The Different Flavors of La Niña
Kevin E. TrenberthNational Center for Atmospheric Research
Boulder, CO. USA
trenbert@ucar.edu
Flavors of El Niño and La Niña
To first order, El Niño and La Niña events represent the opposite swings of the pendulum in the tropical Pacific, with El Niño representing the warm phase of the phenomenon known as ENSO (El Niño-Southern Oscillation), and La Niña the cold phase. El Niño and La Niña events come in many different flavors". The terms are not very specific but identify a class of behavior. A review of terminology and definitions is given in Trenberth (1997) and a suggestion for a quantitative definition is the following: An El Niño (La Niña) event can be said to occur if 5-month running means of sea surface temperature (SST) anomalies in the Niño 3.4 region (5°N to 5°S 170°W to 120°W) exceed +0.4°C (-0.4°C) for 6 months or more. Consequently, such a definition does not take into account SSTs in surrounding areas, the magnitude of the event, or its timing. Yet, it is well established that the tropical surface winds and rainfall patterns respond to the total SST field, not just the anomalies, and so depend on time of year and other important details. To a first approximation, winds in the equatorial regions respond to pressure gradients which are set up by SST gradients, and so the SSTs outside the Niño 3.4 region matter. Moreover, surface convergence is favored to occur in the vicinity of the warmest water in absolute terms, not anomalies. Mostly, this is because down-gradient SSTs occur in every direction from the warmest spot, and so winds blow toward the warmest spot, bringing moisture with them, and resulting in moisture convergence and organized rainfall, mostly through thunderstorms. The rotation of the Earth modifies this simple picture somewhat.
More specifically, because of the several factors involved in the dynamical response of the atmosphere to SST anomalies, and because of the SST dependence on the atmospheric surface winds and radiation, observations reveal that there is no simple linear relationship between SST and rainfall or outgoing longwave radiation (OLR) anomalies. The largest SST anomalies form in the eastern Pacific, in the cold tongue area, while the largest rainfall anomalies typically occur near the warmest water which is near the date line in most ENSO events. On time scales beyond a few weeks, the atmospheric boundary layer convergence, which leads to the organized convective areas, preferentially tends to occur near the warmest region, which may be over land on some occasions (especially in the Indian Ocean-Asian monsoon region). Also, because convection cannot occur everywhere at once, the result is effectively a competition within the atmosphere for where the convergence will occur. Because the warmest observed water is ~29 C only regions where SSTs exceed 28 C can be a competitor. But the whole pattern of SST, including the gradients and absolute values, as well as the proximity of major land masses, are all factors as important as the actual SST anomaly in determining the rainfall anomalies (Trenberth, 1989). Modeling and observational studies reveal that precipitation anomalies are primarily associated with low level moisture convergence rather than local evaporation anomalies.
Because SST gradients are weak in the tropical western Pacific, fairly small changes in SST can have profound implications as the region of warmest water shifts from one region to another perhaps thousands of kilometers away. This changes the favored location of the convergence zones, the Inter Tropical Convergence Zone (ITCZ) and the South Pacific Convergence Zone (SPCZ). Observations show that during El Niño events, the ITCZ tends to move south and the SPCZ moves northeast (Trenberth, 1976) with the result that the convergence zones come together in the western Pacific and widespread convection often extends east of the date line in the vicinity of the equator, eroding the dry zone but, to a large extent, following and lying just west of the region of warmest water. The Hadley circulation is enhanced at the expense of a weakened Walker circulation. The opposite happens with La Niña events and the equatorial region east of the dateline is cold and dry.
A consequence of these arguments is that small changes in SST can considerably alter the atmospheric convection distribution. Changes in location, extent, and time of year of SST anomalies result in differences in tropical rainfall. These project onto the changes in tropical heating and, thus, differences in vertical motion, large-scale overturning, and upper tropospheric convergence and divergence, so that there are differences in atmospheric Rossby wave forcing. In addition, the horizontal size of the anomaly affects which scales are forced and thus the propagation characteristics of the Rossby waves. These different flavors of ENSO events are important.
Links to the Extratropics
A review of global teleconnections associated with tropical SSTs has been given by Trenberth et al. (1998) and the following summarizes the main points. Large influences can clearly be seen in extratropics in all ENSO events. Factors influencing the differences between events includedifferences in SSTs in tropics as discussed above, and thus differences in tropical rainfall and differences in atmospheric Rossby wave forcing, and differences in the medium through which signal propagates. The latter varies with time of year, with location (convection relative to planetary waves), and with transients (synoptic eddy activity). Thus even with the same tropical forcing, the extratropical response can vary enormously.
The differences arise from the changing seasons and the location of the forcing relative to the background climatological planetary waves. In addition, a random component arises from weather and related natural chaotic variations in the extratropical circulation that dominate the circulation regime. Accordingly, the rather small influence from the tropical forcing can only be reliably seen in the extratropics if an average is taken over many synoptic events. A month is too short, and even one season is considered marginal in length. Longer averages are further complicated by different expectations because the background climatological flow changes significantly with the seasons. Consequently, a large random component is added in the extratropics from this aspect (Trenberth and Branstator, 1992).
The above reasoning suggests a linear addition of the tropical forcing influences on the patterns that would otherwise be present in the extratropics although the background climatological flow must be included. This is oversimplified as the changes in the circulation immediately begin to change the jet stream, and the associated storm tracks, so that the heat and vorticity fluxes by the transient eddies in the extratropics are also changed. The impact of these secondary changes, however, can be as large or larger than the influence of the original tropical forcing. Fortunately, some of these influences appear to be fairly systematic although differing considerably regionally, according to the climatological background flow (Trenberth and Hurrell, 1994).
