizontal resolution version of the CCSM4 ocean model coupled to an active freely evolving dynamic-thermodynamic sea ice model (CICE). vMonthly climatological river runoff is based on Dai et al. (2003) discharge estimates. The data analyzed is from the fourth (last) forcing cycle of 60 years (1948-2007).
There were eight ensemble members in the CCSM3 simulations, and where possible, all
215 eight members are used. One of the ensemble members did not contain monthly output and, therefore, is not used in the annual cycle comparisons. The CCSM3 ensemble members were initialized from the 1870 control simulation (the control was switched to 1850 for CCSM4) at
20-year intervals with no tie to a physical feature (see Gent et al., 2006, Table 1 for details on
CCSM3 simulations). The CCSM3 simulations ended in December of 1999; therefore, the 20-
220 year mean used in this study is from 1980-1999.
9

Differences between CCSM3 and CCSM4 are expected due to various changes in the model physics and the spin-up and tuning procedures, which are all intended to produce a more realistic model. Both sets of simulations are the nominal 1-degree resolution in all components; however, this resolution has increased in the atmospheric and land components from
225 approximately 1.4 degrees in CCSM3 to approximately 1 degree in CCSM4. The depth resolution of the ocean model has also increased from 40 levels in CCSM3 to 60 levels in
CCSM4, with the majority of the additional layers in the upper ocean. The dynamical core is different in the atmospheric component (Neale et al. 2011, this issue), the ice model component, with a new radiation scheme and different albedo values, produces different sea ice cover
230 (Holland et al. 2011, this issue), and the addition of new ocean parameterizations include more ocean physics than were present in CCSM3 (Danabasoglu et al. 2011, this issue). Due to differences in model tuning and spin-up (Gent et al., 2011, this issue), the ocean in the CCSM3
20C simulations lose heat over the length of the run, while in the CCSM4 simulations it more realistically gains heat.
235 In the section on the Benguela heat budget, we use monthly averaged fields from the 1degree 1850 control run of CCSM4 (Gent et al., 2011, archived as b40.1850.track1.1deg.006), which is a 1300-year simulation forced by fixed pre-industrial levels of ozone, solar, volcanic, greenhouse gases, carbon, and sulfur dioxide/trioxide. Our analysis focuses on data from a 97year period (model years 863-959)
2
. A sensitivity examination has been carried out to ensure
240 that the climatology of this particular period is similar to that of later periods. b. Observational datasets
2 Selected period follows model years 715 and 851 when some small adjustments to the model parameterization were done (Gent et al., 2011).
10

For the analyses of the mean biases, the observations used include: sea surface salinity
(SSS) from the Polar Hydrographic Climatology (PHC2) dataset (a blending of Levitus et al.
245 (1998) and Steele et al. (2001)), sea surface temperature (SST) from Hurrell et al. (2008), and wind stress from an uncoupled ocean-ice simulation forced with the Coordinated Ocean-ice
Reference Experiments (CORE; Griffies et al. (2009), Large and Yeager (2009)), which represents the wind stress forcing of the ocean by the CORE atmospheric wind data. The time period used from the observations is chosen to match that used in the CCSM3 and CCSM4
250 ensemble means.
For analyses of the Atlantic warm pools, two observational data sets were used for comparison. One observational data set is the World Ocean Atlas 2009 (WOA09) climatological fields (Locarnini et al. 2010; Levitus et al. 1998) with mean temperature data for the twelve calendar months. These WOA09 monthly climatological fields are based on available
255 observations during the period from the year 1773 to the year 2008. Also used is the observational data set developed by Ishii et al (2006) with monthly temperature data interpolated to a 1x1 degree grid. Both of these observational data sets have the same horizontal 1x1 degree grid, and the same vertical levels, with data at the surface, 10m, 20m, 30m, 50m, 75m, 100m,
125m, 150m, 200m, 250m, and at 100m intervals between 300m and 700m. (The depth of the
260
28.5°C isotherm does not exceed 185 meters at any time in the period analyzed from these observational products.)
An advantage of the observational product from Ishii et al (2006) (from now on Ishii) is that it provides monthly data for the recent period thereby allowing for the calculation of means based on different periods (e.g., from 1950 to 2005), whereas the Levitus climatology provides
265 12 monthly climatologies based on available observations for the period 1773-2008 covering
11

more than a couple of centuries. For the warm pool estimates, the averages from the Ishii product were computed based on the period 1950-2005.
For the comparison of rotated Empirical Orthogonal Functions (rEOFs) the Extended
Reconstructed SST version 3b (ERSSTv3b; Smith et al., 2008) observational data set is used.
270
3. Mean Biases in the Tropical Atlantic Ocean a. SST, SSS and TAU mean biases
In general, a shift to warmer surface ocean temperatures is noted in the CCSM4 versus
275 CCSM3 (Danabasoglu et al, 2011), mostly due to the spinup procedure effects described above.
This holds true in the tropical Atlantic basin (Fig. 1a,c). The mean values from 40°S to 40°N of the tropical Atlantic biases are -0.53°C for CCSM3 and 0.61°C for CCSM4, excluding the
Mediterranean Sea and Pacific Ocean. These indicate that the overall mean of the SST bias has not changed in magnitude but has flipped from an overall negative bias in CCSM3 to an overall
280 positive bias in CCSM4. The root mean square (RMS) error of these biases over the same region does show improvement with a value of 1.52°C for CCSM3 and 1.29°C for CCSM4.
The largest SST biases are found in the southern Caribbean Sea along the coast of South
America, and in the southeastern basin along the coast of Africa. The largest negative bias occurs in the southern Caribbean Sea in both CCSM3 and CCSM4 with a value of ~ -2.5 and ~ -4.0
285 degrees C, respectively. The largest positive bias occurs in the southeastern basin near the
African coast in CCSM3 with a value of ~7.5 degrees C and in the Gulf Stream region in
CCSM4 with a value of ~8.0 degrees C. The maximum value along the coast of Africa in
CCSM4 is somewhat improved over CCSM3, but is still large at ~6.5 degrees C.
12

Sea surface salinity (SSS) in the tropical Atlantic has improved in the CCSM4 (Figs.
290 1b,d). The mean of the SSS biases, as calculated above, is -0.517 g/kg for CCSM3 and -0.148 g/kg for CCSM4 indicating an overall reduction of fresh biases. This is different from the global mean bias, which shows little to no improvement from CCSM3 to CCSM4 (Danabasoglu et al.
2011, this issue). The RMS error of these biases also displays improvement with a value of 1.080 g/kg for CCSM3 and 0.775 g/kg for CCSM4. The maximum value of the biases occurs in the
295 same regions in both CCSM3 and CCSM4. The minimum bias value is located off the coast of
Angola with a value of ~ -12 and ~ -6 psu in CCSM3 and CCSM4, respectively, while the largest bias value occurring at the mouth of the Amazon River with a value of ~10 and 11.3 psu in
CCSM3 and CCSM4, respectively.
Large improvements in the fresh bias are apparent in the northern and southern tropics
300 from approximately 20°S to the equator and in the eastern North Atlantic from approximately
15°N-30°N. This improvement is most likely due to a reduction of the positive precipitation bias in this region causing a reduction in the input of freshwater to the ocean (Bates et al. 2011, this issue). Salinity bias improvements are also noted in the south Caribbean Sea; however, the northern portion of the Gulf of Mexico now has a larger saline bias in CCSM4 over CCSM3.
305 Changes in the Gulf Stream and North Atlantic Current path have reduced fresh biases in the central North Atlantic and increased them off the coast of North America. The result is a generally saltier North Atlantic (Danabasoglu et al., 2011, this issue). Persistent in both CCSM3 and CCSM4 are the excessive runoff from the Congo River, which contributes to the fresh bias in the southeast Atlantic basin, and the weak runoff from the Amazon River, which contributes
310 to the salty bias extending from the river mouth north to Bermuda (Danabasoglu et al. 2011, this issue).
13

The biases in the mean state of zonal (TAUX) and meridional (TAUY) wind stress for
CCSM3 and CCSM4 compared to observations are provided in Fig. 3. As will be shown later, wind stress is important to processes controlling heat content changes in the eastern equatorial
315 Atlantic as well as to changes in the volume of the Atlantic warm pools. The overall pattern of easterlies throughout the tropics with centers near 15S and 15N and equatorward flow strongest near Africa (Fig. 3a) is captured by CCSM3 and CCSM4. The direction of the wind stress, for the most part, is correct in the model, it is the strength that causes the biases in Fig. 3 (b,c). In general, the coupled model exhibits weaker wind stress throughout the equatorial region. In the
320 western basin, this is mostly due to weakened easterlies, and in the eastern basin, weakened southerlies are responsible for the decrease in magnitude. In the regions of strongest easterlies
(centered around 15S/N), the wind stress is enhanced over CORE with largest differences in the eastern basin near the African continent. The wind stress due to easterlies in the southern
Caribbean region is much too strong in both CCSM3 and CCSM4. In general, all of these biases
325 have been reduced in CCSM4 compared to CCSM3. b. Seasonal cycles wind stress and SST biases
In Figure 4, we present the seasonal cycle of wind stress from CORE (left panels) along with the bias for each season in the CCSM4 compared to CORE (right panels). In all seasons, wind stress is too weak in the deep tropics with bands that are too strong at higher tropical
330 latitudes. The deep tropics biases are dominated by weakened easterlies in the west and weakened southerlies in the east, as found in the total mean bias discussed above. The largest and most widespread of these biases occurs during the Dec-Jan-Feb (DJF) and Mar-Apr-May
(MAM) seasons, corresponding to the seasons when the equatorial easterlies are the weakest and monsoonal flow in the Gulf of Guinea the strongest. The weakened easterlies are related to the
14

