Home

Contribution à l`étude de la variabilité climatique de l`Océan

1. Vincent D G The south pacific convergence zone SPCZ a review Mon Wea Rev 122 1949 1970 1994 Wyrtki K Fluctuations of the dynamic topography in the Pacific Ocean J Phys Oceanogr 5 450 459 1975 Wyrtki K A southward displacement of the subtropical gyre in the south Pacific during the 1982 1983 El Ni o Tropical ocean atmosphere Newsletter 23 14 15 1984 Xie P and P Arkin Global precipitation a 17 year monthly analysis based on gauge observations satellite estimates and numerical model outputs Bull Amer Meteor Soc 78 2539 2558 1997 LATITUDE Q 20 S 305 30 N 20 N LATITUDE 8 Mean Precipitation 1979 92 m year LATITUDE 140 E 160 80 160 W 140 W 120 W 100 W 80 W LONGITUDE Figure 1 Mean a sea surface temperature in C b sea surface salinity and c precipitation m year averaged over 1979 1992 from Delcroix 1998 The South Western Tropical Pacific zone 24 S 10 S 160 E 140 W is indicated by the rectangle The black dots in a denote Darwin Australia and Tahiti French Polynesia 1400 Number of observations 01 76 01 80 01 84 01 88 01 92 01 96 01 00 Time Figure 2 Monthly number of bucket crosses and thermosalinograph circles measurements of sea surface salinity in the South Western Tropical Pacific from 1976 to 2000 10 S TE Solomon 160 E 170 E 180 W 170 W 160 W 150W 1
2. 1998 Thick lines represent interdecadal trend see text y3 CORR GE ET AL LITTLE ICE AGE SEA SURFACE TEMPERATURE VARIABILITY 3479 Table 1 Instrumental and reconstructed mean SST at Am d e Lighthouse over several time intervals 8 0 SST calculated with RMA equation and quarterly ORSTOM SST from Quinn et al 1998 ORSTOM SST from daily bucket SST measurements GISST2 2 from GOSTA plus CD ROM These are regional SST TE Trace elements Period 5 OSST ORSTOMSST GISST22 TESST 1967 1993 23 5 C 23 4 C 24 7 C 1903 1993 23 0 C 24 5 C 1801 1900 22 2 C 1701 1800 21 8 C 1701 1761 21 6 C 22 0 C is potentially identified as an El Ni o year Ortlieb 2000 but is also affected by a cooling caused by a major volcanic eruption Crowley et al 1997 When the strong interdecadal SSTA clearly highlighted by the 169 month filter see Fig 1 is taken as a baseline it appears that interannual SSTA from 1701 to 1761 are similar to modern values and are of the order of 0 5 C to 1 5 C implying no change in the regional ENSO signature This is consistent with the model results indicating that pre industrial CO levels have at least to be multiplied by four i e twice as much as the present day value before any change in ENSO frequency or amplitude can be detected Collins 2000 Spectral analysis of the 1701 61 SST record did not identify significant spectral peaks but the recurrence of negative SSTA linked to El Nifio at
3. 1999 L appareil install mesure la temp rature et la salinit de surface toutes les 15 secondes une m diane est calcul e sur 5 minutes ces valeurs m dianes sont stock es avec le temps et la position GPS sur le disque dur d un PC situ la passerelle les valeurs m dianes calcul es sur 1 heure sont transmises en temps r el via le syst me GOES La transmission temps r el est conforme aux imp ratifs des programmes international GODAE et national MERCATOR d oc anographie op rationnelle Le thermosalinographe est nettoy avec une 6 solution de Triton 1 lors de chaque escale 4 Noum a et les tuyauteries install es par nos soins qui relient le syst me de refroidissement du navire au thermosalinographe sont chang es tous les ans L appareil est galement envoy chez le fabricant SeaBird USA tous les ans pour talonnage des capteurs Des pr l vements d eau de mer au niveau du thermosalinographe ont t r alis es par les officiers du bord ou par le personnel de la machine raison de deux chantillons minimum par jour Ils visaient minimiser l effet de la d rive temporelle des capteurs de salinit li e l encrassage de la cellule par ajustement des valeurs du thermosalinographe aux valeurs des chantillons Cette d rive peut tre estim e environ 0 1 psu mois pour la salinit Ces tentatives n ont h las pas t tr s fructueuses car de nombreuses mesures issues des chanti
4. Bindoff Interannual and decadal temperature Variability in the southwest Pacific Ocean between 1955 and 1988 J Clim 10 1035 1049 1997 Latif M R Kleeman and C Eckert Greenhouse wanning decadal variability or El Ni o An at tempt to understand the anomalous 1990s J Clim 10 2221 2239 1997 Le Comec F and T Corr ge Determination of uranium to calcium and strontium to calcium r tios in corals by Inductively Coupled Plasina Mass Spectrometry J Anal Atom Spect 12 969 973 1997 Linsley B K R B Dunbar G M Wellington and D A Mucciarone A coral based reconstruction of intertropical convergence zone variability over Central America since 1707 J Geophys Res 99 9977 9994 1994 Min R G R L Edwards F W Taylor J R cy C D Gallup and J W Beck Annual cycles of U Ca in coral skeletons and U Ca thennoinetry Geochim er Cosmochim Acta 39 2025 2042 1995 Paillard D L Labeyrie and P Yiou Macintosh program performs tane series analysis Eos Trans AGU 77 39 379 1996 Pelletier B S Calmant and R Pillet Current tec tonics of the Tonga New Hebrides region Earth Planet Sci Lett 164 263 1998 Quinn TM T J Crowley and F W Taylor New stable isotope results from a 73 year coral from Espiritu Santo Vanuatu Geophys Res Len 22 3413 3416 1996 Reynolds D and T Sinith improved global sea surface temperature analyses using optimum inter polation J Cl
5. Recherche pour le D veloppement Paleofiji cruise The core 2 6 m long was retrieved on the external slope of Nadala Bay fringing reef at a waterdepth of 2 m We collected an average of 15 samples per annual band along the maximum growth axis Organic matter and inorganically precipitated aragonite were removed following the procedure described by Boiseau and Juillet Leclerc 1997 Samples were analyzed with a VG Optima mass spectrometer The data are expressed in the conventional delta notation relative to the V PDB Vienna Pee Dee Belemnite reference standard Coplen 1993 The analytical uncertainty i e the standard deviation calculated from measurements of a carbonate standard is LE BEC ET AL FIJI CORAL O RECORD OF ENSO DRIVEN SSS VARIABILITY 0 067 o 26 and n 818 Taking into account the intra sample isotopic variability from replicate measurements the total uncertainty is 0 082 o The chronology is based on peak matching between 6 O and trace elements data not shown on one hand and on instrumental SST on the other hand The age model uncertainty is 1 month Monthly 6 O data were extrapolated using the AnalySeries software Paillard et al 1996 To calibrate the coral 6 O signal we used monthly SST Reynolds and Smith 1994 SSS Delcroix 1998 and rainfall Xie and Arkin 1996 averaged in a 2 latitude by 10 longitude grid centered on 16 S 175 E Results and Discussion We first compared the cor
6. a time series of SST diamonds and dotted line from Am d e Lighthouse New Caledonia and U Ca circles and solid tine from an adjacent Porites coral and b regression of U Ca ratios against SST and the resulting equation CORREGE ET AL STRONGER ENSO EVENTS IN A MID HOLOCENE MASSIVE CORAL 467 E 5 i H 5 EN 2 S o 3 d D 9 u 4161415 BP 4149415 BP 4133 15 BP 4166415 BP 29 cl SB LL 28 B e7t 27 a J De 26 se Luc 26 25 r T mr T T T 25 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Le gt a 3 mn gt ro z i 4133 15 BP gt 2 2 Figure 3 Mid Holocene coral derived SST data a Raw time series of the Sr Ca solid line and U Ca dashed line data used to reconstruct the paleo SST b Composite SST record derived from the Sr Ca and U Ca analyses and resampled at monthly intervals The ages are based on a U Th date taken from Beck et al 1997 c Mean monthly SST solid line plus and mi nus the associated mean monthly standard deviation dashed lines d Monthly SSTA dashed line with respect to the 47 year period average SST data are filtered with a 25 month Hanning filter and 24 month running annual amplitude solid line error The potential effect of Sr Ca change of seawater through time on the Sr thermometer in coral discussed by Stoll and Schrag 1998 does not appear to affect our Vanuatu record since there is a
7. ge T T Delcroix J R cy J W Beck G Cabioch and F Le Comec Evidence for stronger ENSO events in a mid Holocene massive coral Paleoceanography 15 465 470 2000 Crowley T J T M Quinn F W Taylor C H nin and P Joannot Evidence for a volcanic cooling signal in a 335 year coral record from New Caledonia Paleoceanography 12 633 639 1997 Delcroix T and O Lenormand ENSO signals in the vicinity of New Caledonia South Western Pacific Oceanol Acta 20 481 491 1997 Gagan M K L K Ayliffe D Hopley J A Cali G E Mortimer J Chappell M T McCulloch and M J Head Temperature and surface ocean water balance of the mid Holocene tropical western Pacific Science 279 1014 1018 1998 Gagan M K L K Ayliffe J W Beck JE Cole E R M Druffel R B Dunbar and D P Schrag New views of tropical paleoclimates from corals Quat Sc Rev 19 45 64 2000 Keigwin L D The Little Ice Age and medieval warm period in the Sargasso Sea Science 274 1504 1508 1996 Lau K M and H Weng Interannual decadal interdecadal and global warming signals in sea surface temperature during 1955 97 J Clim 12 1257 1267 1999 Le Cornec F and T Corr ge Determination of uranium to calcium and strontium to calcium ratios in corals by Inductively Coupled Plasma Mass Spectrometry J Anal Atom Spect 12 969 973 1997 Mantua N J S R Hare Y Zhang J M Wallace and R C Francis A Pacific interdecadal c
8. isotopic record Is this interdecadal variability linked to the Pacific inter Decadal Oscillation PDO Mantua et al 1997 and involves exchanges between high and low latitudes Recently Biondi et al 2001 proposed an annual reconstruction of the PDO based on tree rings from the southwestern part of America which extends back to 1661 Comparison of the AL SST record annual means with this PDO indicates an overall good correlation between the two records R 0 45 significant at 1 level The evolutive correlation is shown on Fig 3 During the warmest periods the modem day pattern is present i e mean SST raise at AL when the PDO is negative When SST is at its lowest the relationship seems to reverse This result emphasizes the fact that as pointed out elsewhere Alverson et al 2001 more data and better spatial coverage are needed before we can fully understand how interactions between the tropics and higher latitudes can drive climate change on decadal timescales Acknowledgements We thank F Taylor Y Join and B Pelletier for collecting the Am d e coral core and Pascal Yiou for performing spectral analysis on our record This article was greatly improved by reviews from W Beck Y Gouriou and two anonymous reviewers and by comments from Yves M Tourre Data other than our own were provided by the WDC A for Paleoclimatology and the UK Met Office This work was supported by IRD and NSF Supporting data will be archived
