Abstract

The annual variation of global atmospheric angular momentum (AAM) is dominated by its first and second harmonic components. The first harmonic is associated with maximum global AAM in winter (December– January–February) and minimum in summer, but the second harmonic is important enough to produce a distinct secondary midwinter minimum. Locally, the second harmonic has largest amplitude in the Tropics and subtropics of the upper troposphere. At present, little is known concerning the fundamental cause of this semiannual variation. The problem is investigated here by focusing on the upper-tropospheric winds, whose angular momentum is an excellent proxy of global AAM. The annual variation of the rotational part of these winds (the part that contributes to the global AAM) is diagnosed in a nonlinear upper-tropospheric vorticity-equation model with specified horizontal wind divergence and transient-eddy forcing. The divergence forcing is the more important of the two, especially in the Tropics and subtropics, where it is associated with tropical heating and cooling. Given the harmonics of the forcing, the model predicts the harmonics of the response, that is, the vorticity, from which the harmonics of angular momentum can then be calculated. The surprising but clear conclusion from this diagnosis is that the second harmonic of AAM arises more as a nonlinear response to the first harmonic of the divergence forcing than as a linear response to the second harmonic of the divergence forcing. This result has implications for general circulation model simulations of semiannual variations, not only of global AAM but also of other quantities.

Abstract

The sensitivity of the global atmospheric response to sea surface temperature (SST) anomalies throughout the tropical Indian and Pacific Ocean basins is investigated using the NCEP MRF9 general circulation model (GCM). Model responses in January are first determined for a uniform array of 42 localized SST anomaly patches over the domain. Results from the individual forcing experiments are then linearly combined using a statistically based smoothing procedure to produce sensitivity maps for many target quantities of interest, including the geopotential height response over the Pacific–North American (PNA) region and regional precipitation responses over North America, South America, Africa, Australia, and Indonesia.

Perhaps the most striking result from this analysis is that many important targets for seasonal forecasting, including the PNA response, are most sensitive to SST anomalies in the Niño-4 region (5°N–5°S, 150°W–160°E) of the central tropical Pacific, with lesser and sometimes opposite sensitivities to SST anomalies in the Niño-3 region (5°N–5°S, 90°–150°W) of the eastern tropical Pacific. However, certain important targets, such as Indonesian rainfall, are most sensitive to SST anomalies outside both the Niño-4 and -3 regions.

These results are also relevant in assessing atmospheric sensitivity to changes in tropical SSTs on decadal to centennial scales associated with natural as well as anthropogenic forcing. In this context it is interesting to note the surprising result that warm SST anomalies in one-third of the Indo-Pacific domain lead to a decrease of global mean precipitation.

Abstract

This paper is concerned with estimating the predictable variation of extratropical daily weather statistics (“storm tracks”) associated with global sea surface temperature (SST) changes on interannual to interdecadal scales, and its magnitude relative to the unpredictable noise. The SST-forced storm track signal in each northern winter in 1950–99 is estimated as the mean storm track anomaly in an ensemble of atmospheric general circulation model (AGCM) integrations for that winter with prescribed observed SSTs. Two sets of ensembles available from two modeling centers, with anomalous SSTs prescribed either globally or only in the Tropics, are used. Since the storm track signals cannot be derived directly from the archived monthly AGCM output, they are diagnosed from the SST-forced winter-mean 200-mb height signals using an empirical linear storm track model (STM). For two particular winters, the El Niño of January–February–March (JFM) 1987 and the La Niña of JFM 1989, the storm track signals and noise are estimated directly, and more accurately, from additional large ensembles of AGCM integrations. The linear STM is remarkably successful at capturing the AGCM's storm track signal in these two winters, and is thus also suitable for estimating the signal in other winters.

The principal conclusions from this analysis are as follows. A predictable SST-forced storm track signal exists in many winters, but its strength and pattern can change substantially from winter to winter. The correlation of the SST-forced and observed storm track anomalies is high enough in the Pacific–North America (PNA) sector to be of practical use. Most of the SST-forced signal is associated with tropical Pacific SST forcing; the central Pacific (Niño-4) is somewhat more important than the eastern Pacific (Niño-3) in this regard. Variations of the pattern correlation of the SST-forced and observed storm track anomaly fields from winter to winter, and among five-winter averages, are generally consistent with variations of the signal strength, and to that extent are identifiable a priori. Larger pattern correlations for the five-winter averages found in the second half of the 50-yr record are consistent with the stronger El Niño SST forcing in the second half. None of these conclusions, however, apply in the Euro-Atlantic sector, where the correlations of the SST-forced and observed storm track anomalies are found to be much smaller. Given also that they are inconsistent with the estimated signal-to-noise ratios in this region, substantial AGCM error in representing the regional response to tropical SST forcing, rather than intrinsically lower Euro-Atlantic storm track predictability, is argued to be behind these lower correlations.