335 incorrect SST gradient along the equator, and perhaps the model is under-representing the air-sea temperature difference driving the monsoonal flow. In MAM, a large bias of southerly wind stress occurs in the central basin just north of the equator. This bias is due to the displacement of largest cross equatorial flow westward in CCSM4, while in CORE this maximum occurs in the
340 central basin.
The annual cycle of SST along the equator in the CCSM4 has improved greatly over the
CCSM3 (Fig. 2b,c,d), most likely due to improvements in wind stress forcing from the atmosphere. Most notable of these improvements is found in the warm phase in late boreal winter and spring. In CCSM3, the warm phase begins in the western basin and propagates eastward, neither of which occur in the observations, and are corrected in the CCSM4. Though
345 the phasing in the CCSM4 lags observations by approximately half a month, the character and magnitude are quite similar. c. Biases in the tropical North Atlantic and tropical South Atlantic
Both improvements and degradations in SST bias occur in the transition from CCSM3 to
CCSM4. Of importance to the investigations of this paper are the large reduction in the cold
350 biases centered at 20°N as well as in the Caribbean Sea and Gulf of Mexico. The warm bias of the upwelling regions is worse in CCSM4 compared to CCSM3. This too is a global feature, as the SST warm bias in CCSM3 in all the eastern basin upwelling regions has worsened in
CCSM4. Both the reduction of the cold biases and the worsening of warm biases could be partly
355 due to the overall warming in CCSM4 compared to CCSM3.
1) T
ROPICAL
N
ORTH
A
TLANTIC
B
IASES
A consistent bias of overly strong easterlies in the Caribbean Sea is present in all seasons, with the largest biases in the southern region (Fig. 4, right panels). However, the difference plots
15

in Figure 4 include the mean bias, and much this bias is present in the mean (Fig. 3b).
Comparing the seasonal cycle between observations and the CCSM4, with the respective annual
360 means removed (not shown), we note that the seasonal TAUX biases in this region are slightly too weak in Jun-Jul-Aug (JJA) and slightly too strong, but not as widespread as those in CORE in Sep-Oct-Nov (SON). Easterlies in this region may drive coastal upwelling and thus may explain the negative bias in SST (Fig. 1a).
In the north tropical Atlantic centered at 25N and next to the African coast, where
365 northeasterly winds dominate, the magnitude of wind stress is too strong compared to CORE
(Fig. 4, right panels). This is true in all seasons for zonal wind stress, and in JJA and SON for meridional wind stress (though the bias is still somewhat present in DJF and MAM). As with the
Caribbean Sea, these biases are present in the mean, and when the seasonal departure from the annual mean is examined, we find the zonal wind stress bias to be the strongest in DJF and also
370 present in MAM, and the meridional wind stress weakened in DJF and strengthened in JJA.
2) T
ROPICAL
S
OUTH
A
TLANTIC
B
IASES
The gradient of SST along the equator in both CCSM3 and CCSM4 is influenced by the warm biases in the eastern basin (Fig. 2a). Observations indicate a west-east warm-to-cold gradient. CCSM3 exhibits the opposite gradient. While CCSM4 obtains approximately the right
375 warm pool temperatures in the west, the eastern basin is still much too warm.
Meridional winds along the African coast from the southern tip to the Gulf of Guinea are important for coastal upwelling and thus affect the Benguela Niño and equatorial SST variability.
In the mean, a bias of overly strong southerly winds is present from the tip of Africa to approximately 15S with southerly winds that are too weak extending from there to the Gulf of
380 Guinea (Fig. 3b). In general, this bias is present in all seasons (Fig. 4, right panels); however, the
16

weakened southerlies in the Gulf of Guinea are strongest and most widespread in DJF and
MAM, while the strengthened southerlies in the southern portion are strongest in SON.
Removing the annual mean to examine the seasonal departures, we find the more northern biases to indicate weakened wind stress in DJF, accompanied by southerlies to the south that are too
385 strong, with the opposite sign biases occurring in JJA.
In CCSM4, a large positive bias, not present in CCSM3, occurs in the south Atlantic with its center in the eastern basin near 40S (Fig. 1a). This, too, is a global feature with warm biases appearing in all ocean southern ocean basins with centers in the eastern basin (see Danabasoglu et al, 2011, this issue, their Fig. 6a).
390 4. The Atlantic Warm Pools
In this section the differences in the simulation of the Atlantic warm pools (WP) by
CCSM4 are evaluated against those estimated from observations, and from an ocean-only POP hindcast forced by CORE surface forcings. The CCSM4 data and the CORE-forced POP ocean data are provided in a non-equidistant grid with horizontal resolution of nominally 1 degree (and
395 finer resolution in the Tropics). The vertical resolution of the model is finer than that of the observational data sets. To do a consistent comparison of the depth of the 28.5°C isotherm
(Z28.5), the model data were interpolated to the vertical levels of the observations before calculating the Z28.5 depth. Once the Z28.5 depth was calculated, the result was interpolated to the horizontal grid of the observational products at each month. The same was done for the
400 CORE-forced ocean-only POP hindcast. These monthly values were used to calculate the warm pool time series and other statistics. a. Seasonal cycle
Figure 5 (1em) shows the month of the calendar year when the 28.5°C isotherm is
17

deepest in the long-term mean. When the CCSM4 ensemble mean is compared to the Levitus
405 and the Ishii data sets, a better agreement is obtained between CCSM4 and Ishii. The CCSM4
Z28.5 extends throughout the tropical South Atlantic (TSA) similar to the Ishii Z28.5 (Fig. 6
(2em)). Also, in the tropical North Atlantic (TNA) the CCSM4 produces a Z28.5 across the basin at about 10°N similar to the Ishii Z28.5 (Fig. 7 (3em)). Yet, there are some differences in the timing and extent of the deepest Z28.5, e.g., in the southern Caribbean Sea. We discuss the
410 major differences below. The warm pool extent and timing in the CORE-forced ocean-only hindcast seems to match better the Ishii observational estimate, indicating that some of the major differences between CCSM4 and observations are in the coupled framework (not strictly in the ocean component of CCSM4: POP).
The seasonal cycle of warm pool volume in the tropical Atlantic is shown in Fig. 5 (1em).
415 In all products there are relative maxima in boreal spring and in boreal summer indicating greater extent in April and in September, respectively. The peak in April corresponds to the TSA warm pool, whereas the peak in September corresponds to the TNA warm pool (Figs. 5a-d (1em)). b. Tropical North Atlantic Warm Pool
Figure 6 (2em) shows the mean warm pool in September, when the TNA warm pool has
420 its greatest horizontal extent (Figs. 5 (1em)). Among the similarities in all products is the presence of deepest Z28.5 in the northwestern Caribbean Sea between Cuba and Central
America. In Levitus it is observed the lack of Z28.5 in the southern Caribbean Sea. Similarly, in
CCSM4, the lack of 28.5°C temperatures in the southern Caribbean Sea spans all ensemble simulations. Yet, the POP-CORE simulation has a warm pool very similar in spatial extent to
425 that of the Ishii observational estimate. Among other similarities between POP-CORE and Ishii is the presence of a warm pool in the southern Caribbean Sea. However, the POP-CORE seems
18

to have Z28.5 in the Caribbean Sea deepest in November, whereas the observations (Ishii) have them deepest in October (Fig 5 (1em)). The CCSM4 September warm pool is smaller to the east of Puerto Rico, related to the cold bias in CCSM4, and the bias of stronger easterly wind stress in
430 the TNA.
One of the main differences between the CCSM4 and the Ishii product is the lack of a
Z28.5 in the Caribbean Sea in the CCSM4. Both the Levitus and the Ishii products have a Z28.5 in the Caribbean Sea deeper in October. Even though the CCSM4 shows a Z28.5 in the northern
Caribbean Sea deeper in November, it does not extend to the southern Caribbean Sea. In fact,
435 this is related to the strong meridional gradient of subsurface temperatures across the Caribbean
Sea in CCSM4, which is an indication of the strong zonal wind stress along the Caribbean Sea
(i.e., the Caribbean low-level jet (Muñoz et al., 2008)), in particular in the southern Caribbean
Sea (as observed in Fig. 3). This stronger Caribbean low-level jet would induce stronger upwelling in southern Caribbean Sea thereby simulating temperatures colder than observations.
440 This relationship indicates that there are dynamical aspects setting the lack of Z28.5 in the southern Caribbean Sea.
The September time series of TNA volume (km
3 ) encompassed by the 28.5°C isotherm
(i.e., the TNA-WP) are shown in Figure 6 (2em). The volume was calculated from the 5°N latitude to the north and across the basin. The Pacific data were not included in the volume
445 calculation. The Ishii observational product shows anomalously large volumes in the late 1980s and after 1994. Between 1950 and 2005 there are years of decreased observed volume in the mid-1970s (minimum in 1974) and the mid-1980s (minimum in 1984). The POP-CORE hindcast has a similar evolution than Ishii with relative minima in the mid-1970s and in the mid-
1980s. The CCSM4 ensemble spread encompasses the observational estimate or the POP-CORE
19

450 hindcast after the 1970s. c. Tropical South Atlantic Warm Pool
In the tropical South Atlantic (TSA), April is the month of greatest volume of Z28.5
(Figs. 5, 7 (1em, 3em)). From the ensemble mean (Fig. 6 (2em)) it can be observed that the
CCSM4 has a warm bias with respect to observations. A main difference between the Z28.5
455 from observations and from CCSM4 is the timing of the deepest Z28.5 in the tropical South
Atlantic (Fig. 5 (1em)). In the Gulf of Guinea and Benguela regions, the observations have the deepest Z28.5 in March and April respectively, while the CCSM4 has its deepest Z28.5 in May.
This indicates that the CCSM4 is staying warmer than observations during the decay of the TSA warm pool in the eastern tropical South Atlantic. This prolonged warming into the springtime is
460 also evident in the seasonal cycle along the equator as observed in Fig. bates #.
Furthermore, in the model the volume of water encompassed by the 28.5°C isotherm is largest in boreal spring (whereas in observations is largest in boreal summer-fall peaking in
September, Figs. 5d-f (1em)). This greater TSA warming (in time and in magnitude) during boreal spring seems to be associated with the pattern of wind stress biases in CCSM4 from
465 December to March. As observed from Fig.bates # the wind stress magnitude in CCSM4 is weaker than in observations in an area collocated with the TSA warm pool in the boreal springtime (for example as in Fig. 7a (3em)). The weaker wind stress in the TSA reduces the evaporative cooling in the region. Furthermore, in the tropical southeastern Atlantic the meridional wind stress in CCSM4 is weaker than in observations thereby reducing the upwelling
470 along the eastern boundary, and reducing the intensity of the Benguela Current, Angola Current and the Atlantic South Equatorial Current systems which normally advect cold water from the south.
20