9. 1987 year and the beginning of 1988 an eastward anomaly developed gt 5 cms plate 2d During the 1997 98 El Ni o a westward velocity anomaly gt 10 cm s occurred between 180 W and 165 W from the beginning of 1998 to September 1998 plate 2b Then an eastward velocity anomaly reaching 10 cm s occurred from the beginning of 1999 to July 1999 In all these cases the westward velocity anomaly preceded by 2 3 months the westward extension of the SSS salinity tongue and the eastward retraction of this tongue coincided with the eastward velocity anomaly plate 2a and 2c strongly suggesting that qualitatively zonal geostrophic advection plays a role in the east west movement of the zonal front The meridional current anomalies computed from TOPEX Poseidon altimeter data vary within 2 cms from 1992 to 1999 not shown A northward anomaly develops at 17 S 180 W during the second part of 1988 This is consistent with the westward anomaly observed at the same position and the northwestward displacement of the SSS front although this anomaly is very slight 2 3 cms To obtain a complete picture the effect of the interannual anomalies of the Ekman drift in front displacements has to be estimated While the time function of the first EOF of the zonal component of pseudo windstress is not correlated with the SOI there is a very close correlation Rimax 0 88 with 1 month time lag between the time function of the first EOF of the meridiona
10. 8d Nicet and Delcroix 2000 showed that ENSO related precipitation changes range within 20 50 of the mean annual values 4 14 Evaporation In order to test the potential effect of evaporation changes on SSS variability we used the da Silva et al 1994 data covering the 1976 1993 period The 1976 1993 averaged evaporation values not shown are minimum below the mean SPCZ axis 0 13 m month and maximum south of it between Fiji and New Caledonia 0 15 m month The first EOF not shown extracts a clear seasonal 12 signal in the SWTP with maximum variability south of 16 S and minimum variability along 10 S The amplitude from maximum to minimum of the seasonal variation is about 0 08 m month weaker than the seasonal precipitation cycle This means that the seasonal variation in the Precipitation minus Evaporation difference is mainly dependent on precipitation The first EOF of the low frequency evaporation signal not shown is relatively flat and does not show any correlation with the SOI 4 2 Time space evolution of the frontal zone 4 2 a SST SSS and precipitation fields Time longitude and time latitude plots of the SSS SST and precipitation fields were drawn up to further assess the relationship between changes in these parameters and ENSO We chose to use the time longitude plot at the latitude of 17 S Plate la 1c le where the SSS front is strongest Figure 4a and time series are complete for every lon
11. AL still lies in the typical ENSO band of 2 to 7 years Fig 1 To further investigate ENSO behavior in colder times we compared the December to February DJF SOI derived from tree rings Stahle et al 1998 to the March to May MAM Am d e SST to account for the three month lag present today The two records are statistically correlated R 0 35 correlation coefficient 1700 1710 1720 1730 1740 1750 1760 year Figure 3 Correlation coefficients between mean annual SST at Am d e Lighthouse and the reconstructed SOI dashed line and PDO solid line records for 15 year overlapping subperiods see text for reference of records For each subperiod the calculated correlation coefficient was assigned to the mid point The length of the subperiods was chosen to highlight interdecadal variability Results are not significantly altered by choosing slightly different subperiod length Full and dashed horizontal lines represent 5 and 1 significance levels respectively For the correlation with the northern hemisphere winter DJF reconstructed SOI the mean of three months of Am d e Lighthouse SST namely March April and May was used significant at 1 level However close examination of the records indicates that the correlation evolves with time To visualize this tendency we took a sliding window of 15 years and calculated the correlation coefficient for each block assigned to the mid point of the period
12. CTD observations were collected in the study area Before being entered in the data base the following validation tests were applied to the bucket and TSG measurements a for each voyage data were plotted on a map to detect position errors and obvious outliers b within the tropics SSS lower than 30 or greater than 37 were rejected TSG data were also visually compared to mean monthly values 1 standard deviation For the purpose of the present study an additional validation test was then applied Data were grouped in 2 latitude by 10 longitude bins In a given bin individual measurements greater lower than the climatological monthly value plus or minus 5 4 and 3 5 times the standard deviation were rejected At the end of the additional test 1 of the SSS and 2 of the SST data were rejected The data were then averaged on a monthly basis and gridded 2 latitude x 10 longitude x 1 month using a triangle based linear interpolation The number of SSS data per month in the area under study is displayed in Figure 2 The SST data distribution presents almost the same pattern not shown Between 200 and 400 measurements were taken per month before 1992 when the VOS were equipped with TSGs After that date there is a significant increase up to 1200 month The mean number of measurements per month in every bin is shown in Figure 3 There are fewer than 5 measurements per month in the center of the region and east of 170 E south of 22 S and no
13. Mon Weather Rev 122 1949 1970 Wyrtki K and J Wenzel 1984 Possible gyre gyre interaction in the Pacific ocean Nature 309 538 540 Seasonal and ENSO Variations of Sea Surface Salinity and Temperature in the South Pacific Convergence Zone during 1976 2000 Yves Gouriou and Thierry Delcroix Centre IRD B P A5 98 848 Noumea New Caledonia e mail yves gouriou noumea ird nc Accept dans Journal of Geophysical Research 2001 830 October 2001 Index term 4215 Climate and interannual variability 4223 Descriptive and regional oceanography 4522 El Ni o Key Words ENSO Variability Sea Surface Salinity Sea Surface Temperature South Pacific Convergence Zone Front Abstract Sea surface salinity SSS and temperature SST data collected from voluntary observing ships over 25 years 1976 2000 are analyzed in the South Western Tropical Pacific 10 S 24 S 160 E 140 W This region lies under the South Pacific Convergence Zone SPCZ at the southern edge of the western Pacific warm pool between Tahiti and Darwin the two places whose atmospheric sea level pressure difference is used to define the Southern Oscillation Index SOD Complementary data such as precipitation are used to assist in the analysis The mean and seasonal variations of these parameters are described An Empirical Orthogonal Function EOF analysis of low pass filtered time series is then performed to extract the interannual varia
14. O treatment of recent coral aragonite oxygen and carbon isotopic implications Chem Geol 143 171 180 1997 Boiseau M A Juillet Leclerc P Yiou B Salvat P Isdale and M Guillaume Atmospheric and oceanic evidences of ENSO events in central Pacific Ocean from stable coral isotopic records over the last 137 years Paleoceanography 13 6 671 685 1998 Charles C D D E Hunter and R G Fairbanks Interaction between the ENSO and the Asian monsoon in a coral record of tropical climate Science 277 925 928 1997 Cole J E R B Dunbar T R McClanahan and N A Muthiga Tropical Pacific forcing of decadal SST variability in the Western Indian Ocean over the past two centunes Science 287 617 619 2000 Cole J E and R G Fairbanks The Southern Oscillation recorded in the 6 O of corals from Tarawa atoll Paleoceanography 5 5 669 683 1990 Coplen T B Reporting of stable carbon hydrogen and oxygen isotopic abundances In Reference and intercomparison materials for stable isotopes light elements Int At Energy Agency Tech Doc 825 pp 31 38 Vienna 1993 Corr ge T T Delcroix J R cy J W Beck G Cabioch and F Le Comec Evidence for stronger ENSO events in a mid Holocene massive coral Paleoceanography 15 4 465 470 2000 Delaygue G J Jouzel and J C Dutay Oxygen 18 Salinity relationship simulated by an oceanic general circulation model EPSL 178 113 123 2000 Delcroix T a
15. at the WDC A for Paleoclimatology References Alibert C and M T McCulloch Strontium calcium ratios in modern Porites corals from the Great Barrier Reef as a proxy for sea surface temperature calibration of the thermometer and monitoring of ENSO Paleoceanography 12 345 363 1997 Alverson K G W K Moore G Holdsworth and J E Cole Improving climate predictability and understanding decadal 3480 CORREGE ET AL LITTLE ICE AGE SEA SURFACE TEMPERATURE VARIABILITY variability using proxy climate data CLIVAR Exchanges 6 1 4 5 2001 Barnett T P and M Latif Connections between the Pacific Ocean tropics and midlatitudes on decadal timescale J Clim 13 1173 1194 2000 Beck J W R L Edwards E Ito F W Taylor J R cy F Rougerie P Joannot and C H nin Sea surface temperature from coral skeletal strontium calcium ratios Science 257 644 647 1992 Biondi F A Gershunov and D R Cayan North Pacific decadal climate variability since 1661 J Clim 14 5 10 2001 Blackman D S and J W Tukey The measurement of Power Spectra 190pp Dover Mineola N Y 1958 Bradley R S and P D Jones Little Ice Age summer temperature variations their nature and relevance to recent global warming trends The Holocene 3 367 376 1993 Collins M The El Ni o southern oscillation in the second Hadley Centre coupled model and its response to greenhouse warming J Clim 13 1299 1312 2000 Corr
16. avance ses anomalies constituent des enjeux scientifiques techniques et socio conomiques incontestables Le ph nom ne El Ni o Southern Oscillation ENSO est le signal climatique qui domine l chelle interannuelle dans le Pacifique tropical L analyse d observations in situ et altim triques ainsi que les r sultats de mod les ont permis ces derni res ann es de suivre certains ph nom nes en temps r el et de les pr voir environ 6 mois en avance En d pit de ce succ s relatif notre compr hension d ENSO demeure incompl te li e en partie aux manques d observations sur l oc an Le Pacifique tropical sud ouest est une r gion d int r ts multiples la fois scientifiques et socio conomiques Cette r gion englobe une grande partie du r servoir d eau chaude du Pacifique o se trouvent les eaux les plus chaudes de la plan te gt 28 C Elle est galement situ e entre les deux syst mes m t orologiques du Pacifique qui d finissent l indice d Oscillation Australe Tahiti et Darwin et cette position g ographique unique a sugg r certains auteurs que l origine d ENSO devait y tre recherch e Van Loon and Shea 1985 Holbrook and Bindoff 1997 En dehors de l aspect strictement scientifique cette r gion comprend les TOM Territoires d Outre Mer Fran ais et de nombreux Etats Insulaires et Territoires membres du Programme R gional Oc anien de Environnement PROE d sireux de conna tre leur e
