Forecasting El Niño and La Niña events using decision tree classifier

Theoretical and Applied Climatology - The El Niño-Southern Oscillation (ENSO) phenomenon affects the global climate by changing temperature and precipitation patterns mainly in tropical...     

Locally and remotely forced atmospheric circulation anomalies of Ningaloo Niño/Niña

A recently identified climate mode called Ningaloo Niño (Niña) is associated with positive (negative) sea surface temperature (SST) anomalies off the west coast of Australia and negative (positive) sea level pressure (SLP) anomalies in the overlying atmosphere. By conducting a series of numerical experiments with an atmospheric general circulation model, generation mechanisms of the atmospheric circulation anomalies accompanied by Ningaloo Niño/Niña are examined. Even when SST is allowed to vary interannually only in the eastern South Indian Ocean, negative (positive) SLP anomalies are formed off the west coast of Australia in Ningaloo Niño (Niña) years, supporting the existence of local ocean–atmosphere interaction. When the model is forced by SST anomalies outside of the eastern South Indian Ocean, negative (positive) SLP anomalies are also generated in Ningaloo Niño (Niña) years owing to a Matsuno–Gill type response to atmospheric convection anomalies in the tropical Pacific. It is found that the latter impact is stronger in the current atmospheric general circulation model. Regarding climatic impacts, it is shown that Ningaloo Niño (Niña) induces wet (dry) anomalies over the northwestern part of Australia even when SST anomalies outside of the eastern South Indian Ocean are excluded from the SST forcing.     

On the Ningaloo Niño/Niña

Using both observational and reanalysis data, evolution processes of a regional climate phenomenon off Western Australia named recently “Ningaloo Niño (Niña)” are studied in detail. It is also shown that the Ningaloo Niño (Niña) has significant impacts on the precipitation over Australia. The Ningaloo Niño (Niña), which is associated with positive (negative) sea surface temperature (SST) anomalies and atmospheric anomalies off the western coast of Australia, peaks during austral summer and is classified into two types based on the difference in the evolution process. The first type called a locally amplified mode develops through an intrinsic unstable air–sea interaction off the western coast of Australia; an anomalous cyclone (anticyclone) generated by positive (negative) SST anomalies forces northerly (southerly) alongshore wind anomalies, which induce coastal downwelling (upwelling) anomalies, and enhance the positive (negative) SST anomalies further. The second type called a non-locally amplified mode is associated with coastally trapped waves originating in either the western tropical Pacific, mostly related to El Niño/Southern Oscillation, or the northern coast of Australia. Positive (negative) SST anomalies in both modes are associated with an anomalous low (high) off the western coast of Australia. The sea level pressure (SLP) anomalies in the locally amplified mode are regionally confined with a cell-like pattern and produce a sharp offshore pressure gradient along the western coast of Australia, whereas those in the non-locally amplified mode tend to show a zonally elongated pattern. The difference is found to be related to conditions of the continental SLP modulated by the Australian summer monsoon and/or the Southern Annular Mode.     

Statistical simulations of the future 50-year statistics of cold-tongue El Niño and warm-pool El Niño

Recent extensive studies have suggested that the occurrence of warm-pool El Niño has increased since the late 1970s and will increase in future climate. Occurrence frequencies of cold-tongue and warm-pool El Niño have been investigated in the observational record (1980–2006) and in the future 50 years (2007–2056) based on 100 synthetic SST datasets with estimates of statistical confidence. In the observational record, 80% of the warm-pool El Niño occurred since 1980 over a period of 27 years; only 20% of the warm-pool El Niño occurred prior to 1980 over a period of 110 years. The 100 synthetic datasets, on average, produce 142 months of cold-tongue El Niño in 2007–2056 as opposed to an average 107 months in the same length of the observational data; this is a 20.7% increase in the occurrence of cold-tongue El Niño compared with the observational period. Warm-pool El Niño occurred for 112 months in 2007–2056 as opposed to an average occurrence of 42 months in the observational record; this is 2.5 times the occurrence frequency in the 1980–2006 period in the synthetic datasets. As a result, occurrence frequencies of cold-tongue and warm-pool El Niño in the period of 2007–2056 become quite comparable to each other in the synthetic datasets. It is expected in the next 50 years that warm-pool El Niño will be nearly as frequent as cold-tongue El Niño.     