The April time series of TSA volume (km
3 ) encompassed by the 28.5°C isotherm (i.e., the TSA-WP) are shown in Figure 7 (3em). The volume was calculated from the 5°N latitude to
475 the south and across the basin. The TSA-WP in observations is much smaller than its northern counterpart, the TNA-WP. The TSA-WP volume has periods of relative minima in the early
1950s and the late 1970s. The POP-CORE hindcast has a similar evolution than Ishii. On the other hand, the CCSM4 ensemble simulations have a TSA-WP much larger than observations
(because of the warm bias) and their spread does not encompass the observational estimates.
480 d. Statistics of the Atlantic warm pools
Even though the time series of the CCSM4 ensemble simulations do not correspond to the time series from observations, we can compare basic statistics such as the trend, the standard deviation, and the auto-correlation of the WP time series. Furthermore, a rank histogram was
485 calculated to determine if the CCSM4 ensemble has undervariability with respect to observations and with respect to the POP-CORE simulation.
The trend of each product (including each ensemble simulation) is shown in Table 1. The trend was calculated from the period 1950 to 2005. The September TNA-WP in the CCSM4
20C simulations have a greater trend than that of the observational estimate, and than that of the
490 POP-CORE simulation. However, the April TSA-WP trends in CCSM4 are about the same as those from observations and the POP-CORE simulation. To analyze other statistics the longterm mean and the trends in Table 1em were removed from the corresponding WP time series.
To examine the variability of the Atlantic WPs, the standard deviation and rank histograms were computed from de-mean and de-trended time series. The standard deviations
495 are shown in Table 2. The WP indices from the POP-CORE simulation have the greatest
21

standard deviation. The TNA-WP in September has lower standard deviation in CCSM4 than in
Ishii, but the TSA-WP in April has greater standard deviation in CCSM4 than in Ishii.
Nonetheless, the ensemble spread, as an indication of the intrinsic spread of warm pool realizations, indicates that the spread of the TNA-WP in September can be as large as 25x10
4
500 km
3
at times (e.g., in 1993 and in 2002).
The rank histograms in Fig. 8 (4em) were computed according to Hamill (2001). To create the rank histogram, the five ensemble simulations are considered in addition to an observational estimate (either Ishii or the POP-CORE simulation), thereby having 6 bins in the histogram. For each time step in the WP time series, the ensemble member simulations are
505 ranked from lowest to highest after including the observations. If there is equal probability that the observation will fall in each bin, then the histogram should be uniformly distributed or flat, and one can conclude that on the average, the ensemble spread correctly represents the uncertainty. However, if the histogram is distributed non-uniformly one can refer to either underdispersion or overdispersion of the ensemble, among other categories. From Figure 8
510 (4em) it can be observed that the CCSM4 TNA-WP has underdispersion with respect to both the
Ishii WP and the POP-CORE WP. Yet, the POP-CORE may have excessive interannual variability as indicated by the low auto-correlation values in Table 3. For the TSA-WP, even though the CCSM4 has undervariability with respect to the POP-CORE, the variability with respect to Ishii is unclear.
515 5. Leading modes of Tropical Atlantic Variability
In this section we study the dominant modes of Tropical Atlantic Variability (TAV) by performing a rotated Empirical Orthogonal Function (Rotated EOF, or rEOF) analysis on sea surface temperature (SST) fields. We compare rEOFs from the five 20C CCSM4 ensemble
22

members, the Extended Reconstructed SST version 3b (ERSSTv3b; Smith et al., 2008)
520 observational data set, and the coupled ocean-sea ice experiment forced by the CORE v2 Inter-
Annual Forcing data (Large and Yeager, 2009). The area of interest is the Atlantic Ocean between 30ºS and 30ºN, and the analysis is performed for the period 1948-2005, the era common to these data sets. In contrast to many previous studies, the Caribbean and Gulf of Mexico SSTs are included in these EOF analyses of the tropical Atlantic. First, an EOF analysis is performed
525 on the area-weighted, detrended, monthly anomaly time series of SST, to focus on internal variability and reduce the impact of secular warming trends. The EOFs and their Principal
Components (PCs) are renormalized so that the PCs have unit variance and the EOFs carry the standard deviation. Then a varimax rotation is applied to the dominant 10 EOFs. The rotation technique removes the orthogonality constraint on the EOFs and leads to more localized spatial
530 patterns that might be easier to interpret in terms of dynamical processes (Richman, 1986;
Dommenget and Latif, 2002).
The dominant rEOFs of the observations, the CCSM4 ensemble mean, and the COREforced experiment are shown in Fig. 9 (WW-1), while the spectra of the associated rotated
Principal Components (rPCs) are shown in Fig. 10 (WW-2). The associated levels of variance
535 accounted for by each of these modes (both in relative and absolute sense) are tabulated in Table
4 (WW-1). The dominant modes in observations are the well-known patterns of the Southern
Tropical Atlantic (STA), Northern Tropical Atlantic (NTA) and Subtropical South Atlantic
(SSA) modes (e.g., Huang et al. 2004; Bates 2008, 2010). These modes are clearly represented by the dominant rEOFs of the CCSM4 ensemble members and the CORE-forced experiment,
540 although they account for different levels of variance (Table 4 (WW-1)). Note that the rEOFs are not very well separated, so the relative order of the modes is not of critical importance.
23

The NTA and SSA modes are well represented in CCSM4 (Fig. 9 (WW-1), lower panels). The centers of action are correctly located off West Africa, and in the central South
Atlantic, respectively. The domain-averaged variance of the NTA is underestimated by the
545 ensemble members (0.014 vs. 0.022ºC 2
), making it the weakest mode in all but one of the ensemble members (R008). The variance accounted for by the SSA mode is well represented
(0.019 vs. 0.018ºC 2
). The spectral content of the rPCs of the NTA and SSA modes is consistent with an AR-1 process, as no significant spectral peaks are present in the ensemble mean, nor in the observations (Fig. 10 (WW-2)). Only the NTA mode in the CORE-forced run displays some
550 enhanced energy at the annual frequency.
Lagged correlations between the rPCs and the wind stress (Fig. 11 (WW-3)) shows that both the NTA and SSA modes (lower panels) are associated with a weakening of the trade winds at a 1 month lag, in agreement with observations (e.g., Bates 2008, her Fig. 1b,c). This is consistent with the so-called WES (wind-evaporation-SST) feedback that has been proposed as
555 the dominant mechanism for these modes; a negative wind stress perturbation reduces evaporation and evaporative cooling, inducing a positive SST anomaly that amplifies the wind stress anomaly (Chang et al., 1997; Sterl and Hazeleger, 2003). An interesting result is that a positive SST anomaly in the NTA region is correlated with a strengthening of the southeasterly trade winds in the tropical South Atlantic a few months later. Such a cross-equatorial
560 teleconnection was also found by Bates (2008), and gives credence to the concept of an Atlantic
SST dipole (e.g., Moura and Shukla, 1981).
The largest variability in the observational STA mode (a standard deviation of close to
1ºC) is found off Angola (Fig. 9 (WW-1)). The signal attenuates northward, and achieves amplitudes below 0.3ºC along the equator. This mode of variability is known as the Benguela
24

565
Niño, as observations suggest a generation mechanism similar to the El-Niño phenomenon in the equatorial Atlantic. In particular, observations seem to point at the generation of equatorial
Kelvin waves by wind stress perturbations in the central equatorial Pacific, which subsequently propagate southward along the African coast until they outcrop and generate SST anomalies in the Benguela upwelling zone (e.g., Florenchie et al., 2003, 2004). However, this interpretation is
570 under debate, as modeling studies suggest a dominant role of regional wind stress perturbations in generating coastal upwelling anomalies (e.g., Richter et al., 2010).
The CCSM4 model appears to represent both mechanisms. The dominant rEOF in all but one (R007) of the ensemble members (here called the STA-BG mode) is characterized by strong
(>0.6ºC) SST variability in the southern segment of the Benguela upwelling zone, off the coast
575 of Namibia. The variability extends northward, but does not have an equatorial tongue. The area of highest SST variability corresponds to the region of maximum mean meridional wind stress and a maximum in wind stress bias in the model, compared to observations (Fig. SB-3). Figure
WW-3 shows that a warm phase of this mode is related to a southward wind stress anomaly with a lag of one month, in agreement with the model study of Richter et al. (2010). The spectrum of
580 the corresponding rPC cannot be distinguished from a red-noise process (Fig. 10 (WW-2)).
In contrast, variability in equatorial SSTs (>0.4 ºC) in the CCSM4 ensemble runs is captured by a mode indicated here as STA-EQ. It is highly correlated to wind stress fluctuations in the central equatorial Atlantic about one month earlier (Fig. 11 (WW-3)), suggesting that the mode is maintained by the Bjerkness feedback between zonal equatorial wind anomalies, the tilt
585 of the equatorial thermocline, and the resulting SST anomalies in the eastern tropical Atlantic
(e.g., Keenlyside and Latif, 2007). Wind stress in the central tropical South Atlantic responds to the SST anomaly with a one-month lag. The CORE-forced run has a similar equatorial emphasis
25

of the STA mode, but is lacking an energetic SST variability in the Benguela upwelling region.
This is probably due to an underestimation of wind stress variability resulting from CORE
590 forcing (not shown). The spectrum of the ensemble-mean rPCs display some enhanced energy at a period of about 9 months, a feature that does not seem to be present in the rPC of the observational STA mode (Fig. 10 (WW-2)).
A break-up of the STA in a pattern containing equatorial SST variability and one capturing variability off Angola seems to be characteristic of other coupled models (e.g., Huang
595 et al. 2004; Bates, 2008). Huang et al. (2004) ascribe this disconnection between equatorial and off-equatorial zone to an artificial warm pool created by a southerly bias in the position of the
Intertropical Convergence Zone (ITCZ). In addition, during MAM, when observations indicate the Benguela variability to be most active (Florenchie et al. 2004), the seasonal wind stress bias in CCSM4 is a reduction in magnitude of these southerly coastal winds (Fig. SB-4f).
600 6. Heat budget of the Benguela region
For the purposes of this study the model Benguela region is defined based on the variability of the heat content rate of change (HCR). Standard deviation of anomalous HCR reveals patterns associated with equatorial Kelvin wave and off-equatorial Rossby wave propagation (Fig. 12 (SG-1)). In the east the high values of HCR variability extends poleward
605 from the Equator along the coast reflecting the combined effect of local upwelling and coastal waves. In the east where the thermocline is shallower the regions of high HCR variability are roughly collocated with regions of high SST variability. This is in contrast to the western equatorial Atlantic where SST variability is relatively weak regardless of rather strong HCR variability. The model Benguela region extends from 20°S to the northern edge of the time mean
610
SST front at 13°S (where standard deviation of HCR ≥
250 Wm
-2 ), and from 9°E to the coast of
26

southwestern Africa (Fig. 12 (SG-1)). Although the observed SST front is mostly confined to the meridional extent of the Benguela region, the model SST front is stretched further south (Fig.
12a (SG-1)). This produces one of the strongest and most persistent warm SST biases among various versions of the CCSM (see e.g. Chang et al., 2007). As a result the region of high SST
615 variability is also stretched southward. The model Benguela region, as defined above, covers only the northern part of the high model SST variability zone. This selected northern part is close to the observed region of high SST variability in the Angola-Benguela front (see e.g.
Florenchie et al., 2003).
Anomalous SST in the Benguela region reveals 30 warm events (SSTA>1°C) and 21 cold
620 events (SSTA<-1°C) over the model run’s 1164 months (Fig. 13a (SG-2)). The maximum magnitude of area-averaged anomalous SST is around 2°C which is weaker than in observations
(up to 3°C) (see Fig. 12 (SG-1) for the time-mean model front). The CCSM4 also produces low frequency variability with characteristic periods of 2 to 5 years and a magnitude of up to 0.5°C.
This variability is remotely forced by the Equatorial Pacific as is illustrated by coherent
625 variations of low frequency Benguela SST and anomalous SST in the NINO3 region (Fig. 13a
(SG-2)). Correlation analysis shows that low-frequency SST warming in the Benguela region is related to a basin-scale pattern associated with a warm south tropical Atlantic, weaker than normal southeasterly trade winds, and a southward shift of the ITCZ (northerly cross-equatorial winds). All these changes are typical of development of the Atlantic meridional mode in
630 response to ENSO (Enfield and Mayer, 1997; Xie and Carton, 2004). This forced variability contributes to the Atlantic meridional mode (discussed above in the Section 4), but the latter mode of variability involves wider spectrum of processes including those unrelated to ENSO.
The spatial pattern of SST response stretching from the South African coast westward to Brazil
27

across the southern tropical Atlantic also resembles the second EOF of anomalous mixed layer
635 temperature found by Colberg and Reason (2007a) in model simulations forced by observed winds.
To evaluate relative impacts of heat advection and the surface fluxes on anomalous heat content we next focus on the relationship between the terms of the vertically integrated heat balance equation (1) spatially averaged over the Benguela region. In common notations the
640 vertically-integrated heat balance equation is:
C p
 z z


