So 90 120 Iso Iso 210 240 270 30090 60 90 120 160 160 210 240 270 30ib0 So 90 120 Iso 180 210 240 270 300

Figure 1. Ocean model domains used in the (a) Indian Ocean Run, (b) Indo-Pacific Run, and (c) Pacific Run. Also shown are the long-term mean SST's produced by these runs. Contour intervals are 2°C. Values greater than 28° C are shaded.

3. The Simulated ENSO Cycle

The UCLA CGCM has been shown to produce a realistic seasonal cycle in the tropical eastern Pacific (Yu et al., 1999) and is considered one of the few coupled models capable of producing realistic ENSO simulations in an inter-comparison project of ENSO simulation (Latif et al., 2001). The ENSO cycle simulated by the Pacific Run was thoroughly examined by Yu and Mechoso (2001). It was found that ENSO-like SST variability is produced in the model approximately every 4 years, with a maximum amplitude of about 2°C. Large westerly wind stress anomalies are simulated to the west of the maximum SST anomalies in all major warm events. A multichannel singular spectrum analysis (M-SSA) (Keppenne and Ghil, 1992) was used to extract the simulated ENSO cycle from the simulation. The M-SSA method has been shown to be capable of extracting near-periodicity, and their associated spatiotemporal structures, from short and noisy time series (Robertson et al., 1995). Figure 2 shows the structure and evolution of the simulated ENSO cycle as revealed by the leading M-SSA mode along the equator. The ENSO cycle is found to be characterized by predominantly standing oscillations of SST in the eastern Pacific, almost simultaneous zonal wind stress anomalies

(a) zonal wind stress (b) SST (c) Ocean Heat Content

(a) zonal wind stress (b) SST (c) Ocean Heat Content

Figure 2. Structures of the simulated ENSO cycle along the equatorial Pacific extracted from the Pacific Run by the M-SSA method. Panel (a) shows the structure of zonal wind stress anomalies, (b) the structure of SST anomalies, and (c) the structure of ocean heat content anomalies. The coordinate is the 61-month (5-year) lag used in the M-SSA. Contour intervals are 0.1 dyn/cm2 for (a), 0.10°C for (b), and 0.05°C for (c). Values shown in (a) are scaled by 10. Positive values are shaded. (From Yu and Mechoso, 2001.)

Figure 2. Structures of the simulated ENSO cycle along the equatorial Pacific extracted from the Pacific Run by the M-SSA method. Panel (a) shows the structure of zonal wind stress anomalies, (b) the structure of SST anomalies, and (c) the structure of ocean heat content anomalies. The coordinate is the 61-month (5-year) lag used in the M-SSA. Contour intervals are 0.1 dyn/cm2 for (a), 0.10°C for (b), and 0.05°C for (c). Values shown in (a) are scaled by 10. Positive values are shaded. (From Yu and Mechoso, 2001.)

(a) zonal-mean component (b) zonally asymmetric component

LONGITUDE LONGITUDE

Figure 3. The (a) zonal-mean and (b) zonally asymmetric components of the ocean heat content structure shown in Fig. 2(c). The vertical coordinate spans the 61-month window used in the M-SSA. Contour intervals are 0.05°C. Positive values are shaded. (From Yu and Mechoso, 2001.)

LONGITUDE LONGITUDE

Figure 3. The (a) zonal-mean and (b) zonally asymmetric components of the ocean heat content structure shown in Fig. 2(c). The vertical coordinate spans the 61-month window used in the M-SSA. Contour intervals are 0.05°C. Positive values are shaded. (From Yu and Mechoso, 2001.)

to the west of the SST anomalies, and preceding ocean heat content anomalies near the eastern edge of the basin. These features are similar to those observed during ENSO events.

The ENSO dynamics in the UCLA CGCM were further examined by focusing on the relationship between SST and ocean heat content (i.e. the memory of ENSO). The ocean heat content is defined as the ocean temperature averaged in the upper 300 m. By separating the ocean heat content anomaly of the M-SSA mode into its zonal-mean and zonally asymmetric components (Fig. 3), it is found that the evolution of the zonal-mean component at the equator is 90° out of phase with that of the zonally asymmetric component, as well as that of the SST anomaly [compare Fig. 3(a) to Fig. 2(b)]. The onset of the warm ENSO phase occurs at the time when the mean ocean heat content anomaly grows (i.e. recharge) to a maximum value. The ocean heat content anomalies are then removed (i.e. discharge) as the warm phase develops toward its mature stage. This phase lag indicates that the variation in the zonal-mean ocean heat content provides the oscillation memory for the ENSO cycle. The ENSO dynamics in the UCLA CGCM are consistent with the recharge oscillator theory (Wyrtki, 1975; Cane and Zebiak, 1987; Zebiak, 1989; Jin, 1997; Li, 1997). This theory is conceptually similar to the delayed oscillator theory in suggesting the importance of subsurface ocean adjustment processes in producing the needed delay for ENSO oscillation. The recharge oscillator, however, emphasizes the importance of the buildup (i.e. charge) and release (i.e. discharge) of zonal-mean ocean heat content in the equatorial band for the phase reversal of the ENSO cycle.

