Abstract

Anomalous atmospheric fields associated with the North Atlantic Oscillation (NAO) are analyzed on interannual and intraseasonal time scales in order to examine the extent to which the NAO is a regional phenomenon.

Analyses on the interannual time scale reveal that the NAO signal is relatively confined to the Euro–Atlantic sector in December while it extends toward East Asia and the North Pacific in February. The difference is most clearly seen in the meridional wind anomaly, which shows a wave train along the Asian jet, collocated with an anomalous vorticity source near the jet entrance. Diagnoses using a linear barotropic model indicate that this wave train is interpreted as quasi-stationary Rossby waves trapped on the Asian jet waveguide, and effectively excited by the anomalous upper-level convergence over the Mediterranean Sea. It is found that, when the NAO accompanies the Mediterranean convergence (MC) anomaly, most frequently seen in February, the NAO indeed has a much wider horizontal structure than the classical picture, rather similar to the Arctic Oscillation. In such cases interannual variability of the NAO is tied to the East Asian climate variability such that the positive NAO tends to bring a surface warming over East Asia.

Similar results are obtained from an analysis of individual NAO events based on low-pass-filtered daily fields, which additionally identified that the downstream extension occurs at the decay stage of the NAO event and the MC anomaly appears to be induced by the Ekman pumping associated with the NAO. The signal of the MC anomaly can be detected even at 5 days before the peak of the NAO, suggesting that the NAO influence to East Asia is predictable to some extent; therefore, monitoring the developing NAO event is useful to the medium-range weather forecast in East Asian countries.

Abstract

Atmosphere–ocean coupled processes responsible for generating and maintaining the equatorial warm pool were investigated using models of different complexities. The primary focus was to answer the following question: why is the observed warm pool concentrated around the maritime continent? In this first of a two-part series, the solutions of a simple conceptual model that represents the tropical Pacific and Indian Oceans interacting via the Walker circulation are examined. When the interbasin coupling is sufficiently strong, surface wind convergence over the Maritime Continent associated with easterly trades over the Pacific acts to generate the equatorial westerly over the Indian Ocean, leading to a warm pool spontaneously emerging between the two ocean basins. The conceptual model shows that tropical climate has two equilibria, depending upon the ocean basin widths—a single warm pool regime corresponding to the current climate and a split warm pool regime that accompanies warm pools created in the western parts of each ocean basin. The latter is found to be unstable and hence exhibits large-amplitude vacillations between the ocean basins being further amplified by the Bjerknes feedback. The above two regimes of the equatorial warm pool are identified in the model incorporating the interactive Atlantic Ocean as well, wherein the mean state and variability in the three ocean basins qualitatively agree with the observations.

Abstract

In this second of a two-part study, the two regimes in a simple tropical climate model identified in Part I are verified using a hybrid coupled general circulation model (HCM) that can reproduce the observed climatology and the interannual variability reasonably well. Defining a ratio of basin width between the Pacific and Indian Oceans, a series of parameter sweep experiments was conducted with idealized tropical land geometry. Consistent with the simple model, the HCM simulates two distinct states: the split warm pool regime with large vacillation between the two ocean basins and the single warm pool regime representing current climate. The former is suddenly switched to the latter as the Pacific becomes wider than the Indian Ocean. Furthermore, the vacillation in the split regime reveals a preferred transition route that the warm phase in the Pacific follows that in the Indian Ocean. This route occurs due to convectively coupled Kelvin waves that accompany precipitation anomalies over land. Additional experiments show that the inclusion of the idealized Eurasian continent stabilizes the split regime by reducing the Bjerknes feedback in the Indian Ocean, suggesting the atmosphere–ocean–land interaction at work in maintaining the observed warm pool. No difference in cloud feedback was found between two regimes; this feature may, however, be model dependent.

Both the simple model and the HCM results suggest that the tropical atmosphere–ocean system inherently involves multiple solutions, which may have an implication on climate modeling as well as on the understanding of the observed mean climate.

Abstract

A method is introduced for reducing forecast errors in an extended-range to one-month forecast based on an ensemble Kalman filter (EnKF). The prediction skill in such a forecast is typically affected not only by the accuracy of initial conditions but also by the model imperfections. Hence, to improve the forecast in imperfect models, the framework of EnKF is modified by using a state augmentation method. The method includes an adaptive parameter estimation that optimizes mismatched model parameters and a model ensemble initialized with the perturbed model parameter. The main features are the combined ensemble forecast of the initial condition and the parameter, and the assimilation for time-varying parameters with a theoretical basis.

