Abstract

The hypothesis that extratropical low-frequency variability can be viewed as a linear barotropic response to spatially incoherent forcing is put to test. The recurrent ultra-low-frequency structures [i.e., the empirical orthogonal functions (E0Fs)] of a general circulation model are compared with the leading singular modes of the steady barotropic operator linearized about several mid- and upper-tropospheric levels. In a quasi-stochastically forced system these two sets of functions should approximately coincide. Although at each level one can find singular modes that closely replicate one of the EOFs, there does not exist a general correspondence at any one vertical level between the set of leading singular modes and the set of dominant EOFs, at least not for the same parameter range. The extratropical ultra-low-frequency variability cannot be understood as resulting solely from the organizing influence of horizontal asymmetries in the time-mean flow on randomly forced barotropic disturbances. The neglected processes, such as organized transient eddy forcing and diabatic heating, evidently play a crucial role in accounting for the prevalence of the observed patterns.

Abstract

This study examines the extent to which the thermodynamic interactions between the midlatitude atmosphere and the underlying oceanic mixed layer contribute to the low-frequency atmospheric variability. A general circulation model, run under perpetual northern winter conditions, is coupled to a motionless constant-depth mixed layer in midlatitudes, while elsewhere the sea surface temperature (SST) is kept fixed; interannual tropical SST forcing is not included. It is found that coupling does not modify the spatial organization of the variability. The influence of coupling is manifested as a slight reddening of the spectrum of 500-mb geopotential height and a significant enhancement of the lower-tropospheric thermal variance over the oceans at very low frequencies by virtue of the mixed-layer adjustment to surface air temperature variations that occurs on those timescales. This adjustment effectively reduces the thermal damping of the atmosphere associated with surface heat fluxes (or negative oceanic feedback), thus increasing the thermal variance and the persistence of circulation anomalies.

In studying the covariability between ocean and atmosphere it is found that the dominant mode of natural atmospheric variability is coupled to the leading mode of SST in each ocean, with the atmosphere leading the ocean by about one month. The cross-correlation function between oceanic and atmospheric anomalies is strongly asymmetric about zero lag. The SST structures are consistent with direct forcing by the anomalous heat fluxes implied by the concurrent surface air temperature and wind fluctuations. Additionally, composites based on large amplitude SST anomaly events contain no evidence of direct driving of atmospheric perturbations by these SST anomalies. Thus, in terms of the spatial organization of the covariability and the evolution of the coupled system from one regime to another, large-scale air–sea interaction in the model is characterized by one-way atmospheric forcing of the mixed layer.

These results are qualitatively consistent with those from an earlier idealized study. They imply a subtle but fundamental role for the midlatitude oceans as stabilizing rather than directly generating atmospheric anomalies. It is argued that this scenario is relevant to the dynamics of extratropical atmosphere–ocean coupling on intraseasonal timescales at least: the model is able to qualitatively reproduce the temporal and spatial characteristics of the observed dominant patterns of interaction on these timescales, particularly over the Atlantic.

Abstract

This study extends the investigation of the impact of midlatitude ocean–atmosphere interactions on the atmospheric circulation to the interannual timescale by incorporating SST variability in the tropical Pacific representative of observed conditions. Two perpetual January GCM simulations are performed to examine the changes in the low-frequency atmospheric variability brought about by the inclusion of an interactive slab mixed layer in midlatitudes, in particular the changes in the extratropical response to ENSO-like tropical 90-day mean SST anomalies.

It is found that midlatitude coupling alters the spatial organization of the low-frequency variability in qualitatively the same manner (but not to the same extent) as tropical SST variability—namely, by selectively enhancing (in terms of amplitude, persistence, and/or frequency of occurrence) certain of the preexisting (natural) dominant modes without significantly modifying them or generating new ones. While tropical SST forcing results in a notable amplification of the Pacific–North American (PNA) mode of the model, midlatitude SST anomalies appear to favor the regional zonal index circulations in the eastern and western Pacific (through decreased thermal damping at the surface). As a result, the PNA response to ENSO-like tropical SST forcing is not reinforced but slightly weakened by the presence of interactions with the underlying mixed layer. On the other hand, coupling increases the persistence of the overall extratropical signal and causes it to acquire distinct Western Pacific–like features, thus improving its resemblance to the observed ENSO teleconnection pattern.

The leading mode of covariability between the hemispheric atmospheric circulation and North Pacific SST qualitatively reproduces its observational counterpart, with the atmosphere leading by about one month and surface atmospheric variations consistent with the notion that the atmosphere is driving the ocean. This agreement suggests that, even on interannual timescales, two-way air–sea interactions and ocean dynamics do not play an essential role in establishing the large-scale spatial structure of this observed dominant mode of ocean–atmosphere interaction. In addition, the simulated patterns of covariability in this sector possess the same kind of interannual–intraseasonal duality exhibited by the observations. In the North Atlantic the model essentially recovers the results from Part I of this study.