17. changes not experienced in recent times In particular the strong modulation of the SST annual cycle would indicate that the depth of the thermocline and the associated zonal geostrophic circulation fluctuated more extensively than to day on an interannual basis The cause for these large fluctua tions is unclear but although the record is only 47 years long a visual inspection of Figures 3b and 3d clearly points to a decadal scale variability The three major cooling events at 4166415 B P 4149415 B P and 4133415 B P are 17 and 16 years apart respectively The occurrence of a 14 17 year periodicity in Pacific coral oxygen isotopic records of the last centuries is now well established Dunbar et al 1994 Linsley et al 1994 Quinn et al 1996 On land it has been identified in laminated sediments from an Ecuador ian lake Rodbell et al 1999 where it shows maximum spectral density in the last 1000 years and between 3000 and 4000 B P and in a global surface temperature data set Ghil and Vautard 1991 However reliable modern instru mental SST records are not long enough to fully document the interdecadal mode and its spatial distribution in the Pa cific Ocean Despite this limitation several authors have proposed that the interdecadal mode could significantly modulate the ENSO cycle Holbrook and Bindoff 1997 Latif et al 1997 Zhang et al 1997 Gu and Philander 1997 Weaver 1999 A recent simulation Weav
18. du FIDES en 2000 2001 Il est programm pour mesurer et enregistrer la temp rature et la salinit de surface toutes les 30 minutes La cellule de conductivit permettant de mesurer la salinit est v rifi e tous les 6 mois environ et l appareil est envoy chez le fabricant SeaBird USA tous les ans pour talonnage des capteurs Des pr l vements manuels d chantillon d eau de mer sont effectu s lors de l op ration de nettoyage des capteurs Les mesures de temp rature et salinit de surface n ont pas encore t exploit es scientifiquement en raison de la longueur encore trop faible des s ries temporelles Les donn es r colt es sont en cours de validation one N E th ei Ray p 3 BANS Be he Sa Us LES Figure 1 Photographie du thermosalinographe sur son b ti Wallis 2 2 Mise en place d un thermosalinographe sur un navire de commerce Pour des raisons commerciales le bateau Moana sur lequel il tait pr vu d installer un thermosalinographe voir proposition initiale n a plus effectu de navettes r guli res entre Wallis Fidji et la Nouvelle Cal donie partir de d but 2000 L installation d un thermosalinographe a donc t effectu e sur le M S Kyowa Hibiscus navire effectuant le m me type de trajet avec en plus un aller retour sur le Japon et sur la Polyn sie Frangaise Figure 2 Le sch ma de l installation est d crit en d tail dans Prunier Mignot et al
19. in 1992 The coral was slabbed X rayed and cleaned in an ultrasonic bath prior to sampling On average 12 samples per year were taken con tinuously along the main growth axis using a dental burr and a three axis positioning system The aragonitic powder was then dissolved in 2 spiked nitric acid and Ca Sr and U were analyzed with a Varian Ultramass inductively coupled plasma mass spectrometer ICP MS following a new tech nique Le Cornec and Corr ge 1997 A coral sample from New Caledonia labeled NC20 was reduced to powder sieved with a 40 um mesh and used as a standard Replicate analyses of NC20 yield an external reproducibility 20 of 0 05 mmol mol for Sr Ca 0 7 C and of 0 01 mol mol for U Ca 0 3 C Sr Ca and U Ca analyses were fitted to the in strumental SST record from Am d e Lighthouse for the 1981 1990 period Figures 1 and 2 using the Analyseries program Paillard et al 19961 Each tracer was then regressed against SST and the regression yielded the following equa tions Sr Ca 10 M 10 73 0 0657SST R 0 79 U Ca 10 M 2 106 0 0367SST R 0 89 The precision standard error of estimate Runyon et al 1996 Bevington and Robinson 1992 calculated from these equation is 1 3 C for the Sr Ca thermometer and 0 9 C for the U Ca thermometer This is slightly higher than the preci sion of the Sr Ca thermometer using thermal ionization mass spectrometry which has been estimated to be 0 3 C
20. 1990s thus reflect enhanced displacements of the warm pool 16 In the equatorial band the low frequency displacements of the SSS front were shown to be mainly due to zonal advection by oceanic currents and to a lesser extent to the Evaporation Precipitation budget since the convection zone followed the migration of the warm fresh pool Picaut et al 1996 Delcroix and Picaut 1998 In the SPCZ region the displacement of the SSS front still correlates closely with the SOJ and it moves both zonally and meridionally along a northwest southeast oriented axis Surface geostrophic current anomalies reveal that in 1987 88 and 1998 99 westward and slight northward current anomalies develop for 6 months during the El Ni o event almost in phase with the north westward displacement of the high salinity waters These current anomalies are then reversed and give Way to eastward current anomalies in phase with the south eastward displacement of the fresh waters Remarkably when the SSS front along the equator shifts eastward during an El Ni o event bringing warm and fresh waters to the central Pacific the SSS front in the SWTP moves in the opposite direction toward the north west bringing salty and cold waters to the south western tropical Pacific With the data available it would be unrealistic to try to quantify the relative importance of advection versus evaporation precipitation budget in the displacement of the SSS front Firstly because a
21. 40 W Figure 3 Mean monthly number of sea surface salinity measurements in the South Western Tropical Pacific in 2 Latitude by 10 Longitude cells from 1976 to 2000 a Mean sea surface salinity 1976 2000 10 S NE i A Solomon 34 7 _ 160 E 170 E 480 W 170 W 160 W 150 W 140 W Figure 4 Mean a sea surface salinity b sea surface temperature in C c precipitation in m month Precipitation is averaged between 1979 and 2000 the other fields are averaged between 1976 and 2000 36 3 35 75 2 35 5 1 D 4 35250 sr NER PA EL Wl aAa A PAY EE ee AY o gt 3 35 1 34 75 RC D 2 345 i 3 61 76 01 78 01 80 01 82 01 84 01 86 01 88 01 90 01 92 01 94 01 96 01 98 01 00 Figure 5 Monthly time series of sea surface salinity solid line averaged in the South Western Tropical Pacific and SOI curve dashed line Both time series are low pass filtered with a 3 month Hanning filter Note that the SOI axes is reversed a EOF 1 30 7 Sea surface salinity Seasonal time function tPolynesia 160 E 170 E 180 W 170 W 160 W 150 W 140 W c EOF 1 58 3 Sea surface salinity Interannual time function 01 76 01 78 01 80 01 82 01 84 01 86 01 88 01 90 01 92 01 94 01 96 01 98 01 00 d EOF 1 58 3 Sea surface salinity Spatial pattern pars i 160 E 170 E 180 W 170 W 160 W 150 W 140 W Figure 6 First mode of the empirical orthogonal function in se
22. 5 calendar years B P described previously Beck et al 1997 When passed through a 25 month Hanning filter the fossil record yields interannual SST anomalies Figure 3d which are 2 3 times greater than seen in the modern pe riod The two records on Figure 3d are relatively well corre lated suggesting a common mechanism for these variations Spectral analysis of the fossil SST record reveal periods in the 2 4 and 5 5 6 year bands which are essentially the same as the dominant modern ENSO peaks 5 Discussion and Conclusions Comparison of the modern and fossil records emphasizes the stronger interannual variability which existed at circa 4150 calendar years B P Interestingly this variability oc curred at a time when the overall climate is thought to be very similar to the present day one same mean SST similar solar radiation and no ice volume effect It is therefore im portant to determine whether the large interannual cooling events are caused by purely climatic ie the ocean atmosphere couple or external factors One possible external cause for cool SST are volcanic eruptions Bradley 1988 The southwest Pacific and Vanuatu in particular are tectonically and volcanically active places Pelletier et al 1998 and it has been shown recently Crowley et al 1997 that coral derived SST is a good recorder of volcanism How ever cooling events of the magnitude seen in our paleorecord at 4166 4149 4133 and 4128 calendar year
23. 6 Jan84 26 Jan82 25 Jan82 25 Jan80 Jan80 Jan78 Janza J 2A Jan76 23 Jan76 r j 7 23 170 E 180 E 170 W 160 W 150 W 12 S 14 S 16 S 18 S 20 S 22S e Rainfall m month 17 S f Rainfall m month 175 E Jan00 Jan98 Jan96 03 Jang4 0 3 Jan92 0 25 Jango 0 25 0 2 Jan88 0 2 Jan86 0 15 0 15 Jan84 9 1 Jan82 0 1 0 05 sanso 0 05 Jan78 Jan76 r RARE i i 0 Jan76 Tea oe r 7 r 0 170 E 180 E 170 W 160 W 150 W 12 S 14 S 16 S 18 S 20 S 22 S Plate Time longitude plots along 17 S of the a sea surface salinity c sea surface temperature in C and e precipitation in m month Time latitude plots along 175 E of the b sea surface salinity d sea surface temperature in C and f rainfall in m month Time series has been low pass filtered with a 25 month Hanning filter Superimposed on every diagram is the 25 month filtered SOI Negative positive values of the SOI are towards the left right of every figure The scale of the SOI on the x axis is comprised between 1 5 and 2 5 with the zero close to the 165 W longitude for plots a c e and close to the 18 5 S latitude for plots b d f Jango Jan89 Jan88 a Sea surface salinity 17 S 170 E 180 E 170 W 160 W c Sea surface salinity 17 S ae 170 E 180 E 170 W 160 W 150 W 150 W b Zonal geostrophic current anomaly cm s 17 S Jan01 EP Jan00 Jan99 20 Jan
24. 98 Jan97 A 10 Jan96 Jans Jang4 1 10 Jang3 lt 170E 180 W 170 W 160 W 150 W d Zonal geostrophic current anomaly cm s 17 S Jan90 r j E 20 Jan89 Jan88 SE ea Jang7 ee ene Jan86 170 E 180 W 170 W 160 W 150 W Plate 2 Time longitude plots along 17 S of the a 3 month Hanning filtered sea surface salinity between 1992 and 2000 b anomaly referring to the period 1993 1995 of the zonal component of surface geostrophic current in cms computed from TOPEX Poseidon altimeter data c 3 month Hanning filtered sea surface salinity between 1986 and 1989 d anomaly referring to the period 1986 1989 of the zonal component of surface geostrophic current in cms computed from GEOSAT altimeter data PALEOCEANOGRAPHY VOL 15 NO 4 PAGES 465 470 AUGUST 2000 Evidence for stronger EI Nifio Southern Oscillation ENSO events in a mid Holocene massive coral Thierry Corr ge Thierry Delcroix Jacques R cy Warren Beck Guy Cabioch and Florence Le Cornec Abstract We present a 47 year long record of sea surface temperature SST derived from Sr Ca and U Ca analysis of a massive Porites coral which grew at 4150 calendar years before present B P in Vanuatu southwest tropical Pacific Ocean Mean SST is similar in both the modern instrumental record and paleorecord and both exhibit El Ni o Southern Oscillation ENSO frequency SST oscillations How