H

0

T
 t dz

C p
  z z



H
0


 u

T
 x
 v

T
 y
 w

T
 z

 dz

NSHF

VDIFF ( z

H )

R , (1) where NSHF is the net surface heat flux, and vertical integration is taken down to H= 80m that is
645 below mixed layer year around in this region. We focus only on the terms available from in the history files of CCSM4 output, and so the vertical diffusion VDIFF ( z

H ) is combined in
650
R with other unresolved terms that include lateral diffusion, diffusion introduced by mixed layer model, and errors due to using of monthly (instead of model) sampling to calculate the time derivative in (1)
3
.
Heat balance analysis automatically focuses on high frequency variability (e.g. gray line in Fig. 13a (SG-2)), which is emphasized by the time derivative in equation (1). Anomalous SST events in the Benguela region last for approximately 4 months (Fig. 14a (SG-3)). Equation (1) suggests that anomalous heat content rate of change (HCR) is positively correlated with instantaneous anomalous heat advection and surface flux. In general this is confirmed by
3 Heat advection terms in the right hand side of (1) are computed on original model grid using the
POP numerics.
28

655 correlation analysis (Fig. 14a (SG-3)), which identifies ocean heat advection as the dominant contribution to the heat budget in Benguela region. The largest influence is provided by vertical heat advection (upwelling). The magnitude of its correlation with anomalous HCR at zero lag
(
 
0 ) exceeds 0.7. Vertical advection accounts for 51% of the anomalous HCR variance. The second strongest contribution is from anomalous meridional heat advection
4
660 (CORR(
 
0 )=0.5), which accounts for about 26% of the anomalous HCR variance. The impact of zonal advection is weak due to predominantly zonal orientation of isotherms in the
Angola-Benguela front. Local surface flux accounts for only 12% of the anomalous HCR variance. Surface flux also provides a weak negative feedback on anomalous SST in two months
665 after the peak of HCR via latent heat flux.
To explore possible atmospheric forcing of anomalous heat advection in the Benguela region, the area-averaged time series of anomalous vertical and meridional advection have been lag-correlated with wind stress anomalies elsewhere. This analysis illustrates that anomalously warm vertical advection in the Benguela region (reduced upwelling) occurs in phase with weakening of southeasterly trade winds (Fig. 14b (SG-3)). The maximum correlation is at zero
670 lag, suggesting that impact of local upwelling dominates. This is evident in an anomalous cyclonic wind pattern driven by an anomalously weak southern Atlantic subtropical high (Fig.
14b (SG-3)). The anomalous wind pattern includes a northerly (downwelling) component along the coast. Although the upwelling is attenuated along a major portion of the South African coast its impact on SST is stronger in the Benguela region where the thermocline shoals. An
675 interesting (but not yet well understood) feature of the air pressure pattern is the area of anomalously high pressure over South Africa. A zonal gradient between the high over land and
4 Anomalous meridional heat advection is dominated by an anomalous meridional current acting on the mean meridional gradient of temperature (Colberg and Reason, 2007b).
29

the low over the ocean further accelerates anomalous downwelling winds along the coast. An increase in air pressure over the land during warm Benguela events may be linked to cooling of the land due to above-average rainfall along the coast of Angola and Namibia observed by
680 Rouault et al. (2003). But in CCSM4 the Benguela SST does not correlate significantly with either land temperature or rainfall.
The wind pattern corresponding to anomalous meridional advection (Fig. 14c (SG-3)) is different from that in Fig. 14b (SG-3) in many aspects. The area of weaker trades does not cover the Benguela region itself. There is significant correlation with the zonal Equatorial winds that
685 lead the meridional advection by about a month. This suggests that non-local processes translating wind impacts from the Equatorial region (such as equatorial and coastal Kelvin waves) are responsible for anomalous meridional heat advection in the Benguela region.
Dominant roles of vertical and meridional heat advection is expectable due to the local coastal upwelling and the presence of sharp meridional SST gradient in the Angola-Benguela
690 front. Impact of local and remote winds on the heat budget of the Benguela region seem hold on longer time scales. In particular, Chang et al. (2007) analysis emphasizes the role of anomalously weak equatorial easterly wind. In the ocean this zonal wind bias leads to an erroneous deepening of the equatorial thermocline that is extended poleward along the coast, thus decreasing cooling effect of local upwelling. Predictably, this warm SST bias is reduced if the model equatorial
695 winds are strengthened (Richter et al., 2011). But local coastal upwelling is also anomalously weak in CCSM4. This is evident in below normal southerly wind component right next to the
African coast 30S-20S (Fig. 2d). Weak local upwelling causes warm SST bias in at least two ways. It not only reduces the cooling by vertical advection, but it also affects the meridional heat transport by coastal branch of the cold Benguela Current that is maintained by the cross-coastal
30

700 gradient of sea level (see Grodsky et al., 2011 for more details).
7. Summary
In this study we analyze the variability in the CCSM4 with respect to some of the main aspects of the tropical Atlantic variability (TAV). Various analyses are presented and discussed covering the variability of the heat budget in the Benguela region, the variability of sea surface
705 temperatures, and the basic structure of the tropical Atlantic waters warmer than 28.5°C.
We have performed different sets of analyses to address differences and improvements achieved by the CCSM4 model in the tropical Atlantic Ocean. The analyses and results presented and discussed above will be useful for further evaluations of CCSM4 simulations of the tropical Atlantic climate and for predictive studies of such region. Following we present a
710 summary of our results by section. a. Biases in SST and wind stress
The eastern boundary upwelling region warm biases are due to a combination of ocean, atmosphere, and coupling processes. These regions are characterized by weak ocean currents, weak upwelling, weak along-shore wind, too little stratus cloud, and neighboring mountainous
715 regions. All of these features in combination create SSTs that are too warm. More details on this region and the effects of each of the contributing features are discussed in Large and
Danabasoglu (2006). Additionally, Gent et al. (2008) discusses the improvement in these biases with increased atmospheric resolution, one of the primary reasons a nominal 1 degree atmosphere is used in the coupled model (Gent et al., 2011, this issue).
720 b. Atlantic warm pools
In this set of analyses the warm pool is analyzed by its vertical structure throughout the year in both the tropical North Atlantic (TNA) including the Intra-Americas Sea, and the tropical
31

South Atlantic (TSA) from the new CCSM4 simulations. Furthermore, an observational data set with subsurface temperature estimates for the recent decades, and an ocean-only POP simulation
725 forced with observed wind stress and surface fluxes are used to compare with the CCSM4 simulations spanning the period 1950-2005.
The volume of the WP in the tropical South Atlantic (TSA-WP) peaks in April and in the tropical North Atlantic (TNA-WP) peaks in September. The timing of the WPs in CCSM4 is similar to that of the observations, although the vertical structure indicates that the TSA-WP pool
730 is deeper and wider in the CCSM4 than in observations. This deeper TSA-WP is related to the
CCSM4 warm bias in the TSA region, a common challenge to many coupled models.
Regardless of the warm bias, the ensemble spread of the TSA-WP seems to correctly represent the uncertainty. In the IAS, the CCSM4 warm pool is smaller than in observations, as a result of the CCSM4 cold bias in the southern Caribbean Sea and to the northeast of the Caribbean Sea.
735 The ensemble spread of the TNA-WP is underdispersed compared to the observations. c. Modes of Tropical Atlantic variability
Rotated Empirical Orthogonal Functions (rEOFs) were applied to the sea surface temperature (SSTs) fields of the various ensemble simulations and to an observational data set for the period 1950-2005. The spatial patterns of the main modes of variability in the model are
740 similar to that from the observations. However, the leading rEOF (rEOF1) in the simulations does not have a counterpart in the observations. d. Heat budget of Benguela region
An analysis of approximately 100-year long records from the CCSM4 control run reveals realistic values of anomalous SST in the Benguela region that varies interannually to decadally.
745
The maximum magnitude of interannual events reaches 2°C, which is somewhat smaller than in
32

observations (up to 3°C). This lack of variability is attributed to the shape of the simulated
Angola-Benguela front, which is not as sharp as it is in observations. The CCSM4 also produces interannual (2 to 5 years) variability of Benguela SST. This lower frequency variability is up to
0.5°C and is remotely forced by ENSO.
750 Analysis of the model heat budget in the Benguela region suggests that anomalous vertical advection accounts for about 50% of the anomalous heat content rate variance while the contribution by anomalous meridional heat advection is half as strong. Local surface flux accounts for only 12% of the anomalous HCR variance. The impact of zonal advection is weak.
Anomalously warm vertical advection in the Benguela region (reduced upwelling) occurs
755 in phase with the weakening of southeasterly trade winds. Correlation is maximum at zero lag suggesting that the impact of local upwelling dominates. In contrast to vertical heat advection the anomalous meridional heat advection is forced by zonal equatorial winds, which lead it by about a month. This suggests that non-local processes translating wind impacts from the
Equatorial region (such as Equatorial and coastal Kelvin waves) are responsible for anomalous
760 meridional heat advection in the Benguela region. In distinction from observations of Florenchie et al. (2003) this wave-based teleconnection is not the dominant mechanism of heat content variability in the Benguela region in CCSM4.
765
8. Acknowledgements
We thank all the scientists and software engineers who contributed to the development of the CCSM4. Computational resources were provided by the Climate Simulation Laboratory at
NCAR’s Computational and Information Systems Laboratory (CISL), sponsored by the National
Science Foundation and other agencies. The CCSM is also sponsored by the Department of
33

Energy. We also wish to thank the staff of the Earth System Grid (including Gary Strand from
770 NCAR) and Matthew Maltrud (from LANL) for the support and contribution in downloading and facilitating data used in this study. The Earth System Grid is funded by the U.S. Department of Energy (DOE). Ernesto Muñoz and Wilber Weijer were supported by the Regional and
Global Climate Prediction Program of the DOE Office of Science, and by NSF-OCE award
0928473. Semyon Grodsky was supported by NOAA Climate Variability and Predictability
775 (CVP) Program. We thank Ilana Wainer, Marlos Goes and two anonymous reviewers for helpful comments.
34