Another view of the role of the tropical forcing on the extratropics is that it projects onto certain preferred regimes where more persistent flow patterns recur, presumably associated in some way with the distribution of land and sea and the climatological planetary waves. The suggestion is that one role of tropical SST forcing is to alter the frequency of occurrence and stability of certain pre-existing regimes but with only minor changes in the regimes themselves (Palmer, 1993).
The effects of ENSO events are global in extent. But differences among ENSO events in the tropics are important and can become magnified in the extratropics. The extratropical anomaly patterns cannot be reliably determined by statistical means since the number of examples is not large enough to stratify the ENSO events into subtypes. Consequently, the best hope for predicting the remote influences of a particular ENSO event is from better models. Linear planetary wave models (e.g., Trenberth and Branstator, 1992) are one tool that have yet to be fully exploited in this respect. Atmospheric GCMs can also be useful but it is much more difficult to separate out the signal from the weather-noise, and ensembles must be generated.
Trenberth and Branstator (1992) discuss the criteria for performance of these models if the simulations and/or predictions are to be useful. To be able to properly determine the total extratropical response and thus make forecasts, a number of modeling challenges exist. Based on our current understanding of the problem, as outlined in Trenberth et al. (1998), a modified set of criteria include the need to be able to simulate the following with fidelity:
These highlight factors of importance in the extratropical teleconnections. However, most
models do not yet satisfy these demands. Improved models will allow the predictability that does exist to be better exploited.
The 1988-89 and 1998-99 La Niña
The last major La Niña occurred in 1988-89. In the northern spring-summer of 1988 a major drought occurred over North America and has been estimated to cost $40 billion in damage plus many lives. Trenberth et al. (1988)suggested that the primary cause of the drought was a change in the atmospheric circulation across North America brought about as a teleconnection forced by changes in SSTs in the tropical Pacific. They suggested that the observed cold SSTs along the equator (La Niña) combined with anomalously warm water southeast of Hawaii led to a displacement northwards of the ITCZ in the eastern tropical Pacific as the thunderstorm activity preferentially occurred over the warmer water. This shift in the ITCZ was observed in the OLR.
Trenberth and Branstator (1992) demonstrated that major SST anomalies in the tropical Pacific Ocean, in association with the 1988 La Niña, disrupted atmospheric heating patterns by changing the location and intensity of the ITCZ. The anomalous atmospheric conditions that brought on the drought occurred in April, May, and June of 1988. The evolution of the Pacific SSTs and tropical convection was shown to be consistent with the development of the conditions favorable for initiating the drought circulation pattern in April through June of 1988. On the equator at 110°W, SST anomalies exceeded -2.75°C in only April, May and June and were largest (-4.1°C) in May 1988. Diagnostic calculations by Trenberth and Branstator (1992) of atmospheric diabatic heating confirm that atmospheric heating anomalies existed in the tropical Pacific in association with the major SST anomalies during this time. The link between the anomalous heating and the tropical SSTs supports the view that influences external to the atmosphere were important and that the drought was not generated solely by mechanisms internal to the atmosphere. It is argued that feedback-caused soil moisture anomalies may have been secondary sources for the drought circulation but could not have been the primary instigator.
The 1988 La Niña had some unique aspects to it. In particular, while most La Niña events lead to suppressed convective activity in the eastern Pacific, in 1988 the anomalously warm water southeast of Hawaii sustained a vigorous ITCZ and led to the unique flavor of this event.
In 1998, the cooling in the tropical Pacific has occurred at an unprecedented rate, of about 1°C per week in the center of the Niño 3.4 region from mid May until mid to late June; a 5°C cooling in SSTs from 29°C to 24°C, so that SST anomalies are 3°C below normal at the end of June. This also means that warm waters with above average SSTs remain in surrounding areas, for example at 10°N, at 10°S and from about 80° to 115°W along the coast of South America. This El Niño debris is likely to ensure a unique character to the 1998-99 La Niña, and a different flavor from anything seen before. No doubt the signs of the previous El Niño will wane during 1998, and a La Niña character to the global weather patterns will develop, but it should also be expected that some specific features will also occur.
Acknowledgments
This research was sponsored by NOAA Office of Global Programs under grant NA56GPO247.
References
Palmer, T. N., 1993: A nonlinear dynamical perspective on climate change, Weather, 48, 314-326.
Trenberth, K. E., 1976: Spatial and temporal variations of the Southern Oscillation. Quart. J. Roy. Meteor. Soc., 102, 639-653.
Trenberth, K. E., 1989: TOGA and atmospheric processes. Understanding Climate Change. Geophysical Monogr. 52, IUGG Vol. 7, A. Berger, R. E. Dickinson and J. W. Kidson (Editors). Amer. Geophys. U., Washington, D.C., 117-125.
Trenberth, K. E., 1997: The definition of El Niño. Bull. Amer. Met. Soc. 78, 2771-2777.
Trenberth, K. E., and G. W. Branstator, 1992: Issues in establishing causes of the 1988 drought over North America. J. Climate 5, 159-172.
Trenberth, K. E. and J. W. Hurrell, 1994: Decadal atmosphere-ocean variations in the Pacific. Climate Dynamics, 9, 303-319.
Trenberth, K. E., G. Branstator and P. A. Arkin, 1988: Origins of the 1988 North American drought, Science, 242, 1640-1645.
Trenberth, K. E., G. W. Branstator, D. Karoly, A. Kumar, N-C. Lau, and C. Ropelewski, 1998: Progress during TOGA in understanding and modeling global teleconnections associated with tropical sea surface temperatures. J. Geophys. Res. (special TOGA issue), (June). Vol. 103. 14291-14324.