Abstract

This study examines two possible mechanisms responsible for the selection of the preferred period of the Madden and Julian (40–50 day) oscillation. A global two-level nonlinear model with a positive-only CISK-type cumulus heating parameterization is used to simulate the oscillation, which appears when the SST exceeds a critical value for instability of CISK type. Longitudinal variations of tropical SST are imposed so that a stable and an unstable region coexist. When the cold SST sector is sufficiently stable, the CISK wave propagates efficiently through the stable region in the form of a damped moisture-modified Kelvin wave, and reemerges in the unstable region where its amplitude grows. When the SST in the stable sector is set closer to the instability threshold, the moist Kelvin wave slows down and decays before reentering the unstable region, but the CISK perturbation periodically regenerates over the warm waters in response to a local buildup of instability. This last experiment implies a new mechanism for setting the time scale of the oscillation, alternative to that of simple zonal propagation around the globe. A “discharge-recharge” theory is proposed whereby the 40-day recurrence period in the model is set by the growth and duration times of the convective episode together with the recharge time for the instability. It is shown that the midlatitude baroclinic eddies provide the quasi-stochastic forcing necessary to excite each new intraseasonal episode by organizing a region of subtropical convection, which then grows and expands equatorwards due to the effect of the latent heating. The dynamical picture that emerges from the above results is consistent with observations that suggest a local thermodynamically based time scale for the oscillation, and with case studies indicating that extratropical processes might be responsible for the onset of 40–50-day events.

Abstract

This study examines the dynamical nature of the linear and nonlinear extratropical response of the atmosphere to tropical intraseasonal heating with various temporal and spatial structures. A global two-level model is used. It is found that the midlatitude response is very sensitive to the phase speed of the forcing and, to a lesser extent, to its spatial scale. For a narrow monopolar heating source traveling eastward uniformly with an around-the-world period of 40 days, as in GCM simulations of the Madden–Julian oscillation (MJO), the response is weak and explains very little of the extratropical variance. For dipolar beating with zonal scale around 15 000 km (k = 2–3) and wave period of 40 days propagating slowly eastward within a confined longitudinal domain, a good approximation of the outgoing longwave radiation fields associated with the real MJO, the extratropical response is wavelike and well defined and stands significantly above the background variability. The nonlinear wavetrains evolve in time and exhibit many of the characteristics of their linear counterparts. Nonlinearities, however, play an important role in inhibiting the midlatitude response during certain phases of the oscillation, when cooling coexists west of heating and equatorial winds prevail. Time-mean planetary waves in midlatitudes are not needed to excite a pronounced response. The nonlinear model with zonally asymmetric climatology reveals moderate sensitivity to the longitude of the heating, with a tendency for more intermittent wave trains than in the zonally symmetric case, and does not reproduce the resonant behavior found in the linear model for a particular configuration of the heating and the topographic forcing. The potential asymmetry of the extratropical response to the MJO with respect to the phase of the dipole in intraseasonal convection should he kept in mind when attempting to detect the midaltitude signal from observational data.

Abstract

The extratropical response to El Niño in late fall departs considerably from the canonical El Niño signal. Observational analysis suggests that this response is modulated by anomalous forcing in the tropical west Pacific (TWP), so that a strong fall El Niño teleconnection is more likely when warm SST conditions and/or enhanced convection prevail in the TWP. While these TWP SST anomalies may arise from noise and/or long-term variability, they may also be generated by differences between El Niño events, through variations in the tropical “atmospheric bridge.” This bridge typically drives subsidence west of the date line and enhanced trade winds over the far TWP, which cool the ocean. In late fall, however, some relatively weaker and/or more eastward-shifted El Niño events produce a correspondingly weakened and displaced tropical bridge, which results in no surface cooling and enhanced convection in the TWP. Because the North Pacific circulation is very sensitive to forcing from the TWP at this time of year, the final outcome is a strong extratropical El Niño teleconnection.

This hypothesis is partly supported by regionally coupled ensemble GCM simulations for the 1950–99 period, in which prescribed observed El Niño SST anomalies in the eastern/central equatorial Pacific and an oceanic mixed layer model elsewhere coexist, so that the TWP is allowed to interact with the El Niño atmospheric bridge. To separate the deterministic signal driven by TWP coupling from that associated with inter–El Niño differences and from the “noise” due to intrinsic TWP convection variability (not induced by local SST anomalies), a second large-ensemble (100) simulation of the 1997/98 El Niño event, with coupling limited to the TWP and tropical Indian Ocean, is carried out. Together, the model findings suggest that the extratropical El Niño teleconnection during late fall is very sensitive to convective forcing in the TWP and that coupling-induced warming in the TWP may enhance this El Niño teleconnection by promoting convection in this critical TWP region. A more general implication is that diagnostic studies using December–February (DJF) seasonal averages may obscure some important aspects of climate anomalies associated with forcing in the tropical Pacific.