REFERENCES
780 Barreiro, M., A. Giannini, P. Chang, and R. Saravanan, 2004: On the Role of the South Atlantic
Atmospheric Circulation in Tropical Atlantic Variability, in Earth’s Climate: The Ocean-
Atmosphere Interaction, pp. 143–152, AGU Geophysical Monograph Series 147.
785
Barreiro, M., P. Chang, L. Jia, R. Saravanan, and A. Giannini, 2005: Dynamical elements of predicting boreal spring tropical Atlantic sea-surface temperatures. Dynamics of
Atmospheres and Oceans.
Bates, S.C., 2008: Coupled ocean-atmosphere interaction and variability in the tropical Atlantic
Ocean with and without an annual cycle, J. Clim ., 21 , 5501-5523.
Bates, S.C., 2010: Seasonal influences on coupled ocean-atmosphere variability in the tropical
Atlantic Ocean, J. Clim ., 23 , 582-604.
790 Bates, S.C., B. Fox-Kemper, S.R. Jayne, W.G. Large, S. Stevenson, and S.G. Yeager, 2011:
Mean biases, variability, and trends in air-sea fluxes and upper-ocean in the CCSM4, J.
Clim ., submitted.
Breugem, W.-P., P. Chang, C. J. Jang, J. Mignot, and W. Hazeleger, 2008: Barrier layers and tropical Atlantic SST biases in coupled GCMs. Tellus A , 60 , 885-897.
795 Carton, J. A., and B. Huang, 1994: Warm Events in the Tropical Atlantic, J. Phys. Oceanogr .,
24 , 888–903.
Carton, J. A., X. Cao, B. S. Giese, and A. M. da Silva, 1996: Decadal and interannual SST
800 variability in the tropical Atlantic Ocean, J. Phys. Oceanogr ., 26 , 1165-1175.
Chang, P., L. Ji, and H. Li, 1997: A decadal climate variation in the tropical Atlantic Ocean from thermodynamic air-sea interactions, Nature , 385 , 516-518.
35

Chang, C.-Y., J. A. Carton, S. A. Grodsky, and S. Nigam, 2007: Seasonal climate of the tropical
Atlantic sector in the NCAR Community Climate System Model 3: Error structure and probable causes of errors. J. Clim ., 20 ,1053-1070.
Colberg, F. and C. J. C. Reason, 2007a: Ocean Model Diagnosis of Low-Frequency Climate
805 Variability in the South Atlantic Region, J. Clim ., 20 , 1016-1034.
Colberg, F. and C. J. C. Reason, 2007b: A model investigation of internal variability in the
Angola Benguela Frontal Zone, J. Geophys. Res ., 112 , C07008, doi:10.1029/2006JC003920.
Czaja, A., P. van der Vaart, and J. Marshall, 2002: A Diagnostic Study of the Role of Remote
810 Forcing in Tropical Atlantic Variability, J. Clim ., 15 , 3280-3290.
Dai, A., K. Trenberth, and T. Qian, 2004: A global dataset of palmer drought severity index for
1870-2002: Relationship with soil moisture and effects of surface warming. J.
Hydrometeor.
, 5 , 1117–1130.
Danabasoglu, G., S. Bates, B. P. Briegleb, S. R. Jayne, M. Jochum, W. G. Large, S. Peacock, and
815 S. G. Yeager, 2011: The CCSM4 ocean component. J. Clim ., this issue, submitted.
Danabasoglu, G., S. G. Yeager, Y. -O. Kwon, J. W. Hurrell, A. S. Phillips, and J. J. Tribbia
2011: Variability of the Atlantic Meridional Overturning Circulation in CCSM4 , J. Clim ., submitted.
Davey, M. K., and coauthors, 2002: STOIC: a study of coupled model climatology and 19
820 variability in tropical ocean regions. Climate Dyn., 18 , 403-420.
Deser, C., A. Capotondi, R. Saravanan, and A. S. Phillips, 2006: Tropical Pacific and Atlantic
Climate Variability in CCSM3. J. Clim ., 19 , 2451-2481.
Ding, H., N.S. Keenlyside, M. Latif, 2011: Impact of the equatorial Atlantic on the El Nino
36

Southern Oscillation. Climate Dynamics , in press.
825 Doi, T., T. Tozuka, T. Yamagata, 2010: The Atlantic Meridional Mode and Its Coupled
Variability with the Guinea Dome. J. Climate , 23 , 455–475. doi:
10.1175/2009JCLI3198.1
830
Doi, T., G. A. Vecchi, A. J. Rosati, and T. L. Delworth, 2011: Tropical Atlantic biases in the mean state, seasonal cycle, and interannual variations for a coarse and a high resolution coupled climate model. J. Climate , submitted.
Dommenget, D., and M. Latif, 2000. Interannual to decadal variability in the tropical Atlantic. J.
Clim. 13 , 777–792.
Dommenget, D. and M. Latif, 2002. A cautionary note on the interpretation of EOFs. J. Climate ,
15 , 216-225.
835 Enfield, D., and D. Mayer, 1997: Tropical Atlantic sea surface temperature variability and its relation to El Niño - Southern Oscillation.
J. Geophys. Res ., 102 , 929-945.
Enfield, D. B., A. M. Mestas-Nuñez, and P. J. Trimble, 2001: The Atlantic Multidecadal
Oscillation and its relation to rainfall and river flows in the continental U.S., Geophys.
Res. Lett., 28 , 2077-2080.
840 Enfield, D. B., and S.-K. Lee, 2005: The heat balance of the western hemisphere warm pool, J.
Clim ., 18 , 2662-2681, doi:10.1175/JCLI3427.1.
Florenchie, P., J. R. E. Lutjeharms, C. J. C. Reason, S. Masson, and M. Rouault, 2003: The source of Benguela Niños in the South Atlantic Ocean,
Geophys. Res. Lett ., 30 , 1505, doi:10.1029/2003GL017172.
845 Florenchie, P., C. J. C. Reason, J. R. E. Lutjeharms, and M. Rouault, 2004: Evolution of
Interannual Warm and Cold Events in the Southeast Atlantic Ocean, J. Clim.
, 17 , 2318–
37

2334.
Foltz, G. R., M. J. McPhaden, 2010: Interaction between the Atlantic meridional and Niño modes. Geophys. Res. Lett., 37 , L18604, doi:10.1029/2010GL044001
850 Garzoli, S., and J. Servain, 2003: CLIVAR workshop on tropical Atlantic variability. Geophys.
Res. Lett.
, Vol. 30 , No. 5, 8001, doi:10.1029/2002GL016823
Gent, P. R., and J. McWilliams, 1990: Isopycnal mixing in ocean circulation models, J. Phys.
Oceanogr., 20 , 150–156.
855
Gent, P.R., G. Danabasoglu, L.J. Donner, M.M. Holland, E.C. Hunke, S.R. Jayne, D.M.
Lawrence, R.B. Neale, P.J. Rasch, M. Vertenstein, P.H. Worley, Z.-L. Yang, and M.
Zhang, 2011: The Community Climate System Model version 4, J. Clim ., accepted.
Grodsky, S.A., J. A. Carton, S. Nigam, and Y. Okumura, 2011: Tropical Atlantic Biases in
CCSM4, J. Clim., submitted to special CCSM4 issue.
860
Hamill, T. M., 2001: Interpretation of rank histograms for verifying ensemble forecasts. Mon.
Wea. Rev ., 129 , 550–560.
Hazeleger, W., P. de Vries, and Y. Friocourt, 2003: Sources of the equatorial undercurrent in the
Atlantic in a high-resolution ocean model. J. Phys. Oceanogr ., 33 , 677–693.
Hirst, A.C. and S. Hastenrath, 1983: Atmosphere-ocean mechanisms of climate anomalies in the
Angola-tropical Atlantic sector, J. Phys. Oceanogr ., 13 , 1146-1157.
865 Hu, Z.-Z., and B. Huang, 2007: Physical Processes Associated with the Tropical Atlantic SST
Gradient during the Anomalous Evolution in the Southeastern Ocean, J. Clim ., 20 , 3366–
3378.
Hu, Z.-Z., B. Huang, and K. Pegion, 2008: Leading patterns of the tropical Atlantic variability in a coupled general circulation model. Clim. Dyn ., 30 , 703-726. Doi: 10.1007/s00382-007-
38

870 0318-x
Huang, B., P. S. Schopf, and J. Schukla, 2004: Intrinsic Ocean–Atmosphere Variability of the
Tropical Atlantic Ocean, J. Clim ., 17 , 2058–2077.
Huang, B. and J. Shukla, 2005: Ocean-atmosphere interactions in the tropical and subtropical
Atlantic Ocean, J. Clim ., 18 , 1652-1672.
875 Hurrell, J. W., J. J. Hack, D. Shea, J. M. Caron, and J. Rosinski, 2008: A new sea surface temperature and sea ice boundary dataset for the Community Atmosphere Model. J.
Clim ., 21 , 5145–5153, doi:10.1175/2008JCLI2292.1.
Hurrell, J.W., M. Visbeck, A. Busalacchi, R. A. Clarke, T. L. Delworth, R. R. Dickson, W. E.
880
Johns, K. P. Koltermann, Y. Kushnir, D. Marshall, C. Mauritzen, M. S. McCartney, A.
Piola, C. Reason, G. Reverdin, F. Schott, R. Sutton, I. Wainer, and D. Wright, 2006:
Atlantic Climate Variability and Predictability: A CLIVAR Perspective. J. Clim ., 19 ,
5100-5121.
Huttl-Kabus, S., and C. Boning, 2008: Pathways and variability of the off-equatorial undercurrents in the Atlantic Ocean, J. Geophys. Res ., 113 (C10), C10, 018.
885 Ishii, M., M. Kimoto, K. Sakamoto, and S.-I. Iwasaki, 2006: Steric Sea Level Changes Estimated from Historical Ocean Subsurface Temperature and Salinity Analyses. J. Oceanogr ., Vol.
62 (No. 2), pp. 155-170.
Johns, W. E., R. Zantopp, and G. Goni, 2003: Cross-gyre transport by North Brazil Current
890 rings, in Interhemispheric Water Exchange in the Atlantic Ocean, Elsevier Oceanogr.
Ser., vol. 68, edited by G. J. Goni and P. Malanotte-Rizzoli, pp. 411–442, Elsevier,
Amsterdam, The Netherlands.
Jochum, Markus
39

