Satellite derived wind, rain rate and sea surface temperature data combined with NCEP analysis data are used to examine the structure of the South Pacific Convergence Zone (SPCZ) during La Nina conditions. Annual means indicate that the axes of maximum surface convergence and rain rates are essentially co-located in the Northern Hemisphere Inter-Tropical Convergence Zone and over the tropical (TR) portions of the SPCZ. However, over subtropical (ST) and middle latitudes (ML), the axis of maximum convergence is eastward and equatorward of the axis of maximum rain rates such that surface divergence predominates over the these portions of the SPCZ rain rate maximum. Analysis of NCEP data also exhibit a similar convergence distribution at 1000 hPa; however, at the 925, 850 and 700 hPa levels convergence replaces divergence and is co-located with the axis of maximum rain rates. Seasonal rain rate patterns indicated that the TR and ST SPCZ are more convectively active during Austral Summer and Fall while the ML SPCZ is more active during the Fall and Winter. Vertical wind shear and the variability of the meridional flow indicate the ML SPCZ is strongly associated with the mid-latitude storm track while the TR SPCZ is associated with low wind shear. The ST SPCZ represents a transition zone between an environment dominated by low shear and one that is more baroclinic. Lag correlation analysis using rain rate, 200 hPa divergence and height anomaly data as well as composite analysis of high rain fall events revealed wave like patterns over the ST and ML SPCZ. The estimated wavelengths range from 4700 to 5500 km, values that correspond to wave numbers 7 and 8; individual anomalies tend to move eastward at speeds of 5 to 7 m/s. The vertical structure of the height anomalies suggest baroclinic dynamics are important to the development of the mean rainfall pattern in the ML SPCZ and to a lesser extent the ST SPCZ. During high rain rate events, only a given portion of the SPCZ, approximately 3000 to 4000 km in length, is convectively active.

Data from satellites and reanalysis products are analyzed to study the behavior of wave trains in the three major Southern Hemisphere Convergences zones: the South Pacific, the South Atlantic, and the South Indian. Using composites on high rain-rate days, a wave pattern is identified that is characteristic of high rain events. This wave pattern is then compared to the patterns of variability of brightness temperature using empirical orthogonal functions. A linear regression technique is used to examine the behavior of potential vorticity corresponding to the patterns of maximum variance. Planetary-scale waves, propagating in favorable regions, slow and break, dragging streamers of moisture from the tropics into higher latitudes. These streamers, combined with lifting, lead to the enhanced rain seen in the Southern Hemisphere’s convergence zones. It is concluded that the convergence zones are areas of enhanced streamer activity and that a more thorough study of streamers will yield more information on the structure and behavior of the convergence zones.

The El Niño-Southern Oscillation (ENSO) dominates global interannual climate variability. However, the imprint of anthropogenic climate change hinders understanding of natural ENSO variability. Model predictions of the response of future ENSO variability to anthropogenic forcing are highly uncertain. A better understanding of how ENSO operates during different mean climate states may improve predictions of its future behavior. This study develops a technique to quantify the response of tropical Pacific sea surface temperature and salinity to ENSO variations. This analysis defines expected regional relationships between ENSO forcing and the tropical Pacific climate response. For example, the western tropical Pacific records El Niño events with greater skill than La Niña events; whereas the oceans near the South Pacific Convergence Zone (SPCZ) preferentially record La Niña events. This baseline understanding of regional skill calibrates interpretations of both modern and pre-instrumental coral geochemical climate proxy records. A suite of monthly resolved 18O variations in a fossil corals (Porites spp.) from the tropical western Pacific (Papua New Guinea) and the SPCZ (Vanuatu) are used to develop case studies of ENSO variability under external forcing conditions that differ from the modern climate. A record from Misima, Papua New Guinea (1411-1644 CE) spans a period of reduced solar forcing that coincides with the initiation of the Little Ice Age. This record indicates that the surface ocean in this region experienced a small change in hydrologic balance with no change in temperature, extended periods of quiescence in El Niño activity, reduced mean El Niño event amplitudes, and fewer large amplitude El Niño events relative to signals captured in regional modern records. Several multidecadal (~30-50 year) coral records from Tasmaloum, Vanuatu during the Medieval Climate Anomaly (~900-1300 CE), a period of increased solar forcing, depict ENSO variability that is generally lower than modern times. However, these records often cannot be distinguished from 20th century ENSO variability due to ENSO variability uncertainty associated with record lengths. Neither record can be tied to concurrent changes in solar or volcanic forcing, calling into question the paradigm of ENSO variability being predominantly mediated by external forcing changes on multidecadal time scales. / text

Comprendre les mécanismes moteurs et pouvoir anticiper l'impact environnemental du phénomène El Niño Oscillation Australe (ENOA) constituent des enjeux scientifiques et sociétaux de première importance, notamment pour les Pays Emergents. Dans cette thèse, nous avons documenté et contrasté la signature de différents types d'ENOA - dits canoniques et Modoki - pour plusieurs variables climatiques essentielles (température et salinité de surface, niveau de la mer, courant de surface, précipitation, vent de surface, ...), analysé la pertinence de la théorie dite de recharge / décharge, une des quatre théories majeures d'ENOA, à rendre compte ou non de la nature quasi oscillatoire de ces différents types et quantifié l'impact potentiel des modifications des caractéristiques majeures d'ENOA sur notre interprétation des tendances climatiques à 'long' terme pour ces variables climatiques essentielles.