25. 98 highlighted the dynamical and biogeochimical importance of the Pacific equatorial frontal zone 175 W which separates the warm fresh pool to the west from the colder and saltier waters to the east Figure 1 This front is mainly defined by a zonal gradient of Sea Surface Salinity SSS whereas Sea Surface Temperature SST gradually increasing to 28 5 C from east to west is relatively constant Figure 1 These studies particularly stressed the front s relationship with El Ni o Southern Oscillation ENSO variability demonstrating that its zonal displacements at an interannual time scale are mainly due to advection processes rather than to the Evaporation Precipitation budget The climatic importance of the equatorial salinity front led us to look more closely at a second frontal region located south of the equator 170 W 15 S at the southeastern side of the warm pool Figure 1 This secondary salinity front is the result of the juxtaposition of the high salinity waters formed in the subtropical region 20 S 120 W where evaporation exceeds precipitation and of the low salinity waters of the warm pool area where precipitation exceeds evaporation This frontal region is located under the South Pacific Convergence Zone SPCZ which plays a significant role in global atmospheric circulation The SPCZ extends southeastward from Papua New Guinea 5 S 145 E to about 30 S 120 W over southeastward warm SST low SSS in its western par
26. Alibert and McCulloch 1997 The same procedures were used for the sampling and analy ses of the fossil coral trom Tasmaloum Vanuatu This coral was found between 0 80 and 1 20 m depth in drill core 9A 466 f CORR GE ET AL STRONGER ENSO EVENTS IN A MID HOLOCENE MASSIVE CORAL 27 26 25 N A 3 LSS Sr ca 10 molar amp E e e g Sr Ca 10 73 0 0657 SST n 85 R 0 79 77 19 20 21 22 23 24 25 26 27 SST C Figure 1 Calibration of the Sr Ca thermometer in corals a time series of SST diamonds and dotted line from Am d e Lighthouse New Caledonia and Sr Ca dots and solid line from an adjacent Porites coral and b regression of Sr Ca ratios against SST and the resulting equation collected at an altitude of 4 66 m above sea level Cabioch open shallow marine environment for details see Cabioch et et al 1998 Owing to the calculated uplift rate the pa al 1998 Sr Ca and U Ca analyses were converted to SST leodepth of the Porites can be estimated 10 15 mat 4150 using the abovementioned equations In gt 98 of the sam years B P The surrounding coral assemblages indicate an ples analyzed Sr and U reconstructed SST agree well within U Ca 10 molar 90 LSS 1 05 1 1 1 15 1 2 1 25 1 3 1 35 U Ca 2 106 0 0367 SST 1 4 n 81 R 0 89 1 45 U Ca 10 molar 19 20 21 22 23 24 25 26 27 SST C Figure 2 Calibration of the U Ca thermometer in corals
27. Beck et al 1997 we pre sented the progressive warming of the tropical southwest Pacific during the deglaciation based on analyses of corals drilled on Espiritu Santo Vanuatu 15 40S 167 00 E A 5 year section of an otherwise 47 year long Porites colony revealed a large year long cooling event dated at 4166415 calendar years B P U Th date performed by W Beck in the Department of Geology and Geophysics University of Min nesota which remained unexplained Beck et al 1997 This cold snap prompted us to analyze the entire colony in order Institut de Recherche pour le D veloppement Noum a New Caledonia National Science Foundation Arizona Accelerator Mass Spec trometry Facility Department of Physics University of Arizona Tuc son 3 Laboratoire des Formations Superficielles Institut de Recherche pour le D veloppement Bondy France Copyright 2000 by the American Geophysical Union Paper number 1999PA000409 0883 8305 00 1999P A000409 12 00 465 to investigate whether this was a recurrent feature and if so whether strong ENSO events could be invoked as a possible cause for the cooling In the present contribution we first look at the modern instrumental SST record from Vanuatu in order to characterize the ENSO signal in this area and then compare the modern SST and paleo SST records 2 Material and Method We first collected a live Porites coral from Am d e Light house near Noum a New Caledonia
28. Contribution a l tude de la variabilit climatique de l Oc an Pacifique tropical sud ouest Thierry DELCROIX Yves GOURIOU Fra ois MASIA David VARILLON institut de recherche pour le d veloppement entre de Nounied CONVENTIONS SCIENCES DE LA MER OCEANOGRAPHIE PHYSIQUE N 7 2001 j Institut de recherche pour le d veloppement CONVENTIONS SCIENCES DE LA MER OCEANOGRAPHIE PHYSIQUE N 7 2001 Contribution l tude de la variabilit climatique de l Oc an Pacifique tropical sud ouest Thierry DELCROIX Yves GOURIOU Fra ois MASIA David VARILLON Compte rendu de fin d tude d une recherche financ e par le Fond d investissement pour le D veloppement conomique et Social des Territoires d Outre mer FIDES Contrat n 99T06 financement 2488 00 i Institut de recherche pour le d veloppement IRD Noum a 2001 Delcroix T Gouriou Y Masia F Narillon D Contribution l tude de la variabilit climatique de l Oc an Pacifique tropical sud ouest Noum a IRD Decembre 2001 56 p Conventions Sci Mer Oc anogr Phys 7 HYDROCLIMAT EL NINO VARIABILITE SAISONNIAIRE APPAREIL DE MESURE PREVISION CLIMATIQUE ZONE TROPICALE OCEAN PACIFIQUE SUD OUEST Imprim par la Service de Reprographie Centre IRD de Noum a D cembre 2001 1 INTRODUCTION Observer et comprendre le climat mais aussi pouvoir anticiper suffisamment l
29. E lto F W Taylor J R cy F Rougerie P Joannot and C H nin Sea surface temperature from coral skeletal stron tium calcium ratios Science 257 644 647 1992 Beck J W J R cy F W Taylor RL Edwards and G Cabioch Abrupt changes in early Holo cene tropical sea durface temperature derived from coral record Nature 385 105 707 1997 Berger A Long tenn variations of daily insolation and Quatemary climatic change J Atmos Sci 35 2362 2367 1978 Bevington P R and D K Robinson Data Reduc tion and Error Analysis for the Physical Sci ences McGraw Hill New York 1992 Blackman R B and J W Tukey The Measure ment of Power Spectra Dover Mineola N Y 1958 Bradley R S The explosive volcanic eruption sig nal in Nonhem Hemisphere continental tempera ture records Clim Change 12 221 243 1988 Cabioch G F W Taylor J R cy RL Edwards S C Gray G Faure G S Burr and T Corr ge Environmental and tectonic influence on growth and internal structure of a fringing reef at Tasina loum SW Espiritu Santo New Hebrides island arc SW Pacific Spec Publ Int Assoc Sedimen tol 25 261 277 1998 Crowley T J T M Quinn F W Taylor C H nin and P Joannot Evidence for a volcanic cooling signal in a 335 year coral record from New Cale donia Paleoceanography 12 633 639 1997 Delcroix T Observed surface oceanic and atmos pheric variability in the tropical Pacific at
30. ENTS IN A MID HOLOCENE MASSIVE CORAL 469 that maximum SST occurred on average in March and mini mum SST occurred in September as it does today Figure 4b However the maximum of insolation at 16 S happened in November December at 4150 calendar years B P compared to December January today Berger 1978 and there is a possibility that the seasonal SST cycle was shifted month backward In spite of this probable small phase shift we have elected to present the annual harmonic cycle in the paleo SST record as if it has the same phase as the modern record Figure 3c Over the 47 year long period shown in Figure 3b the mean SST is 27 6 C with a standard deviation of 1 2 C and an annual harmonic amplitude of 0 8 C Figure 3c This mean temperature is very similar to the modern 47 year re gional average 27 7 C Interestingly the annual harmonic amplitude is somewhat smaller than the modern record 1 3 C Parts of the paleorecord exhibit a seasonal range significantly larger than in the modern record while in other portions the range is considerably smaller A 24 month run ning annual amplitude Figure 3d clearly highlights the strong modulation of the annual cycle through time This am plitude modulation explains the relatively large standard de viation observed in the paleorecord Figure 3c This large interannual variability is highlighted in the paleorecord by several long lasting cooling events similar to the one at 416641
31. El Ni o frequency in a climate model forced by future greenhouse warming Nature 398 694 696 1999 Urban F E J E Cole and J T Overpeck Influence of mean climate change on climate variability from a 155 year tropical Pacific coral record Nature 407 989 993 2000 Winter A H Ishioroshi T Watanabe T Oba and J Christy Caribbean sea surface temperatures two to three degrees cooler than present during the Little Ice Age Geophys Res Lett 27 3365 3368 2000 G Cabioch Th Corr ge Th Delcroix and J R cy Institut de Recherche pour le D veloppement BP A5 98848 Noum a New Caledonia e mail correge noumea ird nc F Le Comec Institut de Recherche pour le D veloppement 32 Avenue Varagnat 93143 Bondy cedex France T Quinn College of Marine Science University of South Florida 140 Seventh Avenue South MSL 119 St Petersburg Florida 33701 USA Received March 23 2001 revised July 16 2001 accepted July 17 2001 Centre IRD de Noum a B P A5 Noum a C dex 98848 Nouvelle Cai danie IRD 2001
32. Fig 3 During the warmest periods pre 1720 and post 1735 SST and the SOI are well correlated During the coldest period 1720 1735 there is no correlation between the SST at AL and the DJF SOI This raises the possibility that the western Pacific warm pool cooled below 28 C which constitute an important temperature threshold for organized atmospheric convection In that case the generation of ENSO events could be seriously altered Philander 1990 1999 However interannual SSTA are still present in the coral record during this period and El Nifio events are documented in Peru Ortlieb 2000 If these SSTA are indeed ENSO related then the reliability of the SOI proxy might be altered when backgound conditions get cooler in the western Pacific Alternatively the AL SSTA could have been generated by a phenomenon other than ENSO yet to be determined In addition to the interannual signal the AL SST record contains a strong interdecadal component Fig 1 as opposed to the XXth century instrumental record Fig 2 This is consistent with the findings of Urban et al 2000 who reported stronger decadal cycles in the tropical Pacific during part of the XIXth century when background conditions were cooler The coral SST series from AL is too short to identify unambiguously any spectral peak in the interdecadal band but Quinn et al 1998 did identify a marginally significant 92 94 peak around 14 3 to 15 4 years in their quarterly
33. Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2 1 c 5 os ip 0 bo i 4 gt t 1 15 0 5 i o 1 Q 2 A 1 55 60 65 70 75 80 85 90 95 Figure 4 Modern Sea Surface Temperature SST data a Monthly 1951 1997 SST for a 1 latitude by 1 longitude box cen tered on 16 S and 167 E near Espiritu Santo Vanuatu SW Pacific Ocean b Mean monthly SST solid line plus and minus the associated mean monthly standard deviation dashed lines This graph is plotted with the same scale as Figure 3c for com parison purposes c Comparison of the Southern Oscillation Index SOI dashed line with the monthly SST anomalies SSTA solid line with respect to the 1951 1997 average 27 7 C Both variables are filtered with a 25 month Hanning filter fluence of ENSO on this region we first looked at the monthly 1951 1997 SST averaged over a 1 latitude by 1 longitude box containing Espiritu Santo Reynolds and Smith 1994 Figure 4a The mean SST for that period is 27 7 C with a standard deviation of 1 1 C an annual har monic amplitude of 1 3 C and an associated phase of 67 Figure 4b We applied a 25 month Hanning filter Blackman and Tukey 1958 in order to eliminate signals at periods of 1 year or shorter and clearly highlight interannual variations see Delcroix 1998 Figure 1 The resulting fil tered SST anomalies can then be compared to a filtered South ern Oscillation Index SOI Figure 4c The two signals a
34. N y LU 4 A y f w Ea ANT I FNT UV Lio 24 0 fe pp tp nt pt a pp S O bad 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 Year 33 5 7 5 60 B x a 54 520 z A L FAF las 4 80 7 35 5 1 z 440 E 36 5 di 41 p s ES SC A so oo eo Sa 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 Year Figure 2 Coral 5 O record light line compared to SST A and SSS B dark lines monthly time series from the Fiji area see text The O scale is conventionally reversed and so is the SSS scale for easier reading These are observed time series at zero lag LE BEC ET AL FIJI CORAL 5 O RECORD OF ENSO DRIVEN SSS VARIABILITY 2 5 0 4 Z 15 02462 te JR 0 5 z A 0 SE 3 0 5 S 315 ES RL 2 5 0 4 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 Year 1 0 4 a 0 5 0262 a ws 3 0 0 DE 2 5 2 Z 05 V y 0226 l Api Dapp pc pi 0 4 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 Year 0 4 02 5 A jaz g 8 0 e 2 TS 2 02 ES 0 4 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 Year Figure 3 25 month Hanning filtered time series of O light line SST A SSS B and SOT C dark lines O SST and SSS data are normalized to the 1960 1997 period The light shaded bars represent El Ni o events and the dark shaded ones La Ni a events The scales of SST and SSS are represented in such a way to express their respective effects on O These are observed time series at zero lag that
35. Pazan S E and G Meyers Interannual fluctuations of the tropical Pacific wind field and the Southern Oscillation Mon Wea Rev 110 587 600 1982 Picaut J A Busalacchi M J McPhaden and B Camusat Validation of the geostrophic method for estimating zonal currents at the equator from Geosat altimeter data J Geophys Res 95 3015 3024 1990 Picaut J M Ioualalen C Menkes T Delcroix M J McPhaden Mechanism of the zonal displacements of the Pacific warm pool Implications for ENSO Science 274 1486 1489 1996 Picaut J F Masia and Y du Penhoat An advective reflective conceptual model for the oscillatory nature of ENSO Science 277 663 666 1997 Picaut J M Ioualalen T Delcroix F Masia R Murtugudde and J Vialard The oceanic zone of convergence on the eastern edge of the Pacific warm pool A synthesis of results and 20 implications for El Nifio Southern Oscillation and biogeochimical phenomena J Geophys Res 106 2263 2386 2001 Reverdin G C Frankignoul E Kestenare and M J McPhaden Seasonal variability in the surface currents of the equatorial Pacific J Geophys Res 99 20323 20344 1994 Trenberth KE and T J Hoar The 1990 1995 El Nifio southern oscillation event longest on record Geophys Res Letters 23 57 60 1996 Vialard J and P Delecluse An OGCM study for the TOGA Decade Part I Barrier Layer Formation and Variability J Phys Oceanogr 28 1089 1106 1998
36. SSS is the major factor affecting the interannual 5 O fluctuations Figure 3B The Yasawa coral O records all the salinity anomalies related to major El Ni o and La Ni a events since the early 1960 s Figure 3B C During El Ni o events light shaded bars on Figure 3 the SPCZ shifts equatorward rainfall deficit and large evaporation due to wind reinforcement lead to severe droughts and high SSS in Fiji Delcroix and H nin 1989 Thus El Ni o events are characterized by positive SSS and coral 5 O anomalies During La Ni a events dark shaded bars on Figure 3 the SPCZ migrates southward to Fiji leading to negative SSS and coral O anomalies Zonal advections induced by the displacement of the subtropical gyre center during the ENSO cycle may also contribute to changes in SST and SSS Delcroix and H nin 1989 Corr ge et al 2000 It is worth noting that the SST interannual variability although quite small in amplitude is also linked to the ENSO cycle with a tendency for slightly lower higher SST developing during El Ni o La Ni a events Figure 3A Table 1 summarizes the relationship between the SOI and the interannual anomalies in coral 8O SST SSS rainfall These are rather well correlated to the SOI the usual atmospheric index for basin scale ENSO signal A notable result is the rather good correlation r 0 71 between SSS and 8 O It clearly indicates that 5 O 3899 Table 1 Summary of Correlation Co
37. SWTP Delcroix et al 1996 indicated that the 1974 1989 interannual variations of SSS in the SWTP are closely related to the rainfall regime linked to the displacements of the SPCZ Nevertheless the above cited authors did not entirely rule out the influence of advective processes by current anomalies Since the early analysis of Delcroix and H nin 1989 the VOS program has been maintained and was improved in 1992 when vessels were equipped with thermosalinographs TSG H nin and Grelet 1996 Currently we possess 25 years 1976 2000 of SSS and SST measurements in the SWTP region encompassing 6 El Ni o and 3 La Ni a events The aim of this study is to describe the seasonal and interannual variability of SSS and SST in the SWTP and in particular the above mentioned frontal region In addition some tentative explanations accounting for the observed patterns are investigated using satellite derived precipitation data and surface geostrophic current anomalies computed from altimeter measurements The paper is structured as follows SSS SST precipitation data and their processing and gridding procedure are shown in section 2 Mean field and seasonal cycles of SSS SST and precipitation are briefly described in section 3 to put our analysis in context The ENSO variability of the above parameters is then examined in section 4 using an Empirical Orthogonal Function EOF analysis A conclusion appears in section 5 2 Data and data pr
38. Shimmield C P Chilcott M Jebb A E Fallick and A N Dalgleish Recent changes in climate in the far western equatorial Pacific and their relationship to the Southern Oscillation oxygen isotope records from massive corals Papua New Guinea Earth Planet Sci Let 136 575 590 1995 Vialard J and P Delecluse An OGCM study for the TOGA decade Part Role of salinity in the physics of the western Pacific fresh pool J Phys Oceanogr 28 1071 1088 1998 Wellington G M R B Dunbar and G Merlen Calibration of stable oxygen isotope signatures in Galapagos corals Paleoceanography 11 4 467 480 1996 Xie P and P Arkin Analyses of global monthly precipitation using gauge observations satellite estimates and numerical model predictions J Clim 9 840 858 1996 Yan X H Y He W T Liu Q Zheng and C H Ho Centroid movement of the western Pacific warm pool during the three recent El Ni o Southern Oscillation events J Phys Oceanogr 27 837 845 1997 D Blamart A Juillet Leclerc N Le Bec Laboratoire des Sciences du Climat et de l Environnement CEA CNRS 91198 Gif sur Yvette c dex France e mail Nolwenn Le bec lsce cnrs gif fr T Corr ge T Delcroix Institut de Recherche pour le D veloppement BP AS Noum a New Caledonia Received June 5 2000 accepted October 12 2000 GEOPHYSICAL RESEARCH LETTERS VOL 28 NO 18 PAGES 3477 3480 SEPTEMBER 15 2001 Little Ice A
39. a gt 2 5 C Figure 7b The first EOF on rainfall which represents 51 of the variance 18 for EOF 2 reflects seasonal variations with maximum minimum precipitation during the austral summer winter Figure 8a The maximum amplitude of variation 0 25 m month from peak to peak is situated around 13 S 170 W close to maximum SSS variation Figure 8b and figure 6b Seasonal variability diminishes gradually southward to about 0 06 m month from peak to peak The second EOF of these parameters are not discussed here as we found no obvious related physical mechanisms 4 Interannual variability 4 1 EOF Analysis As shown in the previous section SSS SST and rainfall are influenced by seasonal variability but modulation at longer periods is clearly visible in the original time series Figure 5 We now present the low frequency time series and compare them to the SOI smoothed with a 25 month Hanning filter 4 1 a Sea surface salinity The first EOF on SSS represents 58 of the data variance The time function Figure 6c is well correlated with the SOI Ro 0 70 with no time lag Rmax 0 73 with a two month time lag The spatia pattern Figure 6d indicates that maximum variability is centered near 15 S 175 W along the mean axis of the SPCZ It should be noted that the seasonal and ENSO related signals display very similar spatial patterns Figure 6b and 6d indicating the key role of the SPCZ During El Ni o e
40. a surface salinity in the South Western Tropical Pacific a time function and b spatial pattern of the seasonal time series The seasonal time series is defined as the difference between the monthly time series minus the 25 month filtered time series c time function solid line and d spatial pattern of the interannual time series The interannual time series is computed by filtering the monthly time series with a 25 month Hanning filter Superimposed on the time function is the 25 month filtered SOI dashed line The units are so defined that the product between the spatial pattern and the time function denotes PSU a EOF 1 81 2 Sea surface temp rature Seasonal time function 10 8 6 4 2 0 01 76 01 78 01 80 01 82 01 84 01 86 01 88 01 90 01 92 01 94 01 96 01 98 01 00 b EOF 1 81 2 Sea surface temperature Spatial pattern 160 E 170 E 180 W 170 W 160 W 150 W 140 W c EOF 1 45 1 Sea surface temperature Interannual time function 01 76 01 78 01 80 01 82 01 84 01 86 01 88 01 90 01 92 01 94 01 96 01 98 01 00 d EOF 1 45 1 Sea surface temperature Spatial pattem 10 S p 14 s K y om 0 05 P Man TRI Ar f 18 Sj 7 N are vo Fr nch Poesia 4 2254 OS 160 E 170 E 180 W 170 W 160 W 150 W 140 W Figure 7 Same as Figure 6 for the sea surface temperature The units are so defined that the product between the spatial pattern and the t