Keenlyside, N. S., and M. Latif, 2007: Understanding equatorial Atlantic interannual variability.
895 J. Clim., 20 , 131-142.
Kilbourne, K. H., T. M. Quinn, T. P. Guilderson, R. S. Webb, and F. W. Taylor, 2007: Decadal-
900 to interannual-scale source water variations in the Caribbean Sea recorded by Puerto
Rican coral radiocarbon. Clim. Dyn ., 29 , 51-62.
Kröger, J., A. J. Busalacchi, J. Ballabrera-Poy, and P. Malanotte-Rizzoli, 2005: Decadal variability of shallow cells and equatorial sea surface temperature in a numerical model of the Atlantic, J. Geophys. Res ., Vol. 110 , No. C12, C12003 10.1029/2004JC002703.
Kucharski F, A. Bracco, J.H. Yoo, F. Molteni, 2007: Low-Frequency variability of the Indian monsoon–ENSO relationship and the tropical Atlantic: the ‘‘weakening’’ of the 1980s and 1990s. J Clim, 20, 4255–4266.
905 Kucharski, F., A. Bracco, J. Y. Yoo, and F. Molteni, 2008: Atlantic forced component of the
Indian monsoon interannual variabiity, Geophys. Res. Lett., 35, doi:10.1029/2007GL033037
Kucharski F, A. Bracco, J.H. Yoo, A. Tompkins, L. Feudale, P. Ruti, A. Dell’Aquila, 2009: A
Gill-Matsun-type mechanism explains the Tropical Atlantic influence on African and
910 Indian Monsoon rainfall. Quart J R Met Soc, 135, 569–579
Large, W. G., and G. Danabasoglu, 2006: Attribution and Impacts of Upper-Ocean Biases in
CCSM3, J. Clim ., 19 , 2325-2346.
Large, W., and S. Yeager, 2009: The global climatology of an interannually varying air–sea flux data set. Clim. Dyn.
, 33 , 341–364.
915 Laurian, A., and S. Drijfhout, 2010: Response of the South Atlantic circulation to an abrupt
40

collapse of the Atlantic Meridional Overturning Circulaiton. Clim. Dyn.
,
DOI:10.1007/s00382-010-0890-3.
Lazar, A., T. Inui, A. J. Busalacchi, P. Malanotte-Rizzoli, and L. Wang, 2002: Seasonality of the ventilation of the tropical Atlantic thermocline., J. Geophys. Res., 107 , 181–187.
920 Lee, S.-K., D. B. Enfield, and C. Wang, 2007: What drives the seasonal onset and decay of the western hemisphere warm pool? J. Clim ., 20 , 2133–2146, doi:10.1175/JCLI4113.1
Levitus, S., T. P. Boyer, M. E. Conkright, T. O'Brien, J. Antonov, C. Stephens, L. Stathoplos, D.
Johnson, & R. Gelfeld, 1998: World Ocean Database 1998 Volume 1: Introduction.
NOAA Atlas NESDIS 18, U.S. Government Printing Office, Washington, D.C.
925 Locarnini, R. A., A. V. Mishonov, J. I. Antonov, T. P. Boyer, and H. E. Garcia, 2010. World
Ocean Atlas 2009, Volume 1: Temperature. S. Levitus, Ed. NOAA Atlas NESDIS 68,
U.S. Government Printing Office, Washington, D.C., 184 pp. (URL: ftp://ftp.nodc.noaa.gov/pub/WOA09/DOC/woa09_vol1_text.pdf
)
Losada, T., B. Rodrıguez-Fonseca, I. Polo. S. Janicot, S. Gervois, F. Chauvin, P. Ruti, 2009:
930 Tropical response to the Atlantic Equatorial mode: AGCM multimodel approach, Clim
Dyn., DOI 10.1007/s00382-009-0624-6
Lübbecke, J. F., C. W. Böning, N. S. Keenlyside, and S.-P. Xie, 2010, On the connection between Benguela and Equatorial Atlantic Niños and the role of the South Atlantic
Anticyclone, J. Geophys. Res. Oce., 115, C09015
935 Mahajan, S., R. Saravanan, and P. Chang, 2010: Free and Forced Variability of the Tropical
Atlantic Ocean: Role of the Wind–Evaporation–Sea Surface Temperature Feedback. J.
Clim., 23 , 5958–5977. doi: 10.1175/2010JCLI3304.1
Malanotte-Rizzoli, P., K. Hedstrom, H. Arango, and D. Haidvogel: 2000: Water mass pathways
41

940 between the subtropical and tropical ocean in a climatological simulation of the North
Atlantic Ocean circulation., Dyn. Atmos. Oceans , 32 , 331–371.
Maximenko, N. A., and P. P. Niiler, 2005: Hybrid decade-mean global sea level with mesoscale resolution. In N. Saxena (Ed.) Recent Advances in Marine Science and Technology,
2004, pp. 55-59. Honolulu: PACON International
( http://apdrc.soest.hawaii.edu/projects/DOT/index.html
)
945 Meehl, G. H., et al., 2007: Global climate projections. In Climate Change 2007: The Physical
Science Basis: Contribution of Working Group I to the Fourth Assessment Report of the
Intergovernmental Panel on Climate Change, edited by S. Solomon et al., pp. 747–846,
950
Cambridge Univ. Press, Cambridge, U. K.
Misra, V., S. Chan, R. Wu, and E. Chassignet, 2009: Air-sea interaction over the Atlantic warm pool in the NCEP CFS. Geophys. Res. Lett ., Vol. 36 , L15702, doi:10.1029/2009GL038737.
Mo, K. C., M. Chelliah, M. L. Carrera, R. W. Higgins, and W. Ebisuzaki, 2005: Atmospheric
Moisture Transport over the United States and Mexico as Evaluated in the NCEP
Regional Reanalysis. J. Hydrometeor , 6 , 710–728.
955 Molinari, R., S. Bauer, D. Snowden, G. Johnson, B. Bourles, Y. Gouriou, and H. Mercier, 2003:
A comparison of kinematic evidence for tropical cells in the Atlantic and Pacific oceans.
Elsevier Oceanography Series, 68 , 269–286.
Moura and Shukla, JAS 1981
960 Muñoz, E., C. Wang, and D. Enfield, 2010: The Intra-Americas springtime sea surface temperature anomaly dipole as fingerprint of remote influences. J. Clim ., 23 , 43-56.
42

Muñoz, E., B. Kirtman, and W. Weijer, 2011: Varied representation of the Atlantic Meridional
Overturning Circulation in ocean reanalyses. Deep-Sea Research, Part II . In press. Url : http://dx.doi.org/10.1016/j.dsr2.2010.10.064
965
Muñoz, E., A. J. Busalacchi, S. Nigam, and A. Ruiz-Barradas, 2008: Winter and summer structure of the Caribbean low-level jet. J. Clim., 21 , 1260-1276.
Murtugudde, R., J. Ballabrera, J. Beauchamp, and A. Busalacchi, 2002: Relationship between zonal and meridional modes in the tropical Atlantic. Geophys. Res. Lett.
, 22 , 4463–4466.
970
Nicholson, S.E., 2010: A low-level jet along the Bengueal coast, an intergral part of the
Benguela current ecosystem. Clim. Change , 99 , 613-624.
Niiler, P. P., N. A. Maximenko, and J. C. McWilliams, 2003: Dynamically balanced absolute sea level of the global ocean derived from near-surface velocity observations. Geophys. Res.
Lett ., 30 (22), 2164, doi:10.1029/2003GL018628.
975
Okumura, Y., and S.-P. Xie, 2006: Some overlooked features of tropical Atlantic climate leading to a new Niño-like phenomenon.
J. Clim ., 19 , 5859-5874.
Reynolds, R. W., N. A. Rayner, T. M. Smith, D. C. Stokes, and W. Wang, 2002: An Improved In
Situ and Satellite SST Analysis for Climate, J. Clim ., 15 , 1609–1625.
Richman, M. B., 1986: Rotation of principal components. J. Clim., 6 , 293-335.
980
Richter, I., and S.-P. Xie, 2008: On the origin of equatorial Atlantic biases in coupled general circulation models. Clim. Dyn ., 31 , 587-598.
Richter, I., S. K. Behera, Y. Masumoto, B. Taguchi, N. Komori, and T. Yamagata, 2010: On the triggering of Benguela Niños: Remote equatorial versus local influences,
Geophys. Res.
Lett ., 37 , L20604, doi: 10.1029/2010GL044461.
Richter, I., S.-P. Xie, A. T. Wittenberg, and Y. Masumoto, 2011: Tropical Atlantic biases and
43

985 their relation to surface wind stress and terrestrial precipitation. Clim. Dyn ., in press, doi://10.1007/s00382-011-1038-9.
Robertson, A. W., J. D. Farrara, and C. R. Mechoso, 2003: Simulations of the Atmospheric
Response to South Atlantic Sea Surface Temperature Anomalies, J. Clim ., 16 , 2540–
2551.
990 Rouault, M., 2010: Intrusion of tropical water in the Benguela upwelling, a numerical study,
Tropical Atlantic and PIRATA-15 meeting 2-5 March 2010 Miami, Florida, http://www.aoml.noaa.gov/phod/pne/pirata15/Rouault.pdf.
Rouault, M., P. Florenchie, N. Fauchereau, and C. J. C. Reason, 2003: South East tropical
995
Atlantic warm events and southern African rainfall, Geophys. Res. Lett., 30 , 8009, doi:10.1029/2002GL014840.
Rouault, M., S. Illig, C. Bartholomae, C. J. C. Reason, and A. Bentamy, 2007: Propagation and origin of warm anomalies in the Angola Benguela upwelling system in 2001, J. Mar.
Systems , 68 , 473–488.
1000
Rouault, M., J. Servain, C. J. C. Reason, B. Bourl`es, M. J. Rouault, and N. Fauchereau, 2009:
Extension of PIRATA in the tropical South-East Atlantic: an initial one-year experiment,
Afr. J. Mar. Sci ., 31 , 63–71.
Ruiz-Barradas, A., J. A. Carton, and S. Nigam, 2000: Structure of Interannual-to-Decadal
Climate Variabilty in the Tropical Atlantic Sector, J. Clim ., 13 , 3285–3297.
1005
Saravanan, R., and P. Chang, 2000: Interactions between the Pacific ENSO and tropical Atlantic climate variability. J. Clim.
, 13 , 2177-2194.
Schouten, M. W., R. P. Matano, and T. P. Strub, 2005: A description of the seasonal cycle of the equatorial Atlantic from altimeter data, Deep-Sea Res. I , 52 , 477–493.
44