4. Teleconnection of ENSO

By applying the M-SSA to the SST variability produced by the Indo-Pacific Run, it is found

Figure 4. Structures of SST and surface wind stress anomalies of the leading M-SSA mode obtained from the Indo-Pacific Run. Contours represent SST anomalies. Surface wind stress anomalies are indicated by vectors. Contour intervals are 0.1°C. The structures shown are taken from the mature phase of ENSO in the leading M-SSA mode.

Figure 4. Structures of SST and surface wind stress anomalies of the leading M-SSA mode obtained from the Indo-Pacific Run. Contours represent SST anomalies. Surface wind stress anomalies are indicated by vectors. Contour intervals are 0.1°C. The structures shown are taken from the mature phase of ENSO in the leading M-SSA mode.

that the simulated ENSO cycle is accompanied by significant SST anomalies in many parts of the Indo-Pacific Ocean. Figure 4 shows the SST anomaly pattern of the leading M-SSA mode (i.e. the ENSO mode) extracted from this run. The major anomaly features in this figure include: (1) the Pacific ENSO, (2) the basinwide warming/cooling in the Indian Ocean,

(3) an anomaly dipole in the northwestern Pacific (10°N-30°N and 120°E-160°E), and

(4) an anomaly dipole in the northeastern Pacific (10°N-30°N and 170°E-120°W). The basinwide warming (cooling) during an El Nino (La Nina) event is a well-known remote response of the Indian Ocean to ENSO (e.g. Yu an Rienecker, 1999, 2000; Murtugudde et al., 2000) via the "atmospheric bridge" mechanism (Lau and Nath, 1996) or the ENSO-induced tro-pospheric temperature and moisture variations (Chiang and Sobel, 2002; Neelin et al., 2003). The SST dipole in the northwestern Pacific is accompanied by an anomalous anticyclonic (cyclonic) surface circulation during El Nino (La Nina) and is very similar to the Pacific-East Asia teleconnection pattern discussed by Wang et al. (2000) and Lau and Nath (2006). They suggested that this teleconnection pattern was initially forced by ENSO, and was later maintained by a positive thermodynamic feedback between the circulation anomaly and the ocean mixed layer in the northwestern Pacific.

The northeastern SST anomaly dipole simulated by the CGCM is similar to the zonal SST dipole observed during the 1997-98 ENSO event (Liu et al., 1998), which consisted of centers of anomalous warming along the coast of California and anomalous cooling further to the west in the central Pacific. The question arises as to whether or not the development of this subtropical SST anomaly feature is associated with that of ENSO. Furthermore, if such an association exists, what are the mechanisms that link the two phenomena? Yu et al. (2000) analyzed the Pacific Run to address these two questions. To concentrate on the relationship between ENSO and this SST dipole, an empirical orthogonal function (EOF) analysis was applied to the model SST anomalies in the Pacific domain between 30°S and 50° N. The leading EOF mode (not shown) is characterized by the ENSO in the tropics and the zonal SST dipole in the subtropics, similar to those shown in the Pacific portion of Fig. 4.

To examine their relationship, the lagged correlation coefficients were calculated between an index representing the strength of the northeastern Pacific SST dipole and an index representing the intensity of ENSO. The ENSO index is the averaged SST anomalies over the central equatorial Pacific (160°-130°W and 40°S-40°N), where the simulated SST anomalies are largest in Fig. 4. The dipole index is defined as the difference between the SST anomalies averaged over the eastern center (150°W-130°W and 24°N-36°N) of the dipole, and those averaged over the western center (180° E-160°W and 24°N-36°N). It was found that the maximum correlation coefficient between these two indices is 0.65 when ENSO leads the dipole by one month. These analyses suggested that the subtropical SST anomaly dipole is forced by ENSO.

To understand the generation mechanism of the SST dipole, Fig. 5 illustrates the relationships between the principal component of the leading EOF mode of SST and the anomalies in sea-level pressure, surface wind stress, and surface heat flux. For the sake of discussion, Fig. 5 is interpreted for the case of a warm ENSO event. Figure 5(a) shows that during El Niño events, the Aleutian Low is enhanced and results in an anomalous cyclonic circulation over the subtropical Pacific. Along the southwesterly branch of the cyclonic circulation, the surface heat flux is reduced [Fig. 5(b)]. This branch brings warm and moist air from the tropics to the subtropics, which is consistent with a reduction in sensible and latent heat flux off the North America coast. Similarly, the northwesterly branch of the cyclonic circulation brings dry and cold air from higher latitudes, which results in increased surface heat flux from the central Pacific. The regions of reduced and enhanced surface heat flux roughly coincide with the warm and cold centers of the subtropical SST anomaly dipole [Fig. 5(c)]. Figure 5, therefore, suggests that the SST anomaly dipole is driven primarily by anomalous surface heat

0 0

Post a comment