41. al O to the regional SST Figure 2A Both records display a marked seasonality The regression of the monthly resolved 8 0 record against SST yields the following equation SO 0 174 40 009 SST 0 032 40 253 r 0 62 1 Since only 38 of the O variance is due to SST the 5 O signal must also be affected by change in the isotopic composition of the water 5 O induced by the precipitation vs evaporation mass balance and or by oceanic advection In Fiji the effects of SST and net freshwater flux are combined high low SST and rainy dry season are concomitant leading to low high coral 5 O values Over the 1960 1998 period the average SSS annual amplitude is very small less than 0 2 However notable interannual variations are clearly visible in Figure 2B To remove the seasonal component and highlight the inter annual variability in SSS 5 O and SST the three records were filtered using a 25 month Hanning filter Blackman and Tukey 1958 Delcroix 1998 Maximum interannual anomalies from peak to peak are 0 77 C 0 96 and 0 60 o respectively for SST SSS and 60 SST isotopic effect on 5 O is relatively weak whereas SSS isotopic effect on Owa is quite substantial Figure 3 indicating coral 32 0 5 60 o A 30 0 f f 20 O t j G in i hon AS A 5 a L i A ADI sap avi A 3 ai w 280 T VV A PY UVA VU VAL VV VU VA 430 3 AVE T
42. al records of salinity are scarce and short in the tropical Pacific and there is a clear need for a reliable salinity proxy to extend our knowledge of ENSO through time Here we present 40 years of O data from a Fiji coral 16 48 S 177 27 E The coral 5 O signal integrates both sea surface temperature SST and sea surface salinity SSS variations On a seasonal timescale 5 O is mainly driven by SST changes whereas on an interannual ENSO timescale it is almost exclusively affected by SSS variability Since interannual fluctuations of SSS are rather well correlated to the Southern Oscillation Index in Fiji coral O can be used to reconstruct paleo salinity data with some level of confidence This may help for tracking ENSO influences back in time Introduction Much useful work to understand and model ENSO has been done neglecting salinity variations see the special TOGA Decade J Geophys Res Oceans volume in June 1998 However it is now well recognized that near surface salinity changes may play a major role in the mixed layer dynamics and thermodynamics of the western Pacific warm pool Lukas and Lindstrom 1991 Vialard and Delecluse 1998 a region of enhanced ENSO related air sea interactions Continuous instrumental records seldom exceed a few decades in the tropics However the understanding of the tropical ocean and atmosphere requires continuous and century long records of key climate variables such as sea surfac
43. and surface current anomalies in the equatorial Pacific during the 1986 1989 El Ni o and La Ni a J Geophys Res 99 25093 25107 1994 Delcroix T C H nin V Porte and P Arkin Precipitation and sea surface salinity in the tropical Pacific Ocean Deep Sea Res 43 1123 1141 1996 19 Delcroix T B Dewitte Y du Penhoat F Masia and J Picaut Equatorial waves and warm pool displacements during the 1992 1998 El Ni o Southern Oscillation events observation and modeling J Geophys Res 105 26045 25062 2000 H nin C and J Grelet A merchant ship thermo salinograph network in the Pacific Ocean Deep Sea Res I 43 1833 1855 1996 Latif M R Kleeman C Eckert Greenhouse warming decadal variability or El Ni o An attempt to understand the anomalous 1990s J Climate 10 2221 2239 1997 Legler D M and J J O Brien Tropical Pacific wind stress analysis for TOGA in JOC Time Series of Ocean Measurements IOC Tech Ser vol 4 UNESCO Paris 1988 Meehl G A The annual cycle and interannual variability in the tropical Pacific and Indian Ocean regions Mon Wea Rev 115 27 50 1987 Morli re A and J P Rebert Rainfall Shortage and El Nifio Southern oscillation in New Caledonia southwestern Pacific Mon Wea Rev 114 1131 1137 1986 Nicet J B and T Delcroix ENSO Related precipitation changes in New Caledonia southwestern tropical Pacific 1969 98 Mon Wea Rev 128 3001 3006 2000
44. ands 13 15 S 180 170 W Oceanologica Acta 22 249 263 1999 Blackman R B and J W Tukey The measurement of power spectra Dover Publications 190 pp 1958 Corr ge T T Delcroix J R cy W Beck G Cabioch and F Le Cornec Evidence for stronger El Nifio Southern Oscillation ENSO events in a mid Holocene massive coral Paleoceanogr 15 465 470 2000 da Silva A A C Young and S Levitus Atlas of Surface Marine Data 1994 Volume 1 Algorithms and Procedures NOAA Atlas NESDIS 6 U S Department of Commerce Washington D C 1994 Delcroix T Observed surface oceanic and atmospheric variability in the tropical Pacific at seasonal and ENSO timescales A tentative overview J Geophys Res 103 18 611 18 633 1998 Delcroix T and C H nin Mechanisms of thermal structure and sea surface thermohaline variabilities in the southwestern tropical Pacific during 1975 1985 J Mar Res 47 777 812 1989 Delcroix T and C H nin Seasonal and interannual variations of sea surface salinity in the tropical Pacific Ocean J Geophys Res 96 22135 22150 1991 Delcroix T and O Lenormand ENSO signals in the vicinity of New Caledonia South Western Pacific Oceanol Acta 20 481 491 1997 Delcroix T and J Picaut Zonal displacement of the western equatorial Pacific fresh pool J Geophys Res 103 1087 1098 1998 Delcroix T J P Boulanger F Masia and C Menkes GEOSAT derived sea level
45. arison of velocity anomalies deduced from the TOPEX Poseidon altimeter data with the current anomalies given by Doppler current measurements made at the equator with the Tropical Atmosphere Ocean TAO array gave good results Delcroix et al 1994 Delcroix et al 2000 Given the absence of direct current measurements away from the equator this type of comparison cannot be made in the SWTP As geostrophic computation is directly related to the invers of the Coriolis parameter computation errors decrease away from the equator We are thus reasonably confident about the anomalies of geostrophic velocities computed in the SWTP The da Silva et al 1994 data base derived from the Comprehensive Ocean Atmosphere Data Set COADS includes one of the longest available evaporation time series It was computed using bulk formulas and ship report data In da Silva et al s atlas this parameter is available on a monthly 7 basis from 1976 to 1993 on a 1 x1 horizontal grid The data were averaged over a spatial 2 x10 grid Finally surface wind stress t will be used to evaluate the magnitude of the Ekman drift ue t pfh Ve tx pfh where fis the Coriolis parameter p the sea water density and A the depth of the Ekman layer In this study we computed the surface wind stress from the Florida State University FSU pseudo windstress Legler and O Brien 1988 available on a monthly basis on a 2 latitude by 2 longitude grid Data were a
46. ations the O record of massive corals can be used as a reliable tool to track salinity anomalies back in time Given the present day ENSO influence on SSS this feature may prove useful for inferring regional past ENSO variability We are now extending the 5 O record of the Fiji coral to the beginning of the century together with Sr Ca measurements as a temperature proxy We will also generate similar records from other ENSO sensitive areas such as Tuvalu and Tokelau Acknowledgments This work was supported by the IRD program Climate Variability and Regional Impacts and Pal oc an Project We thank Jacques R cy leader of the Pal oc an Project and Guy Cabioch leader of the Paleofiji cruise We are most thankful to the officers and crew of the IRD R V Alis Guy Cabioch Yvan Join Jo l Orempiiller and St phanie Reynaud Vaganay provided assistance in coral drilling Jocelyne Bonneau helped for the sample collection The SURTROPAC ECOP group at IRD Noum a was instrumental in collecting in situ SSS data from voluntary observing ships for three decades We express our gratitude to the Government of Fiji for allowing us to work in their economic zone We also thank Michael Evans and an anonymous reviewer for their relevant comments This is LSCE contribution number 0514 References Blackman D S and J W Tukey The measurement of Power Spectra 190pp Dover Mineola N Y 1958 3900 Boiseau M and A Juillet Leclerc H
47. bility All parameters show an interannual signal which correlates well with the SOI The South Western Tropical Pacific Ocean is saltier and colder during El Ni o than during La Ni a events In the southwestern part there is a shortage excess in precipitation during El Ni o La Ni a events The greatest anomalies appeared during the last La Nifia in 1996 and 1999 as regards SST and in 1999 and 2000 as regards SSS SST and precipitation ENSO related anomalies are an order of magnitude smaller than seasonal anomalies while the SSS ENSO related signal is twice as strong as the seasonal signal These facts reflect the northeastward southwestward shift of the SPCZ during El Nifio La Nifia events While consistent with precipitation changes the ENSO related variability in SSS can also be partly explained by the displacement of the salinity front that separates fresh warm pool waters from salty subtropical waters Computation of surface geostrophic current anomalies from GEOSAT 1987 1988 and TOPEX Poseidon 1993 2000 indicates that westward current anomalies developed during the 1987 88 and 1997 98 El Ni o and are linked to the displacement of the salinity front The South Western Tropical Pacific salinity front moves westward eastward in contrast to the equatorial salinity front which moves eastward westward during an El Ni o La Nifia event 1 Introduction Recent studies Picaut et al 1996 1997 2001 Vialard and Delecluse 19