First, the method is validated in the imperfect Lorenz ’96 model constructed by parameterizing the small-scale variable of the perfect model. The results indicate a reduction in the ensemble-mean forecast error and the optimization of the ensemble spread. It is found that the time-dependent parameter estimation contributes to reduce the forecast error with a lead time shorter than one week, whereas the model ensemble is effective for improving a forecast with a longer lead time. Moreover, the parameter assimilation is useful when model imperfections have a longer time scale than the forecast lead time, and the model ensemble appears to be relevant in any time scale. Preliminary results using a low-resolution atmospheric general circulation model that implements this method support some of the above findings.

Abstract

In association with extreme anomalies in the extratropical atmosphere, numerical experiments using an atmospheric general circulation model are performed to investigate the relative impact of the anomalous snow with SST anomalies on the atmospheric circulation. Large negative anomalies in the Eurasian snow cover and global SST anomalies observed in 1988/89 are employed as the respective boundary forcings because winter atmospheric states largely shifted in 1989.

The model is integrated for half a year from 1 September. Five-member ensemble states are obtained by conducting the light snow (LSNW) run, in which the snowfall was suppressed over eastern Eurasia during the first 3 months with prescribed SSTs, and another experiment, which employed observed SST anomalies instead of snow anomalies (the SST run). The LSNW run simulated dipole (positive in midlatitudes and negative in polar regions) anomalies in 500-hPa height similar to those observed in 1989, although the amplitude was smaller over the North Pacific. Surface warming over Eurasia found in winter 1989 is also reproduced through albedo feedback. On the other hand, the SST run reveals large height anomalies over the North Pacific in addition to the significant dipole similar to that in the LSNW run, but failed to reproduce observed surface warming as well as negative snow anomalies over Eurasia. SST anomalies in the equatorial Pacific corresponding to La Niña in 1988/89 are responsible for simulated height anomalies over the North Pacific in the SST run, whereas an influence of extratropical SST anomalies appears to be tenuous relative to either tropical SST anomalies or Eurasian snow anomalies. The amplitude of response in the LSNW run is roughly 60% of that in the SST run.

An analysis of the dynamics emphasizes that, in the upper troposphere, interactions of anomalies themselves with climatological zonal asymmetries as well as changes in transient eddy activities contribute to the height response found in the model. This suggests that the nonlinearities in the atmosphere are also important in addition to the snow and SST anomalies for the extreme anomalies in winter 1989 atmospheric circulation.

Abstract

Historical wintertime sea surface temperature (SST) data show that a sandwich pattern dominates on the decadal timescales in the North Atlantic, at least after the 1970s. The authors investigated how such decadal SST anomalies can survive against local thermal feedback, which acts to dampen them rapidly. At the surface, winter SST anomalies have a negligible projection with the subsequent summer anomalies while they show a significant projection with the SST anomalies in the next winter. On the other hand, observed summer temperature anomalies below the mixed layer tend to have the same sign as the previous winter SST anomalies, although the magnitude of the former is roughly one-third of the latter. This suggests that a reemergence mechanism of SST anomalies associated with the seasonal cycle of the mixed layer depth (MLD), which has been verified by Alexander and Deser, helps maintain the decadal SST anomalies. In order to examine this scenario, a mixed layer model driven by daily atmospheric data generated by a T42 atmospheric general circulation model was used. The mixed layer model well reproduces the climatology of both SST and MLD in the North Atlantic. An experiment in which a thermal forcing having the observed decadal pattern is added only for the initial winter shows that the SST anomalies disappear until July but reappear in the subsequent winters. This result supports the inference based on the observational evidence, and is explained as follows: (i) SST anomalies are partly detrained to deeper levels in spring when the mixed layer shoals rapidly, (ii) temperature anomalies beneath the shallow mixed layer are preserved during summer, (iii) they are entrained into the surface in the succeeding fall and winter when the mixed layer is again deepened. The recurrence of SST anomalies was found in two centers of the decadal anomaly pattern (30°–45°N, 80°–50°W and 45°–60°N, 50°–20°W), but not in another center in the subtropics (10°–25°N, 40°–10°W) where the MLD reveals only a small seasonality. The magnitude of recurrent SST anomalies is affected by two factors: MLD difference between winter and summer and the persistence of SST anomalies from winter to spring as they determine the amount and the magnitude of detrained temperature anomalies into the mixed layer, respectively.

The above results indicate that the effective damping time for the winter SST anomalies is much longer than the local damping time of several months.