Anomalous winter climate conditions in the Pacific rim during recent El Niño Modoki and El Niño events

Present work compares impacts of El Niño Modoki and El Niño on anomalous climate in the Pacific rim during boreal winters of 1979–2005. El Niño Modoki (El Niño) is associated with tripole (dipole) patterns in anomalies of sea-surface temperature, precipitation, and upper-level divergent wind in the tropical Pacific, which are related to multiple “boomerangs” of ocean-atmosphere conditions in the Pacific. Zonal and meridional extents of those “boomerangs” reflect their independent influences, which are seen from lower latitudes in the west to higher latitudes in the east. In the central Pacific, more moisture is transported from the tropics to higher latitudes during El Niño Modoki owing to displacement of the wet “boomerang” arms more poleward toward east. Discontinuities at outer “boomerang” arms manifest intense interactions between tropical and subtropical/extratropical systems. The Pacific/North American pattern and related climate anomalies in North America found in earlier studies are modified in very different ways by the two phenomena. The seesaw with the dry north and the wet south in the western USA is more likely to occur during El Niño Modoki, while much of the western USA is wet during El Niño. The moisture to the southwestern USA is transported from the northward shifted ITCZ during El Niño Modoki, while it is carried by the storms traveling along the southerly shifted polar front jet during El Niño. The East Asian winter monsoon related anticyclone is over the South China Sea during El Niño Modoki as compared to its position over the Philippine Sea during El Niño, causing opposite precipitation anomalies in the southern East Asia between the two phenomena.     

La Niña diversity and Northwest Indian Ocean Rim teleconnections

The differences in tropical Pacific sea surface temperature (SST) expressions of El Niño-Southern Oscillation (ENSO) events of the same phase have been linked with different global atmospheric circulation patterns. This study examines the dynamical forcing of precipitation during October–December (OND) and March–May (MAM) over East Africa and during December–March (DJFM) over Central-Southwest Asia for 1950–2010 associated with four tropical Pacific SST patterns characteristic of La Niña events, the cold phase of ENSO. The self-organizing map method along with a statistical distinguishability test was used to isolate La Niña events, and seasonal precipitation forcing was investigated in terms of the tropical overturning circulation and thermodynamic and moisture budgets. Recent La Niña events with strong opposing SST anomalies between the central and western Pacific Ocean (phases 3 and 4), force the strongest global circulation modifications and drought over the Northwest Indian Ocean Rim. Over East Africa during MAM and OND, subsidence is forced by an enhanced tropical overturning circulation and precipitation reductions are exacerbated by increases in moisture flux divergence. Over Central-Southwest Asia during DJFM, the thermodynamic forcing of subsidence is primarily responsible for precipitation reductions, with moisture flux divergence acting as a secondary mechanism to reduce precipitation. Eastern Pacific La Niña events in the absence of west Pacific SST anomalies (phases 1 and 2), are associated with weaker global teleconnections, particularly over the Indian Ocean Rim. The weak regional teleconnections result in statistically insignificant precipitation modifications over East Africa and Central-Southwest Asia.     

Favorable connections between seasonal footprinting mechanism and El Niño

Previous studies suggested that the wintertime SST in the North Pacific that are generated by the concurrent North Pacific Oscillation (NPO) are able to force El Niño during subsequent winter via the so-called ‘seasonal footprinting mechanism’ (SFM). We examine how the NPO effectively generates the El Niño via the SFM in the observations and models. The occurrence ratio for El Niño under conditions of NPO forcing during the previous winters is about 41 % for the period of 61 years (1949–2009), indicating that the atmospheric forcing from the mid-latitudes through the SFM does not always trigger an El Niño. We observed certain favorable conditions under which the SFM may effectively induce El Niño. We directly compared these observations with two cases: when the wintertime NPO leads to El Niño during the following winter through the SFM, and when the wintertime NPO is not followed by El Niño. Our analysis demonstrates that the spatial structures of the NPO, associated wind speed and net heat flux in the northeast Pacific, differ between the two cases. Such differences determine the existence of a footprint SST in the northeastern Pacific during the late spring and summer, which plays a key role in initiating the El Niño via the projection of westerly wind stress anomalies onto the equatorial Pacific during the same seasons. By conducting linear baroclinic model experiments, it is found that the positions of La Niña SST forcing during the previous winter are able to modify the spatial structures of the NPO, which produces favorable conditions for the El Niño during subsequent winter via the SFM.     