Servain, J., I. Wainer, J. McCreary, A. Dessier, 2001: Relationship between the equatorial and meridional modes of climatic variability in the tropical Atlantic. Geophys. Res. Lett ., 26 ,
1010 485–488.
Servain, J., 1991: Simple Climatic Indices for the Tropical Atlantic Ocean and Some
Applications, J. Geophys. Res ., 96 , 15137–15146.
Shannon, L. V., A. J. Boyd, G. B. Bundrit, and J. Taunton-Clark, 1986: On the existence of an El
Niño–type phenomenon in the Benguela system, J. Mar. Sci.
, 44 , 495–520.
1015 Shannon, L. V., J. J. Agenbag, and M. E. L. Buys, 1987: Large and mesoscale features of the
Angola-Benguela front, South Afr. J. Mar. Sci ., 5 , 11–34.
Smith, T. M., R. W. Reynolds, T. C. Peterson, and J. Lawrimore, 2008: Improvements to
NOAA’s Historical Merged Land-Ocean Surface Temperature Analysis (1880-2006).
J.
Clim ., 21 , 2283-2296.
1020 Snowden, D., and R. Molinari, 2003: Subtropical cells in the Atlantic Ocean: An observational summary, Elsevier Oceanography Series, 68 (ISBN 0-444-51267-5).
Solomon, A., and I. Wainer, 2006: Pacific tropical-extratropical thermocline water mass exchanges in the NCAR CCSM3. Ocean Modelling , doi: 10.1016/j.ocemod.2006.04.003.
Steele, M., R. Morley, and W. Ermold, 2001: PHC: A global ocean hydrography with a high-
1025 quality Arctic Ocean. J. Clim.
, 14 , 2079–2087.
Sterl, A., and W. Hazeleger, 2003: Coupled variability and air-sea interaction in the South
1030
Atlantic Ocean, Clim. Dyn ., 21 , 559–571, doi: 10.1007/s00,382–003–0348–y.
Stramma, L., and M. England, 1999: On the water masses and mean circulation of the South
Atlantic Ocean, J. Geophys. Res. (Oceans) , 104 , 20863–20884, doi:
10.1029/1999JC900139.
45

Talley, L., 1996: Antarctic intermediate water in the South Atlantic. In The South Atlantic:
Present and Past Circulation, pp. 219–238.
Tian, B., G. J. Zhang, and V. Ramanathan, 2001: Heat Balance in the Pacific Warm Pool
Atmosphere during TOGA COARE and CEPEX. J. Climate , 14 , 1881-1893.
1035 Tokinaga, H., and S.-P. Xie, 2011: Weakening of the equatorial Atlantic cold tongue over the past six decades. Nature Geoscience, 4 (4), 222-226, doi:10.1038/ngeo1078.
Tozuka, T., T. Doi, T. Miyasaka, N. Keenlyside, T. Yamagata, 2011: Key factors in simulating the equatorial Atlantic zonal sea surface temperature gradient in a coupled general circulation model. J. Geophys. Res., 116, C06010, 12 PP., 2011
1040 doi:10.1029/2010JC006717.
Visbeck, M., G. Reverdin, F. Schott, J. Carton, A. Clarke, B. Owens, and D. Stammer, 2001:
Atlantic Climate Variability Experiment, in "Observing the Oceans in the 21st Century" edited by C.J. Koblinsky and N. R. Smith, GODAE project office, Melbourne, 445-452. von Storch, H., 2002: Statistical analysis in climate research. Cambridge University Press, 496
1045 pages.
Wahl, S., M. Latif, W. Park, and N. Keenlyside, 2011: On the Tropical Atlantic SST warm bias in the Kiel Climate Model. Clim. Dyn.
, 36 , 891-906.
Wainer, I., A. Lazar, and A. Solomon, 2006: Tropical extra-tropical thermocline water mass exchanges in the Community Climate Model v.3 Part I: the Atlantic Ocean. Ocean
1050 Science , 2 , 137-146.
Wang, C., and D. B. Enfield, 2003: A further study of the tropical western hemisphere warm pool. J. Clim ., 16 , 1476– 1493.
Wang, C., S.-P. Xie, and J.A. Carton, 2004: A global survey of ocean-atmosphere interaction and
46

1055 climate variability.
Earth’s Climate: The Ocean-Atmosphere Interaction, Geophys.
Momogr ., Vol. 147, Amer. Geophys. Union, 1-19.
Wang, C., D. B. Enfield, S.-K. Lee, and C. W. Landsea, 2006: Influences of Atlantic Warm Pool on western hemisphere summer rainfall and Atlantic hurricanes. J. Clim ., 19 , 3011-3028.
Wang, C., S.-K. Lee, and D. Enfield, 2008: Climate response to anomalously large and small
Atlantic warm pools during the summer. J. Clim ., 21 , 2437-2450.
1060 Xie, S.-P., and J. A. Carton, 2004: Tropical Atlantic variability: patterns, mechanisms, and impacts, in “Ocean-Atmosphere Interaction and Climate Variability”, edited by C. Wang,
S.-P. Xie, and J. A. Carton, AGU Press.
1065
Zebiak, S. E., 1993: Air-sea interaction in the equatorial Atlantic region, J. Clim ., 6 , 1567–1686.
Zhang, D., M. J. McPhaden, and W. E. Johns, 2003: Observational evidence for flow between the Subtropical and Tropical Atlantic: the Atlantic Subtropical Cells. J. Phys. Oceanogr.,
33 , 1783–1797.
Zuidema, P., P. Chang, C.R. Mechoso, and L. Terray, 2011: Coupled ocean-atmosphere-land processes in the tropical Atlantic, Clivar Exchanges , 16 (1), 12-14.
47

1070 CAPTIONS
Figure 1: Sea surface temperature (°C, left panels) and salinity (psu, right panels) model minus observations differences for the period of 1980-1999 for CCSM3 and 1986-2005 for CCSM4.
These plots correspond to Figure 6 from Danabasoglu et al. (2011, this issue) with a focus on the
1075 tropical Atlantic.
Figure 2: (top) Total mean of zonal and meridional wind stress from observations. The middle panel are the differences of these observations from CCSM4, and bottom panel are the differences with CCSM3. The units are N/m
2
and the time period matches those for the means in
1080 Figure 1.
Figure 3: a) Mean SST along the equator from observations (black line), CCSM4 (red line), and
CCSM3 (blue line). Seasonal cycle of SST along the equator calculated as the mean of each month minus the total mean from observations (b), CCSM4 (c), and CCSM3 (d). Units are °C
1085 and the means are calculated over 1980-1999 for CCSM3 and 1986-2005 for CCSM4. The mean from observations spans both periods including 1980-2005.
Figure 4: The left panels are the seasonal mean of the wind stress (vectors) and their magnitude
(shades); the right panels are the differences of these observations from CCSM4 wind stress. The
1090 units are N/m 2 , and the time period used is 1986-2005.
Figure 5 ( 1em ): (a-d) Horizontal distribution of the month of deepest 28.5°C isotherm from the
48

long-term mean from 1950 to 2005. The numbers 1 to 12 correspond to the months from
January to December. The Pacific data has been masked. Panel (a) corresponds to the CCSM4
1095 ensemble mean. Panel (c) corresponds to the POP ocean model forced with CORE surface forcing. Panels (b) and (d) correspond to the observational products, Ishii and Levitus, respectively. (e) Seasonal cycle of the volume of the 28.5°C isotherm between 40°S-40°N and above 250 meters of depth.
1100 Figure 6 ( 2em) : The tropical South Atlantic (TSA) Warm Pool in April. (a-d) Mean depth
(meters) of the 28.5°C isotherm in April. The CCSM4 ensemble mean (panel a) is the mean of five different simulations. (e) Time series of the volume (10
4
km
3 ) encompassed by the 28.5°C isotherm in April south of 5°N. The black line is the Ishii observational product; the blue line is the ocean POP simulation forced by CORE forcing; the red line is the CCSM4 ensemble mean
1105 with the ensemble spread in gray.
Figure 7 ( 3em) : The tropical North Atlantic (TNA) Warm Pool in September. (a-d) Mean depth
(meters) of the 28.5°C isotherm in September. The CCSM4 ensemble mean (panel a) is the mean of five different simulations. (e) Time series of the volume (10 4 km 3 ) encompassed by the
1110
28.5°C isotherm in September north of 5°N. The black line is the Ishii observational product; the blue line is the ocean POP simulation forced by CORE forcing; the red line is the CCSM4 ensemble mean with the ensemble spread in gray.
Figure 8 ( 4em) : Rank histograms of the CCSM4 ensemble spread against the POP ocean
1115 simulation forced by CORE (purple), and against the Ishii observational estimate (blue). The top
49

panel corresponds to the index of the tropical North Atlantic (TNA) Warm Pool in September.
The bottom panel corresponds to the index of the tropical South Atlantic (TSA) Warm Pool in
April. The black line represents a uniform distribution.
1120 Figure 9 ( WW-1 ): Dominant rotated EOFs (rEOFs) of SST for the ERSSTv3b data set (left), the mean of the five 20C ensemble members of the CCSM4 (center), and the CORE-forced oceanice simulation (right). The rEOFs are based on a varimax rotation of the 10 dominant EOFs of detrended, area-weighted, monthly SST anomalies. The North Tropical Atlantic (NTA) and
Subtropical South Atlantic (SSA) modes are found in all data sets. In CCSM4, the South
1125 Tropical Atlantic (STA) variability is represented by the STA-EQ and STA-BG modes, with SST variability in the equatorial region and the Benguela upwelling zone, respectively. The rEOFs carry the standard deviation. Negative, zero, and positive contours are thin dashed, thick solid, and thin solid, respectively, with contour interval of 0.1ºC.
1130 Figure 10 ( WW-2 ): Power spectra of the rotated PCs (rPCs) for the different modes featured in
Figure WW-1. A 13-point Daniell filter is applied to smooth the spectra. For CCSM4 (black) the spectra are averaged over the five 20C ensemble members. The spectra of the ERSST (dark gray) and CORE (light gray) data sets are offset by factors 0.25 and 0.0625, respectively. The thin lines are 95% confidence limits, based on a best-fit AR-1 model to the time series, and a 2500-
1135 member ensemble of AR-1 processes with these same parameters.
Figure 11 ( WW-3 ): Correlations between wind stress and the four dominant modes of SST in the
20C ensemble member 005 of CCSM4. Contours: peak correlation of monthly wind stress
50

magnitude anomalies and rPCs (interval 0.05; negative values in gray, positive in white; only
1140 values significantly different from zero at the 99% level are shown); shading: lag for which this peak correlation is achieved (in months; negative values: rPC lags wind stress magnitude); and vectors: the vectorized correlation between the rPCs and wind stress components at this lag
1145
(maximum vector lengths represent (square-root) correlations of 0.62, 0.75, 0.55 and 0.66 for rEOF 1, and the STA, NTA and SSA modes, respectively).
Figure 12 ( SG-1 ): Standard deviation (STD) of anomalous heat content rate of change in the upper 80m (shading, Wm
-2
), STD of anomalous SST (black contours), and time mean SST (gray
1150 contours). Box is the model Benguela region. All data are from the 1deg 1850 control run of
CCSM4.
Figure 13 ( SG-2 ): (a) Monthly (gray) and yearly smoothed (solid black) anomalous SST in the
Benguela region, model NINO3 anomalous SST (dashed black) shown 9 months ahead of
1155
Benguela SST. (b) Correlation of yearly smoothed Benguela SST with SST and wind stress elsewhere.
Figure 14 ( SG-3 ): Heat budget of the Benguela region.
(a) Lagged autocorrelation of anomalous SST and lagged correlation of anomalous heat content rate of change (HCR) with anomalous vertical (VERT), meridional (MER), zonal (ZON) heat advection, and anomalous net surface heat flux (NHF). All variables are spatially averaged over
1160 the Benguela region box and vertically integrated in the upper 80m.
(b) Lagged correlation of anomalous vertical heat advection in the Benguela region with wind
51

stress elsewhere. Arrows show maximum correlation. Shading and color scale in panels b) and c) show time lag (in month) corresponding to maximum correlation. Wind stress leads for positive lags. Correlations exceeding 0.3 are shown in red. Temporal regression of anomalous vertical
1165 heat advection on anomalous mean sea level pressure elsewhere at zero lag is overlain as contours. Contour values show pressure anomalies (mbar) corresponding to 100 Wm
-2 anomalous vertical heat advection in the Benguela region.
(c) The same as in (b) but for anomalous meridional heat advection. Pressure pattern is not
1170
1175
1180 shown.
52