Abstract

Warm and cold phases of El Niño–Southern Oscillation (ENSO) exhibit a significant asymmetry in their transition/duration such that El Niño tends to shift rapidly to La Niña after the mature phase, whereas La Niña tends to persist for up to 2 yr. The possible role of sea surface temperature (SST) anomalies in the Indian Ocean (IO) in this ENSO asymmetry is investigated using a coupled general circulation model (CGCM). Decoupled-IO experiments are conducted to assess asymmetric IO feedbacks to the ongoing ENSO evolution in the Pacific. Identical-twin forecast experiments show that a coupling of the IO extends the skillful prediction of the ENSO warm phase by about one year, which was about 8 months in the absence of the IO coupling, in which a significant drop of the prediction skill around the boreal spring (known as the spring prediction barrier) is found. The effect of IO coupling on the predictability of the Pacific SST is significantly weaker in the decay phase of La Niña. Warm IO SST anomalies associated with El Niño enhance surface easterlies over the equatorial western Pacific and hence facilitate the El Niño decay. However, this mechanism cannot be applied to cold IO SST anomalies during La Niña. The result of these CGCM experiments estimates that approximately one-half of the ENSO asymmetry arises from the phase-dependent nature of the Indo-Pacific interbasin coupling.

Abstract

To better understand the predictability of the wavelike circumglobal teleconnection (CGT) pattern prevailing during boreal summer, two sets of experiments are performed using a nonlinear dry atmospheric model. Each experiment consists of a 10-member ensemble of 26-yr integrations driven by the diabatic heating derived from reanalysis data: one with the monthly climatological mean heating (CLIM) and the other with the monthly heating for 1979–2004 (HIST). Both do well in reproducing the observed summer mean state, as well as the low-frequency variance distribution. The CGT pattern identified in the monthly meridional wind anomalies at 200 hPa shows zonally oriented wave packets over Eurasia. The simulated CGT has a nearly identical phase structure with the observations and indicates little difference between the CLIM and HIST results. While this indicates that the origin of CGT lies in the internal dry dynamics, the ensemble mean of the CGT in HIST is partly controlled by the slow variation in the heating field, as indicated by the high potential predictability of the simulated CGT pattern. Diagnoses using the linearized model demonstrate that the heating anomaly most responsible for the CGT-like steady response is located over the eastern Mediterranean region, where the heating may be coupled with the CGT pattern. In addition to the heating near the CGT, remote heating and cooling anomalies over North America and equatorial Africa are found to be effective at exciting stationary Rossby waves trapped on the Atlantic and Asian jets. It is thus suggested that the mechanisms generating the heating anomalies over these regions are the key to the predictability of the CGT pattern.

Abstract

The atmospheric boundary layer (ABL) response to mesoscale eddies in sea surface temperature (SST) in the Kuroshio Extension was investigated using a high-resolution (T213L30) atmospheric general circulation model. A control run was performed first by integrating the model for 40 days, driven by the satellite-derived, eddy-resolving SST during January 2006. The spatial pattern of surface wind anomalies—that is, a deviation from large-scale winds—reveals a positive correlation with the spatial pattern of mesoscale SST anomalies. The momentum budget analysis of the anomalous zonal wind was performed to investigate the formation of the ABL response. The most dominant term was the pressure gradient force; the advection term was comparable but in the opposite sense. Vertical mixing acts to weaken the anomalous zonal wind near the surface; however, the downward (upward) vertical turbulent flux anomalies were dominant near the ABL top over the warm (cold) SST anomalies, suggesting that the vertical mixing mechanism is effective. The role of the vertical mixing was further examined by a sensitivity experiment in which the turbulent diffusion coefficient for momentum was spatially smoothed. While the pressure gradient force and the advection terms were almost unchanged in the momentum budgets, the deceleration due to turbulence was enhanced because of the absence of the momentum input from the free atmosphere. The result is a reduction in the amplitude of the surface zonal wind anomalies to approximately half in the sensitivity experiment.

Abstract

Paleo proxy records have suggested that El Niño–Southern Oscillation (ENSO) variability during the mid-Holocene [8200 to 4200 years ago (8.2–4.2 ka)] was weaker than during the instrumental periods, but the mechanisms remain unclear. We examined processes of ENSO suppression using a coupled general circulation model (CGCM) that simulates ENSO amplitude and skewness under the present climate reasonably well. Two long simulations were performed: one using the preindustrial condition (CTRL) and the other using the 8-ka insolation having a greater seasonal cycle (MH8K). Consistent with proxy records and previous modeling studies, the ENSO amplitude weakened by 20% in MH8K compared to CTRL, mainly because of reduced thermocline feedback during the mature and decay phases. The weak thermocline feedback, likely a result of the loose equatorial thermocline in the mid-Holocene, suppresses the occurrence of extreme El Niño events and consequently explains the reduction in both ENSO amplitude and asymmetry. In MH8K, strengthened trade winds over the western-central Pacific Ocean act to cool the surface via evaporation while warmer water in the southern subtropical Pacific is transported beneath the equatorial thermocline, both contributing to diffuse the thermocline. Multimodel simulations for the mid-Holocene showed mean state changes and ENSO weakening similar to MH8K, but most models did not show reduced ENSO skewness, probably because of the failure in reproducing extreme El Niño events under the present climate.