Changes in El Niño and La Niña teleconnections over North Pacific–America in the global warming simulations

The change in the teleconnections of both El Niño and La Niña over the North Pacific and American regions due to a future greenhouse warming has been analyzed herein by means of diagnostics of the Intergovernmental Panel on Climate Change-Fourth Assessment Report (IPCC-AR4) coupled general circulation models (CGCMs). Among the IPCC-AR4 CGCM simulations, the composites of the eight-member multimodel ensemble are analyzed. In most CGCMs, the tropical Pacific warming due to the increase of CO₂ concentration in the atmosphere promotes the main convection centers in the equatorial Pacific associated with both El Niño and La Niña to the east. The eastward shift of the convection center causes a systematic eastward shift of not only El Niño but also La Niña teleconnection patterns over the North Pacific and America, which is demonstrated in the composite maps of 500 hPa circulation, surface temperature, and the precipitation against El Niño and La Niña, as observed in a comparison between the pre-industrial and CO₂ doubling experiments. Thus, a systematic eastward migration of convection centers in the tropical Pacific associated with both El Niño and La Niña due to a future global warming commonly causes the eastward shift of the atmospheric teleconnection patterns over the Northern Hemisphere.     

Why were some La Niñas followed by another La Niña?

This paper investigates why some La Niña events are followed by another La Niña and some others are not. We propose two preconditions that result in continuation of a La Niña. The first one is that La Niña must be a strong event (a major La Niña). This ensures that the reflected Rossby wave signal at the eastern boundary of the Pacific has a strong westward propagating cold ocean temperature anomaly over the off-equatorial region. The off-equator cold anomaly may not be conducive to the equatorial recharge process, and as a result, may favor the persistence of cold ocean subsurface temperature anomaly and prevent the transition from La Niña to El Niño. The second precondition is whether there are eastward propagating downwelling Kelvin waves during the decay phase of a major La Niña. Eastward propagating downwelling Kelvin waves could lead to demise for a tendency for a follow-up La Niña. The equatorial Kelvin wave activities are associated with fluctuations of surface wind in the equatorial far-western Pacific. The analysis suggests that both the surface wind in the equatorial far-western Pacific and the recharge/discharge of the equatorial Pacific are indicators for occurrence or no occurrence of a follow-up La Niña event.     

Impact of El Niño onset timing on the Indian Ocean: Pacific coupling and subsequent El Niño evolution

A relation between the timing of the El Niño onset and its subsequent evolution is examined by emphasizing its association with the Indian Ocean (IO) SST variation. Two types of El Niño events based on the timing of their onset are classified and their characteristics are examined and compared. In general, spring onset (SP) events grow greater in magnitude and their evolutions have a faster transition. On the contrary, summer onset (SU) events are relatively weaker in magnitude and have a slower transition. Moreover, in contrast to the SU events, the SP events have a strong tendency for accompanying an IO dipole and basin-wide type of warming pattern in the El Niño developing and mature phases, respectively. It is demonstrated here that the distinctive evolutions in transition phase of the two events are resulted from the difference in IO SST. The warm IO SST in the SP El Niño event, lead an anomalous easterlies over the western Pacific, which forces a fast termination of El Niño events.     

Spatial variability and mechanisms underlying El Niño-induced droughts in Mexico

The El Niño Southern Oscillation plays a key role in modulating interannual rainfall variability in Mexico. While El Niño events are linked to drought in Mexico, uncertainty exists about the spatial pattern and causal mechanisms behind El Niño-induced drought. We use lead/lag correlation analysis of rainfall station data to identify the spatial pattern of drought associated with the summer before, and the spring following, the peak of warm SST anomalies in the eastern equatorial Pacific. We also use atmospheric fields from the North American Regional Reanalysis to calculate the anomalous moisture budget and diagnose the mechanisms associated with El Niño-induced drought in Mexico. We find that reduced rainfall occurs in Mexico in both the summer before and the spring after a peak El Niño event, especially in regions of climatologically strong convection. The teleconnection in the developing phase of El Niño is primarily driven by changes in subsidence resulting from anomalous convection in the equatorial Pacific. The causes of drought during the decaying phase of El Niño events are varied: in some years, descent anomalies dominate other moisture budget terms, while in other years, drying of the boundary layer on the Mexican plateau is important. We suggest that the latter may result from the interaction of weakened southeasterly winds in the Intra-Americas Sea with high topography along the Atlantic coast of Mexico. Weakened winds are likely driven by a reduced sea level pressure gradient between the Atlantic and the Pacific. Changes in easterly wave activity may contribute to drought in the developing phase of El Niño, but may be less important in the decaying phase of El Niño.     