48. cal version 1 2 Voir aussi http www ird nc ECOP En ad quation avec les objectifs scientifiques du programme international CLIVAR et avec un financement MERCATOR compl mentaire les mesures r colt es grace a cette nouvelle installation ont aussi t mises disposition de la communaut scientifique attach e l tude de la variabilit climatique dans le Pacifique tropical Un CD ROM rassemblant l ensemble des mesures de salinit de surface effectu es depuis 1969 dans le Pacifique tropical par l IRD et par d autres organismes IFREMER NOAA JAMSTEC etc a donc t r alis Figure 3 et diffus plus de 200 scientifiques de part le monde Ce CD ROM a t mis en acc s libre en septembre 2000 via le site Web de notre laboratoire http www ird nc ECOP Le site est r actualis tous les 6 12 mois avec les donn es en temps diff r qui continuent tre r colt es et valid es Les proc dures de validation et les codes de qualit attribu s aux donn es sont d crits dans le CD ROM 2 4 Valorisation scientifique et publications Quatre articles de rang A utilisant les donn es nouvellement r colt es ont t publi s ou sous presse en 2000 2001 Il s agit e Dune tude sp cifique sur la variabilit saisonni re et li e ENSO dans le Pacifique tropical sud ouest sur la p riode 1976 2000 Gouriou et Delcroix 2001 e D une tude sur la variabilit de la salinit de surface Fid
49. croix T et O Lenormand 1997 ENSO signals in the vicinity of New Caledonia south western Pacific Oceano Acta 20 481 491 Gouriou Y et T Delcroix 2001 Seasonal and ENSO variations of sea surface salinity and temperature in the South Pacific Convergence Zone during 1976 2000 J Geophys Res in press Holbrook N J and N L Bindoff 1997 Temperature variability in the Southwest Pacific Ocean between 1955 and 1988 J Climate 10 31035 1049 Le Bec N A Juillet Leclerc T Correge D Blamart and T Delcroix 2000 A coral 5880 record of ENSO driven sea surface salinity variability in Fiji South Western Tropical Pacific Geophys Res Letter 27 3897 3900 Nicet J B et T Delcroix 2000 ENSO related precipitation changes in New Caledonia South Western Tropical Pacific 1969 1998 Mon Wea Rev 128 3001 3006 Prunier Mignot M D Varillon L Foucher J M Ihily B Buisson F Masia C H nin M loualalen and T Delcroix 1999 Manuel d installation et de maintenance d un thermosalinographe embarqu Users guide for thermosalinograph installation and maintenance aboard a ship Notes techniques Sciences de la Mer Oc anogr Phys Centre ORSTOM de Noum a 13 102 pages 10 van Loon H and J Shea 1985 The Southern Oscillation IV The precursors south of 15 S to the extremes of the Oscillation Mon Weather Rev 113 2063 2074 Vincent D G 1994 The South Pacific Convergence Zone SPCZ a review
50. d reconstruction of intertropical LE BEC ET AL FIJI CORAL O RECORD OF ENSO DRIVEN SSS VARIABILITY convergence zone variability over Central America since 1707 J Geophys Res 99 C5 9977 9994 1994 Lukas R and E Lindstrom The mixed layer of the western equatorial Pacific Ocean J Geophys Res 96 3343 3358 1991 Madec G P Delecluse M Imbard and C L vy OPA 8 ocean general circulation model reference manual Notes du P le de mod lisation PSL 1998 PDF version available at http www ipsl jussieu fr modelisation notes OPA8 1 Total pdf McConnaughey T C 13 and O 18 isotopic disequilibria in biological carbonates I Patterns Geochim and Cosmochim Acta 53 163 171 1989 Paillard D L Labeyrie and P Yiou Macintosh program performs time series analysis Eos Trans AGU 77 379 1996 An electronic supplement of this reference is available at http www agu org eos_elec 96097e html Picaut J M loualalen C Menkes T Delcroix and M J McPhaden Mechanism of the zonal displacements of the Pacific warm pool implications for ENSO Science 274 1486 1489 1996 Quinn T M T J Crowley and F W Taylor New stable isotope results from a 173 year coral from Espiritu Santo Vanuatu Geophys Res Lett 23 23 3413 3416 1996 Reynolds D and T Smith Improved global sea surface temperature analyses using optimum interpolation J Clim 7 929 948 1994 Tudhope A W G B
51. dures described elsewhere Le Cornec and Corr ge 1997 Corr ge et al 2000 Both Sr Ca and U Ca ratios are well correlated to modern day instrumental SST at AL Corr ge et al 2000 and can be used with confidence to infer past SST For each pseudo monthly sample Sr Ca and U Ca gave paleo SST estimates that were averaged and extrapolated to derive a composite monthly SST record Fig 1 These monthly SST were then passed through a 25 month and a 169 month Hanning filter Blackman and Tukey 1958 respectively to highlight interannual and interdecadal SST variability Oxygen and carbon isotopes were measured previously at a resolution of 4 samples per year on the whole colony Quinn et al 1998 which started to grow circa 1657 A D Fig 2 We chose to perform high resolution analyses 12 samples per year of trace elements on the first 60 years of the eighteenth century because this period encompasses the coldest years of the last 350 years according to oxygen isotopes Fig 2 The interest of trace elements is that whereas the oxygen isotopic record of corals is integrative of both temperature and isotopic composition and hence salinity of surface water Gagan et al 1998 trace elements can provide an almost pure temperature signal Beck et al 1992 Alibert and McCulloch 1997 The combined use of Sr and U which are incorporated in corals at different ratios with respect to temperature is a guarantee of the robustness of th
52. e SST reconstruction Results and Discussion At present ENSO events induce a signature in the vicinity of 1997 which consists in colder than average SST during the warm phase i e El Ni o of ENSO and in warmer than average SST during the cold phase i e La Nifia Correlation of the SST anomaly SSTA at AL with the Southern Oscillation Index SOI is about 0 6 and is maximum when the SOI precedes the SSTA by three months Delcroix and Lenormand 1997 For the last 30 years the mean SST at AL is 23 4 C based on daily bucket measurements with a maximum peak to trough i e La Ni a to El Ni o amplitude of interannual SST anomaly reaching 1 5 C The original oxygen isotopic data indicates a gradual warming from the XVIIIth century to the present see Table 1 and Fig 2 However cooler SST are often linked to drier conditions in New Caledonia yielding to positive 8 O anomalies 3477 3478 8 9 1 1 AAD 1 2 9 1 IR I TE BT DRE 9 2 At THY 1 3 Sr Ca 10 molar erw OL BO N a 10 fl t J Al y 9 4 i TERTI iW I 1 4 9 5 i 1 5 9 6 9 7 1 6 28 26 24 22 AT UNN SST 20 TN 18 N c _ SSTA 1 QU he he ke Tea 1700 1710 1720 1730 1740 1750 1760 year Figurel Reconstructed Sea Surface Temperature SST at Am d e Lighthouse between 1701 and 1761 AD a Raw times series of Sr Ca solid line and U Ca dashed line analyses b Composite monthly SST
53. e temperature SST sea surface salinity SSS and rainfall In this context massive reef building corals are Copyright 2000 by the American Geophysical Union Paper number 2000GL01 1843 0094 8276 00 2000GL01 1843 05 00 increasingly used as natural archives to tentatively provide multicentury climate reconstruction The oxygen isotopic composition O of coral aragonite skeletons is a function of both SST and 5 0 Epstein et al 1953 McConnaughey 1989 Coral 5 O has been widely used as a paleothermometer assuming a fairly constant S O value Dunbar et al 1994 Wellington et al 1996 Charles et al 1997 Boiseau et al 1998 Cole et al 2000 Conversely in regions where temperature variations are small the O of coral has been used to reconstruct changes in 6 O and by extension rainfall variability Cole and Fairbanks 1990 Linsley et al 1994 Tudhope et al 1995 In sites where the climatic and environmental setting is more complex coral tO can reflect a composite signal Quinn et al 1996 Klein et al 1997 E In this paper we present the calibration of a Fiji south western tropical Pacific ocean 16 48 S 177 27 E coral O time series against the instrumental climatic record and validate the robustness of O as a proxy for salinity Climatic and Oceanic Setting The western Pacific warm pool is characterized by the warmest SST 228 C in the open ocean Deep active atmosph
54. easonal oscillations and on the low frequency variations to focus on the ENSO like oscillations In this procedure the twelve first and twelve last months of the time series cannot be filtered and are therefore excluded from the analysis In the following section we describe the EOF performed on the high frequency variations 3 2 Seasonal variability The first EOF on SSS represents 31 of the data variance 12 for EOF 2 A seasonal cycle is apparent in the SSS time function Figure 6a with low salinities around March and high salinities around September Maximum salinity variability has a diagonally oriented shape from Samoa 14 S 172 W to French Polynesia 16 S 145 W roughly along the mean axis of the SPCZ Figure 6b The amplitude of the seasonal signal from maximum to minimum is about 0 5 in the Samoa region peak to peak variation of about 2x0 25 This amplitude is minimum in the south close to New Caledonia peak to peak variation of 0 1 The first EOF on SST represents 81 of the data variance 3 for EOF 2 The time function variability is clearly dominated by seasonal fluctuations Figure 7a As expected high low temperatures are found during the austral summer winter The spatial pattern presents zonally oriented isolines reflecting low amplitude variations in the north east corner of the study area 1 C from peak to peak and high amplitude variations in the south west close to New Caledoni
55. efficients r Values Between the Coral 6 O Record and the Climatic Records all Data are Filtered with a 25 month Hanning Filter Correlations are Significant at the 95 Confidence Level SOI Rainfall SST SSS Rainfall 0 58 sig SST 0 64 0 75 vs SSS 0 62 0 67 0 47 ox coral O 0 45 0 41 0 41 0 71 denotes that a 3 month lag time was introduced the SOI or rainfall preceding the SSS and coral 5 O signals by 3 months can be used as a proxy for reconstructing salinity variability on an interannual timescale in the studied region We then calculated the relationship linking the Oa and SSS interannual variations using the filtered data SO 0 428 0 020 SSS 19 843 40 722 r 0 71 2 The slope of this line s 0 428 0 020 is close to the slope s 0 384 0 008 of the On vs SSS calibration calculated in the Fiji area using the method developed by Delaygue et al 2000 In this latter study the One iS Simulated using the OPA oceanic general circulation model Madec et al 1998 where the atmospheric fluxes are prescribed by the isotopic version of the NASA GISS atmospheric model Jouzel et al 1987 This result from a model reinforces our observational conclusion that the Yasawa coral O is a reliable tracer of seawater O interannual variations and consequently of salinity To conclude we have shown that in the Fiji area where ENSO driven interannual variations in SST are small compared to SSS vari