Annual global surface temperature and global land surface temperature trends are calculated for all possible periods of the historical record between 1850 and 2009. Two-dimensional parameter diagrams show the critical influence of the choice of start and end years on the calculated trend and associated temperature changes and suggest time scales required to establish robust trends.

The largest trends and associated temperature changes are all positive and have occurred over periods ending in recent years. Substantial positive changes also occurred from the early twentieth century until the mid-1940s. The continents exhibit greater long-term warming than the global average overall, but less warming in the early part of the century (segments ending in the 1940s). The recent period of short-term cooling beginning in the late 1990s is neither statistically significant nor unusual in the context of trend variability in the full historical record.

Global-mean and land surface temperature changes for periods ending in recent years and longer than about 90 years are extremely unlikely to have occurred by chance. In contrast, short-term trends over less than a few decades are generally not statistically significant. This implies significant contributions of decadal variability to trends estimated over such short time periods.

Abstract

The winter extratropical teleconnection of El Niño–Southern Oscillation (ENSO) in the North Atlantic–European (NAE) sector remains controversial, concerning both the amplitude of its impacts and the underlying dynamics. However, a well-established response is a late-winter (January–March) signal in sea level pressure (SLP) consisting of a dipolar pattern that resembles the North Atlantic Oscillation (NAO). Clarifying the relationship between this “NAO-like” ENSO signal and the actual NAO is the focus of this study. The ENSO–NAE teleconnection and NAO signature are diagnosed by means of linear regression onto the sea surface temperature (SST) Niño-3.4 index and an EOF-based NAO index, respectively, using long-term reanalysis data (NOAA-20CR, ERA-20CR). While the similarity in SLP is evident, the analysis of anomalous upper-tropospheric geopotential height, zonal wind, and transient-eddy momentum flux, as well as precipitation and meridional eddy heat flux, suggests that there is no dynamical link between the phenomena. The observational results are further confirmed by analyzing two 10-member ensembles of atmosphere-only simulations (using an intermediate-complexity and a state-of-the-art model) with prescribed SSTs over the twentieth century. The SST-forced variability in the Northern Hemisphere is dominated by the extratropical ENSO teleconnection, which provides modest but significant SLP skill in the NAE midlatitudes. The regional internally generated variability, estimated from residuals around the ensemble mean, corresponds to the NAO pattern. It is concluded that distinct dynamics are at play in the ENSO–NAE teleconnection and NAO variability, and caution is advised when interpreting the former in terms of the latter.

Abstract

The core region of the North American summer monsoon is examined using spatially averaged daily rainfall observations obtained from gauges, with the objective of improving understanding of its climatology and variability. At most grid points, composite and interannual variations of the onset and end of the wet season are well defined, although, among individual stations that make up a grid average, variability is large. The trigger for monsoon onset in southern and eastern Mexico appears to be related to a change in vertical velocity, while for northwestern Mexico, Arizona, and New Mexico it is related to a reduction in stability, as indicated by a decrease in the lifted index. The wet-season rain rate is a combination of the wet-day rain rate, which decreases with distance from the coast, and the wet-day frequency, which is largest over the Sierra Madre Occidental. Thus the maximum total rate lies slightly to the west of the highest orography. As has been previously noted, onset is not always well correlated with total seasonal precipitation, so in these areas, variations of wet-day frequency and wet-day rain rate must be important. Correlations are small between the wet-day frequency and the wet-day rate, and the former is better correlated than the latter with the seasonal rain rate. Summer rainfall in central to southern Mexico exhibits moderate negative correlations with the leading pattern of sea surface temperature (SST) anomalies in the equatorial Pacific, which projects strongly onto El Niño. The influence of equatorial SSTs on southern Mexico rainfall seems to operate mainly through variability of the wet-day frequency, rather than through variations of the wet-day rain rate.

Abstract

The authors systematically investigate two easily computed measures of the effective number of spatial degrees of freedom (ESDOF), or number of independently varying spatial patterns, of a time-varying field of data. The first measure is based on matching the mean and variance of the time series of the spatially integrated squared anomaly of the field to a chi-squared distribution. The second measure, which is equivalent to the first for a long time sample of normally distributed field values, is based on the partitioning of variance between the EOFs. Although these measures were proposed almost 30 years ago, this paper aims to provide a comprehensive discussion of them that may help promote their more widespread use.

The authors summarize the theoretical basis of the two measures and considerations when estimating them with a limited time sample or from nonnormally distributed data. It is shown that standard statistical significance tests for the difference or correlation between two realizations of a field (e.g., a forecast and an observation) are approximately valid if the number of degrees of freedom is chosen using an appropriate combination of the two ESDOF measures. Also described is a method involving ESDOF for deciding whether two time-varying fields are significantly correlated to each other.

A discussion of the parallels between ESDOF and the effective sample size of an autocorrelated time series is given, and the authors review how an appropriate measure of effective sample size can be computed for assessing the significance of correlations between two time series.