The El Niño stochastic oscillator

A stochastic model is fitted to the observed NINO3.4 time series between 1951–1995. The model is nothing more than the complex version of a first-order autoregressive process. The autocorrelation of this stochastic oscillator model is an exponentially decaying cosine, specified by three parameters: a period, a decay time, and a phase shift. It fits the observed NINO3.4 autocorrelation quite well. Anomalies during an El Niño can be characterized to a large extent by a single, irregularly oscillating, index. Equatorial wave dynamics and delayed-oscillator models have been used to explain this behaviour, and it has been suggested that El Niño might be a stable phenomenon excited by weather noise. Assuming this is the case, the stochastic oscillator has a direct physical interpretation: the parameters of the oscillation can be linked to dynamical models of the delayed-oscillator type, and the noise terms represent random influences, such as intraseasonal oscillations. Long Monte Carlo simulations with the stochastic oscillator show substantial decadal variability and variation in predictability. The observed decadal variability is comparable, except for the rather large rise in the long-term mean around 1980. The observed seasonal dependence of El Niño behaviour is not compatible with the natural variability of a stationary stochastic oscillator. Formulating the model in terms of standardized anomalies takes into account some of the seasonal dependence. A stochastic oscillator forecast model has a skill approaching that of more comprehensive statistical models and may thus serve as an appropriate baseline for the skill of El Niño forecasting systems.     

Precipitation response to La Niña and global warming in the Indo-Pacific

Recent studies have highlighted the nonlinear rainfall response to El Niño sea surface temperature (SST) events in the Indo-Pacific region and how this response might change over coming decades. Here we investigate the response to La Niña SST anomalies with and without global warming by performing idealised SST-forced experiments with an atmospheric general circulation model. The La Niña SST anomaly is multiplied by a factor \(1 \le \alpha \le 4\) and added to climatological SSTs. Similar experiments using El Niño SST anomalies were previously performed, in which large nonlinearities in the precipitation response were evident. We find that: (i) Under current climatic conditions, as \(\alpha\) increases, the precipitation responds in three ways: the intertropical convergence zone (ITCZ) dries and moves poleward, the maximum precipitation along the equator moves west, and the South Pacific convergence zone (SPCZ) narrows, intensifies, and elongates. For weak ( \(\alpha = 1\) ) La Niña events, the precipitation anomalies approximately mirror those from the El Niño events along the ITCZ and SPCZ, though there are some marked differences in the central-eastern Pacific. For stronger La Niña events ( \(\alpha > 1\) ), precipitation responds nonlinearly to SST anomalies, though the nonlinearities are smaller and differ spatially from the nonlinearities in the El Niño runs. (ii) The addition of a global warming SST pattern increases rainfall in the western Pacific and SPCZ, enhances the narrowing of the SPCZ, and increases the nonlinear response in the western Pacific. However, large La Niña events reduce the impact of global warming along the central-eastern equatorial Pacific as the global warming and La Niña SST anomalies have opposite signs in that region. (iii) The response to La Niña SST anomalies is driven primarily by changes in the atmospheric circulation, whereas the response to the global warming SST pattern is mainly driven by increases in atmospheric moisture. (iv) Large changes in La Niña-driven rainfall anomalies can occur in response to global warming, even if the La Nina SST anomalies relative to the warmer background state are completely unchanged.     

The El Niño Southern Oscillation and economic growth in the developing world

The El Niño Southern Oscillation (ENSO) affects weather around the globe, particularly in regions where developing countries typically lie. These countries are known to be most vulnerable to weather anomalies, and ENSO thereby has the potential to influence their economic growth. In this study, we investigate the effect of ENSO on economic growth in 69 developing countries, using annual data from 1961 to 2015. We find regime-dependent nonlinearity in the growth response to ENSO shocks. An El Niño event, equivalent to a 1 °C deviation in sea surface temperatures in the Niño3.4 region of the equatorial Pacific, results in one-to-two percent annual growth reduction during the El Niño regime, but the effect is absent during the La Niña regime. In addition, we find that the effect of El Niño is twice-as-large in the tropics relative to temperate areas, and particularly pronounced in Africa and Asia-Pacific. The findings of this study have two important implications. From the modeling standpoint, we find that the growth impacts of ENSO shocks are nonlinear, and vary across regions and climatic zones. From the policy-making standpoint, our findings suggest opportunities for short-term adjustments to climate shock management and international aid programs, depending on the existing state and the intermediate-term patterns of the ENSO cycle.     