1185 FIGURES
1190
1195
1200
Figure 1: Sea surface temperature (°C, left panels) and salinity (psu, right panels) model minus observations differences for the period of 1980-1999 for CCSM3 and 1986-2005 for CCSM4.
These plots correspond to Figure 6 from Danabasoglu et al. (2011, this issue) with a focus on the tropical Atlantic.
53

1205
1210
Figure 2: (top) Total mean of zonal and meridional wind stress from observations. The middle panel are the differences of these observations from CCSM4, and bottom panel are the differences with CCSM3. The units are N/m2 and the time period matches those for the means in Figure 1.
54

1215
1220
1225
1230
Figure 3: a) Mean SST along the equator from observations (black line), CCSM4 (red line), and
CCSM3 (blue line). Seasonal cycle of SST along the equator calculated as the mean of each month minus the total mean from observations (b), CCSM4 (c), and CCSM3 (d). Units are °C and the means are calculated over 1980-1999 for CCSM3 and 1986-2005 for CCSM4. The mean from observations spans both periods including 1980-2005.
55

1235
1240
Figure 4: The left panels are the seasonal mean of the wind stress (vectors) and their magnitude
(shades); the right panels are the differences of these observations from CCSM4 wind stress. The units are N/m2, and the time period used is 1986-2005.
56

1245
1250
Figure 5 (1em). (a-d) Horizontal distribution of the month of deepest 28.5°C isotherm from the long-term mean from 1950 to 2005. The numbers 1 to 12 correspond to the months from
January to December. The Pacific data has been masked. Panel (a) corresponds to the CCSM4 ensemble mean. Panel (c) corresponds to the POP ocean model forced with CORE surface forcing. Panels (b) and (d) correspond to the observational products, Ishii and Levitus, respectively. (e) Seasonal cycle of the volume of the 28.5°C isotherm between 40°S-40°N and above 250 meters of depth.
57

1255
1260
Figure 6 (2em). The tropical South Atlantic (TSA) Warm Pool in April. (a-d) Mean depth
(meters) of the 28.5°C isotherm in April. The CCSM4 ensemble mean (panel a) is the mean of five different simulations. (e) Time series of the volume (10
4
km
3 ) encompassed by the 28.5°C isotherm in April south of 5°N. The black line is the Ishii observational product; the blue line is the ocean POP simulation forced by CORE forcing; the red line is the CCSM4 ensemble mean with the ensemble spread in gray.
58

1265
1270
1275
Figure 7 (3em). The tropical North Atlantic (TNA) Warm Pool in September. (a-d) Mean depth
(meters) of the 28.5°C isotherm in September. The CCSM4 ensemble mean (panel a) is the mean of five different simulations. (e) Time series of the volume (10
4
km
3
) encompassed by the
28.5°C isotherm in September north of 5°N. The black line is the Ishii observational product; the blue line is the ocean POP simulation forced by CORE forcing; the red line is the CCSM4 ensemble mean with the ensemble spread in gray.
59

1280
1285
1290
1295
Figure 8 (4em). Rank histograms of the CCSM4 ensemble spread against the POP ocean simulation forced by CORE (purple), and against the Ishii observational estimate (blue). The top panel corresponds to the index of the tropical North Atlantic (TNA) Warm Pool in September.
The bottom panel corresponds to the index of the tropical South Atlantic (TSA) Warm Pool in
April. The black line represents a uniform distribution.
60

1300
1305
1310
Figure 9 ( WW-1 ): Dominant rotated EOFs (rEOFs) of SST for the ERSSTv3b data set (left), the mean of the five 20C ensemble members of the CCSM4 (center), and the CORE-forced oceanice simulation (right). The rEOFs are based on a varimax rotation of the 10 dominant EOFs of detrended, area-weighted, monthly SST anomalies. The North Tropical Atlantic (NTA) and
Subtropical South Atlantic (SSA) modes are found in all data sets. In CCSM4, the South
Tropical Atlantic (STA) variability is represented by the STA-EQ and STA-BG modes, with SST variability in the equatorial region and the Benguela upwelling zone, respectively. The rEOFs carry the standard deviation. Negative, zero, and positive contours are thin dashed, thick solid, and thin solid, respectively, with contour interval of 0.1ºC.
61

1315
1320
1325
Figure 10 ( WW-2 ): Power spectra of the rotated PCs (rPCs) for the different modes featured in
Figure WW-1. A 13-point Daniell filter is applied to smooth the spectra. For CCSM4 (black) the spectra are averaged over the five 20C ensemble members. The spectra of the ERSST (dark gray) and CORE (light gray) data sets are offset by factors 0.25 and 0.0625, respectively. The thin lines are 95% confidence limits, based on a best-fit AR-1 model to the time series, and a 2500member ensemble of AR-1 processes with these same parameters.
62

1330
1335
1340
1345
Figure 11 ( WW-3 ): Correlations between wind stress and the four dominant modes of SST in the
20C ensemble member 005 of CCSM4. Contours: peak correlation of monthly wind stress magnitude anomalies and rPCs (interval 0.05; negative values in gray, positive in white; only values significantly different from zero at the 99% level are shown); shading: lag for which this peak correlation is achieved (in months; negative values: rPC lags wind stress magnitude); and vectors: the vectorized correlation between the rPCs and wind stress components at this lag
(maximum vector lengths represent (square-root) correlations of 0.62, 0.75, 0.55 and 0.66 for rEOF 1, and the STA, NTA and SSA modes, respectively).
63

1350
1355
1360
Figure 12 ( SG-1 ): Standard deviation (STD) of anomalous heat content rate of change in the upper 80m (shading, Wm
-2
), STD of anomalous SST (black contours), and time mean SST (gray contours). Box is the model Benguela region. All data are from the 1deg 1850 control run of
CCSM4.
64

1365
1370
Figure 13 ( SG-2 ): (a) Monthly (gray) and yearly smoothed (solid black) anomalous SST in the
Benguela region, model NINO3 anomalous SST (dashed black) shown 9 months ahead of
Benguela SST. (b) Correlation of yearly smoothed Benguela SST with SST and wind stress elsewhere.
65

1375
1380
1385
Figure 14 ( SG-3 ): Heat budget of the Benguela region.
(a) Lagged autocorrelation of anomalous SST and lagged correlation of anomalous heat content rate of change (HCR) with anomalous vertical (VERT), meridional (MER), zonal (ZON) heat advection, and anomalous net surface heat flux (NHF). All variables are spatially averaged over the Benguela region box and vertically integrated in the upper 80m.
(b) Lagged correlation of anomalous vertical heat advection in the Benguela region with wind stress elsewhere. Arrows show maximum correlation. Shading and color scale in panels b) and c) show time lag (in month) corresponding to maximum correlation. Wind stress leads for positive lags. Correlations exceeding 0.3 are shown in red. Temporal regression of anomalous vertical heat advection on anomalous mean sea level pressure elsewhere at zero lag is overlain as contours. Contour values show pressure anomalies (mbar) corresponding to 100 Wm
-2 anomalous vertical heat advection in the Benguela region.
(c) The same as in (b) but for anomalous meridional heat advection. Pressure pattern is not shown.
66

1400
1405
1390
TABLES
1395 Table 1. Linear trend (10
4
km
3
/year) of the September TNA and April TSA warm pool indices for each CCSM4 ensemble simulation, the POP ocean simulation forced by CORE, and the observational estimate of Ishii. The columns R005 through R009 correspond to the ensemble simulations.
Linear trned
R005 R006 R007 R008 R009 POP Ishii
September 0.347 0.337 0.355 0.360 0.353 0.283 0.221
April 0.184 0.190 0.276 0.196 0.251 0.254 0.207
67

1420
1425
1410
1415
Table 2. Standard deviation of the September TNA and April TSA warm pool indices for each
CCSM4 ensemble simulation, the POP ocean simulation forced by CORE, and the observational estimate of Ishii. The columns R005 through R009 correspond to the ensemble simulations.
Standard deviation
R005
September 6.42
April 5.56
R006
5.00
4.90
R007
5.38
5.35
R008
5.89
4.61
R009
6.31
5.85
POP
8.90
8.14
Ishii
7.52
4.76
68

1440
1430
Table 3. Spearman (R a
S) and Pearson (R a
P) auto-correlations of the September TNA and April
TSA warm pool indices for each CCSM4 ensemble simulation, the POP ocean simulation forced
1435 by CORE, and the observational estimate of Ishii. The columns R005 through R009 correspond to the ensemble simulations.
Autocorrelation
R005 R006 R007 R008 R009 POP Ishii
September
R a
S
R a
P
April
R a
S
R a
P
0.38
0.46
0.18
0.22
0.30
0.35
0.32
0.39
0.31
0.34
0.12
0.10
0.43
0.39
0.18
0.03
0.56
0.49
0.20
0.24
0.17
0.14
0.02
-0.05
0.34
0.33
0.13
0.02
69

1460
1445
Table 4 (WW-1). Leading rotated EOFs (rEOFs) of SST for the ERSSTv3b data set, the five
20C ensemble members of CCSM4 (R005-R009), the ensemble mean (Ens), and the COREforced ocean-ice simulation. The rEOFs are based on a varimax rotation of the 10 dominant
EOFs of the detrended, area-weighted, monthly SST anomaly time series. The North Tropical
1450 Atlantic (NTA) and Subtropical South Atlantic (SSA) modes are found in all data sets. In the
CCSM4 ensemble members, the South Tropical Atlantic (STA) variability is represented by the
STA-EQ and STA-BG modes, with SST variability in the equatorial region and the Benguela upwelling zone, respectively. Lightest gray cells indicate relative ordering of the modes, while medium and dark gray cells indicate relative and absolute (domain-averaged, ºC2) levels of
1455 variance accounted for by the modes.
70

1475
1480
1465
1470
71

1485
72