Abstract

The time- and space scales of tropical deep convection are estimated via analysis of 3-hourly Global Cloud Imagery (GCI) data for 3 yr at 35–70-km resolution. The emphasis is on estimating local time- and space scales rather than traditional zonal wavenumber–frequency spectra. This is accomplished through estimation of local spatial lag autocorrelations, the conditional probability of convection at neighboring points, and the expected duration of convective events. The spatial autocorrelation scale is found to be approximately 130 km, and the mean duration of convective events approximately 5.5 h, in the convectively active areas of the Tropics. There is a tendency for the spatial autocorrelation scales to be shorter over the continents than oceans (95–155 versus 110–170 km). The expected duration of convective events likewise tends to be shorter (4–6 versus 5–7 h). In the far western Pacific, these differences are sharp enough to legitimize the notion of the Indonesian archipelago as an extended maritime continent with a distinctive shape. Consistent with many other studies, the diurnal variation of the convection is also found to be strikingly different over the continents and oceans. The diurnal amplitude over land is comparable to the long-term mean, raising the possibility of significant aliasing across timescales. The simple analysis of this paper should be useful in evaluating and perhaps even improving the representation of convective processes in general circulation models.

Abstract

Single column models (SCMs) provide an economical framework for developing and diagnosing representations of diabatic processes in weather and climate models. Their economy is achieved at the price of ignoring interactions with the circulation dynamics and with neighboring columns. It has recently been emphasized that this decoupling can lead to spurious error growth in SCM integrations that can totally obscure the error growth due to errors in the column physics that one hopes to isolate through such integrations. This paper suggests one way around this “existential crisis” of single column modeling. The basic idea is to focus on short-term SCM forecast errors, at ranges of 6 h or less, before a grossly unrealistic model state develops and before complex diabatic interactions render a clear diagnosis impossible.

To illustrate, a short-term forecast error diagnosis of the NCAR SCM is presented for tropical conditions observed during the Tropical Ocean and Global Atmosphere (TOGA) Coupled Ocean–Atmosphere Response Experiment (COARE). The 21-day observing period is divided into 84 6-h segments for this purpose. The SCM error evolution is shown to be nearly linear over these 6-h segments and, indeed, apart from a vertical mean bias, to be mainly an extrapolation of initial tendencies. The latter are then decomposed into contributions by various components of the column physics, and additional 6-h integrations are performed with each component separately and in combination with others to assess its contribution to the 6-h errors. Initial tendency and 6-h error diagnostics thus complement each other in diagnosing column physics errors by this approach.

Although the SCM evolution from one time step to the next is nearly linear, the finite-amplitude adjustments made multiple times within each time step to the temperature and humidity to remove supersaturation and convective instabilities make it necessary to consider nonlinear interactions between the column physics components. One such particularly strong interaction is identified between vertical diffusion and deep convection. The former, though nominally small, is shown to have a profound impact on both the amplitude and timing of the latter, and thence on the small imbalance between the total diabatic heating and adiabatic cooling of ascent in the column. The SCM diagnosis thus suggests that misrepresentation of this interaction, in addition to that of the interacting components themselves, might be a major contributor to the NCAR GCM's tropical simulation errors.

Abstract

While it is obvious that the mean diabatic forcing of the atmosphere is crucial for maintaining the mean climate, the importance of diabatic forcing fluctuations is less evident in this regard. Such fluctuations do not appear directly in the equations of the mean climate but affect the mean indirectly through their effects on the time-mean transient-eddy fluxes of heat, momentum, and moisture. How large are these effects? What are the effects of tropical phenomena associated with substantial heating variations such as ENSO and the MJO? To what extent do variations of the extratropical surface heat fluxes and precipitation affect the mean climate? What are the effects of the rapid “stochastic” components of the heating fluctuations? Most current climate models misrepresent ENSO and the MJO and ignore stochastic forcing; they therefore also misrepresent their mean effects. To what extent does this contribute to climate model biases and to projections of climate change?

This paper provides an assessment of such impacts by comparing with observations a long simulation of the northern winter climate by a dry adiabatic general circulation model forced only with the observed time-mean diabatic forcing as a constant forcing. Remarkably, despite the total neglect of all forcing variations, the model reproduces most features of the observed circulation variability and the mean climate, with biases similar to those of some state-of-the-art general circulation models. In particular, the spatial structures of the circulation variability are remarkably well reproduced. Their amplitudes, however, are progressively underestimated from the synoptic to the subseasonal to interannual and longer time scales. This underestimation is attributed to the neglect of the variable forcing. The model also excites significant tropical variability from the extratropics on interannual scales, which is overwhelmed in reality by the response to tropical heating variability. It is argued that the results of this study suggest a role for the stochastic, and not only the coherent, components of transient diabatic forcing in the dynamics of climate variability and the mean climate.