Factor analysis for El Niño signals in sea surface temperature and precipitation

Maximum likelihood factor analysis (MLFA) is applied to investigate the variables of monthly Tropical Pacific sea surface temperatures (SST) from Niño 1+2, Niño 3, Niño 3.4, and Niño 4 and precipitation over New South Wales and Queensland of eastern Australia, Kalimantan Island of Indonesia, and California and Oregon of the west coast of the United States. The monthly data used were from 1950 to 1999. The November-February SST with time leads of 0, 1, 2, and 3 months to precipitation are considered for both El Niño warm phases and non El Niño seasons. Interpretations of the factor loadings are made to diagnose relationships between the SST and precipitation variables. For El Niño signals, the rotated FA loadings can efficiently group the SST and precipitation variables with interpretable physical meanings. When the time lag is 0 or 1 month, the November–February El Niño SST explains much of the drought signals over eastern Australia and Kalimantan. However, when the time lag is 2 or 3 months, the same SST cannot adequately explain the precipitation during January–May over the two regions. Communality results of five factors for precipitation indicate nearly 100% explanation of variances for Queensland and California, but the percentages are reduced to only about 30% for Oregon and Kalimantan. Factor scores clearly identify the strongest El Niño relevant to precipitation variations. Principal component factor analysis (PCFA) is also investigated, and its results are compared with MLFA. The comparison indicates that MLFA can better group SST data relevant to precipitation. The residuals of MLFA are always smaller than the PCFA. Thus, MLFA may become a useful tool for improving potential predictability of precipitation from SST predictors.     

Nonlinear precipitation response to El Niño and global warming in the Indo-Pacific

Precipitation changes over the Indo-Pacific during El Niño events are studied using an Atmospheric General Circulation Model forced with sea-surface temperature (SST) anomalies and changes in atmospheric CO₂ concentrations. Linear increases in the amplitude of the El Niño SST anomaly pattern trigger nonlinear changes in precipitation amounts, resulting in shifts in the location and orientation of the Intertropical Convergence Zone (ITCZ) and the South Pacific Convergence Zone (SPCZ). In particular, the maximum precipitation anomaly along the ITCZ and SPCZ shifts eastwards, the ITCZ shifts south towards the equator, and the SPCZ becomes more zonal. Precipitation in the equatorial Pacific also increases nonlinearly. The effect of increasing CO₂ levels and warming SSTs is also investigated. Global warming generally enhances the tropical Pacific precipitation response to El Niño. The precipitation response to El Niño is found to be dominated by changes in the atmospheric mean circulation dynamics, whereas the response to global warming is a balance between dynamic and thermodynamic changes. While the dependence of projected climate change impacts on seasonal variability is well-established, this study reveals that the impact of global warming on Pacific precipitation also depends strongly on the magnitude of the El Niño event. The magnitude and structure of the precipitation changes are also sensitive to the spatial structure of the global warming SST pattern.     

Westerly wind events,diurnal cycle and central Pacific El Niño warming

In the past three decades, the strongest central Pacific (CP) El Niño event was observed in 2009–2010 by satellites. When intensity of this CP El Niño reached its maximum, large diurnal variations of sea surface temperature (SST) were also observed from tropical atmosphere ocean moorings in the central equatorial Pacific. Solar radiation in the equatorial central Pacific is larger than 140 W/m², which leads to the amplitude of diurnal cycle of SST primarily determined by large-scale wind patterns. Intraseasonal westerly wind events (WWEs) can lead to an eastward displacement of the warm pool and also can weaken the trade winds in central Pacific. When the occurrence of equatorial WWEs is more than 20 days in a month, monthly mean wind speed in central equatorial Pacific has high possibility of wind speed less than 3 m/s, thus has pronounced diurnal cycle of SST. The diurnal cycle of SST will rectify daily mean SST. Reduced mixing at the base of the mixed layer and suppression of entrainment due to the accumulated effect of diurnal cycle may lead to warmer SST in the following month. This study suggests the occurrence of more diurnal SST events may contribute to the increasing intensity of the CP El Niño events.