Abstract

An important question in assessing twentieth-century climate change is to what extent have ENSO-related variations contributed to the observed trends. Isolating such contributions is challenging for several reasons, including ambiguities arising from how ENSO itself is defined. In particular, defining ENSO in terms of a single index and ENSO-related variations in terms of regressions on that index, as done in many previous studies, can lead to wrong conclusions. This paper argues that ENSO is best viewed not as a number but as an evolving dynamical process for this purpose. Specifically, ENSO is identified with the four dynamical eigenvectors of tropical SST evolution that are most important in the observed evolution of ENSO events. This definition is used to isolate the ENSO-related component of global SST variations on a month-by-month basis in the 136-yr (1871–2006) Hadley Centre Sea Ice and Sea Surface Temperature dataset (HadISST). The analysis shows that previously identified multidecadal variations in the Pacific, Indian, and Atlantic Oceans all have substantial ENSO components. The long-term warming trends over these oceans are also found to have appreciable ENSO components, in some instances up to 40% of the total trend. The ENSO-unrelated component of 5-yr average SST variations, obtained by removing the ENSO-related component, is interpreted as a combination of anthropogenic, naturally forced, and internally generated coherent multidecadal variations. The following two surprising aspects of these ENSO-unrelated variations are emphasized: 1) a strong cooling trend in the eastern equatorial Pacific Ocean and 2) a nearly zonally symmetric multidecadal tropical–extratropical seesaw that has amplified in recent decades. The latter has played a major role in modulating SSTs over the Indian Ocean.

Abstract

The relative impacts of tropical diabatic heating and stratospheric circulation anomalies on wintertime extratropical tropospheric variability are investigated in a linear inverse model (LIM) derived from the observed zero lag and 5-day lag covariances of 7-day running mean departures from the annual cycle. The model predicts the covariances at all other lags. The predicted and observed lag covariances are generally found to be in excellent agreement, even at the much longer lag of 21 days. This validates the LIM’s basic premise that the dynamics of weekly averages are effectively linear and stochastically driven, which justifies further linear diagnosis of the system.

Analysis of interactions among the LIM’s variables shows that tropical diabatic heating greatly enhances persistent variability over most of the Northern Hemisphere, especially over the Pacific Ocean and North America. Stratospheric effects are largely confined to the polar region, where they ensure that the dominant pattern of sea level pressure variability is the annular Arctic Oscillation rather than the more localized North Atlantic Oscillation. Over the North Atlantic, both effects are important, although some of the stratospheric influence is ultimately traceable to tropical forcing. In general, the tropically forced anomalies extend through the depth of the troposphere and into the stratosphere, whereas stratospherically generated anomalies tend to be largest at the surface and relatively weak at midtropospheric levels. Some persistent variability is, however, found even in the absence of these “external” forcings, especially near the amplitude maxima of the leading eigenmodes of the internal extratropical tropospheric evolution operator. One of these eigenmodes has a circumglobal zonal wavenumber-5 structure with maxima over the Arabian Sea and the central Pacific, and two others are associated with north–south dipole variations across the North Atlantic jet. Overall, tropical influences are generally found to be larger than stratospheric influences on extratropical tropospheric variability and have a pronounced impact on the persistent, and therefore the potentially predictable, portion of that variability.

Abstract

Linear stochastically forced models have been found to be competitive with comprehensive nonlinear weather and climate models at representing many features of the observed covariance statistics and at predictions beyond a week. Their success seems at odds with the fact that the observed statistics can be significantly non-Gaussian, which is often attributed to nonlinear dynamics. The stochastic noise in the linear models can be a mixture of state-independent (“additive”) and linearly state-dependent (“multiplicative”) Gaussian white noises. It is shown here that such mixtures can produce not only symmetric but also skewed non-Gaussian probability distributions if the additive and multiplicative noises are correlated. Such correlations are readily anticipated from first principles. A generic stochastically generated skewed (SGS) distribution can be analytically derived from the Fokker–Planck equation for a single-component system. In addition to skew, all such SGS distributions have power-law tails, as well as a striking property that the (excess) kurtosis K is always greater than 1.5 times the square of the skew S. Remarkably, this K–S inequality is found to be satisfied by circulation variables even in the observed multicomponent climate system. A principle of “diagonal dominance” in the multicomponent moment equations is introduced to understand this behavior.

To clarify the nature of the stochastic noises (turbulent adiabatic versus diabatic fluctuations) responsible for the observed non-Gaussian statistics, a long 1200-winter simulation of the northern winter climate is generated using a dry adiabatic atmospheric general circulation model forced only with the observed long-term winter-mean diabatic forcing as a constant forcing. Despite the complete neglect of diabatic variations, the model reproduces the observed K–S relationships and also the spatial patterns of the skew and kurtosis of the daily tropospheric circulation anomalies. This suggests that the stochastic generators of these higher moments are mostly associated with local adiabatic turbulent fluxes. The model also simulates fifth moments that are approximately 10 times the skew, and probability densities with power-law tails, as predicted by the linear theory.