56. er 1999 which used an extension of the delayed oscillator model showed that extratropical subduction of cooler water which propagates toward the equator Gu and Philander 1997 could alter ENSO on decadai to interdecadal timescales This model generates SST time series for the eastern Pacific which mirror well our paleorecord with significant changes in the seasonal amplitude through time What we see in the fossil record could then represent phase shifts in the ENSO mode quite similar to those which occurred in the twentieth cen tury Zhang et al 1997 but perhaps with stronger ex changes between the tropics and extratropics Acknowledgments We thank Jocelyne Bonneau Dany Bouttefort Claude Ihily Yvan Join Michel Lardy and Jean Louis Laurent for help during the course of this work We also thank Didier Paillard and Henning Kuhnert for help with the use of the Analyseries program Amy Clement John Chappell Richard Grove and Michael Evans made fruitful comments on an earlier version of this manuscript We thank our two reviewers Christina Gallup and George Philander for their relevant comments This work was supported by IRD formely ORSTOM 470 References Alibert C and M T McCulloch Strontium calcium ratios in modem Porites corals from the Great Bar ner Reef as a proxy for sea surface temperature Calibration of the thermometer and monitoring of ENSO Paleoceanography 12 345 363 1997 Beck J W R L Edwards
57. eric convection is located over the warm pool and over the Intertropical and South Pacific Convergence Zones ITCZ and SPCZ see Figure 1 In these zones precipitation exceeds evaporation leading to the occurrence of a fresh pool SSS lt 35 0 Delcroix et al 1996 The fresh pool is limited by a well marked zonal salinity front located at the eastern edge of the maximal rainfall region in the equatorial band Picaut et al 1996 On a seasonal timescale Yan et al 1997 the movements of the warm pool are essentially meridional The warm pool and the SPCZ move southward and reach Fiji in the austral summer November to April rainy season On an interannual timescale the eastern edge of the warm pool and the salinity front migrate zonally in the equatorial band Picaut et al 1996 in phase with the SOI Southern Oscillation Index During El Ni o events the SPCZ tends to merge with the ITCZ leading to dry conditions in Fiji 3897 3898 Figure 1 Synthetic map of the western tropical Pacific Ocean The mean position of the 28 C surface isotherm indicates the margin of the western Pacific warm pool The mean position of the Intertropical Convergence Zone ITCZ and South Tropical Convergence Zone SPCZ is also shown Material and Methods A massive colony of Porites sp was drilled in Yasawa Island 16 48 S 177 27 E on the western side of the Fiji archipelago in July 1998 during the IRD Institut
58. evel data Wyrtki 1975 1984 It implies a stronger than usual SEC from April to July 1983 concomitant with the northeastward displacement of the SSS front and consequently an increase of SSS in the SWTP 5 Conclusion The analysis of SSS SST and precipitation changes over 25 years 1976 2000 in the SPCZ region 10 S 24 S 160 E 140 W brings out the close relationship between the time variability of these parameters and El Ni o or La Ni a This is consistent with previous studies limited to shorter time periods Delcroix and H nin 1989 1991 Delcroix et al 1996 The present study deals with an unprecedented long time series encompassing six El Ni o 1976 77 1982 83 1987 1991 92 1993 1997 98 and three La Nifia 1988 89 1986 1998 00 events Our analysis confirms that SSS is higher SST and precipitation are lower during El Ni o events than during La Ni a events The amplitude of this interannual signal is an order of magnitude less than the amplitude of the seasonal signal for SST and precipitation whereas it is twice the amplitude of the seasonal signal for SSS It is worth noting that the greatest anomalies of the time series 1976 2000 occurred during the last La Nifia events in 1999 and 2000 for sss and in 1996 and 1999 for SST As all ENSO variations are mainly due to the southeastward northward displacement of the warm pool during La Ni a El Ni o events the greatest changes in SST and SSS amplitude during the
59. ever several strong decadal frequency cooling events and a marked modulation of the seasonal SST cycle with power at both ENSO and decadal frequencies are observed in the paleorecord which are unprecedented in the modem record 1 Introduction In the last two decades we have witnessed a change in the mode of El Nifio Southern Oscillation ENSO with more frequent and stronger El Ni o events Trenberih and Hoar 1996 1997 Trenberth and Hurrell 1994 Whether this modulation of ENSO is a result of superimposition of other natural oceanic cycles or is induced by some external factor such as increasing atmospheric greenhouse gases is still de bated Harrison and Larkin 1997 Holbrook and Bindoff 1997 Latif er al 1997 Guilderson and Schrag 1998 Zhang et al 1998 and is a question of some importance to predictive models of climate change One way to differentiate among the potential causal agents is to examine ENSO be havior from a time period where increasing greenhouse gas ses is not an issue Since instrumental sea surface temperature SST records do not extend that far back in time we have to rely on proxies to generate long high resolution SST data sets It is now well established that some trace elements Sr and U in particular incorporated in the aragonitic skeleton of scleractinian corals provide a robust paleothermometer Beck et al 1992 Min et al 1995 Alibert and McCulloch 1997 In a previous contribution
60. ge sea surface temperature variability in the southwest tropical Pacific Thierry Corr ge Terry Quinn Thierry Delcroix Jacques R cy and Guy Cabioch Abstract We present a 60 year near monthly record of tropical sea surface temperature SST during the Little Ice Age derived from coupled Sr Ca and U Ca analysis of a massive coral from New Caledonia southwest tropical Pacific The record indicates that from 1701 to 1761 surface temperatures were on average 1 4 C cooler than during the past 30 years This cooling was accompanied by strong interannual to interdecadal oscillations that changed the background state Correlations between SST changes and the Southern Oscillation and the Pacific Decadal Oscillation are evolutive and appear to depend on the background state Introduction The climate of the Pacific zone is mostly under the influence of El Nifio Southern Oscillation ENSO events As longer records of ENSO related sea surface temperature SST changes become available and theoretical and modelling studies progress the role of the background state of the ocean atmosphere system is increasingly being acknowledged Philander 1999 Lau and Weng 1999 Understanding how interdecadal and long term variations in the background climate state affect the phasing and strength of ENSO is thus a necessary step to improve climate predictions Timmerman et al 1999 Barnett and Latif 2000 Collins 2000 Here we report
61. gitude and the time latitude plot at the longitude of 175 E Plate 1b 1d 1f as time series are complete there from 11 S to 23 S In these plots data have been low pass filtered with a 25 month Hanning filter to eliminate periods lt 12 months and the filtered SOI is superimposed on these plots Plate la and plate 1b clearly show that the salinity front separating the low salinity water under the SPCZ from the high salinity waters of the south central Pacific oscillates back and forth in the zonal meridonal direction following the SOI The correlation between the sor and the displacement in longitude of the 35 5 isoline is Ro 0 59 Rmax equals also 0 59 with a one and two month lag The correlation with the shift in latitude of the 35 isoline is Ro 0 62 Rmx 0 84 with a five month lag As the displacements occurred in both east west and north south directions they indicate that the front moved in a northeast southwest direction Saline waters gt 35 25 shown in yellow to red in the plots invaded the north west portion of the area during El Ni o events at the end of 1977 1983 1987 1993 and 1998 while low salinity waters moved south eastward during La Ni a periods Confirming the EOF analysis the salinity front stayed close to its mean position in longitude around 170 W during the 1993 1995 prolonged warm events As the mean SST field is quasi zonally oriented Figure 4b it is mainly in the north south direction that we can ob
62. good agreement between the Sr and U recon structed SST A 47 year long composite SST curve was then constructed and monthly SST values were extrapolated us ing the Analyseries program Figure 3 Supporting data for Figure 3b are electronically archived at World Data Center A for Paleoclimatology NOAA NGDC 325 Broadway Boulder co 80303 e mail pa leo mail ngdc noaa gov URL http www ngdce noaa gov paleo 3 Modern Instrumental Record The general surface circulation pattern in the southwest Pacific Ocean can be described as a large scale anticyclonic gyre centered near 15 S see Delcroix and H nin 1989 p 7911 North of this latitude a westward surface geostrophic flow tends to bring cooler water whereas south of 15 S an eastward flow carries warmer water During an El Nifio event the center of the gyre is shifted southward by few degrees of latitude and Vanuatu is then affected by stronger than aver age westward flow Wyrtki and Wenzel 1984 Delcroix and H nin 1989 Vanuatu is located on the southwestern fringe of the oceanic domain notably affected by SST changes asso ciated with ENSO Delcroix 1998 To assess the exact in 468 CORREGE ET AL STRONGER ENSO EVENTS IN A MID HOLOCENE MASSIVE CORAL 30 30 a _ 29 29 oO o 28 28 He B 27 27 26 26 25 HI IE 25 55 60 65 70 75 80 85 90 95 30 29 28 27 26 reel Le L 26 25 4 1 r R 7 T 25
63. h 2 and 3 month lag Figure 8c The spatial pattern Figure 8d indicates that rainfall interannual variability has a converse signal on the southwestern and the northeastern sides of the SWTP There is a precipitation shortage excess over Samoa Fiji Vanuatu Solomon and New Caledonia during El Nifio La Nifia events while there is a slight precipitation excess shortage over French Polynesia This pattern is in line with the global analysis made by Meehl 1987 indicating that during strong monsoon events over India corresponding roughly to La Nifia events rainfall increased to the west of the SPCZ axis and decreased to the east of the SPCZ axis This is also in agreement with regional analysis made for New Caledonia from a set of rain gauge stations Morli re and Rebert 1986 Nicet and Delcroix 2000 and for Samoa Wallis Alory and Delcroix 1999 The amplitude of variation due to low frequency modulation is less pronounced than that of seasonal variation For example the precipitation shortage due to the 1982 83 El Ni o in Fiji 18 S 178 W is about 0 04 m month and the precipitation excess during the 1988 89 La Ni a event in the same region is 0 06 m month This can be contrasted with the amplitude from minimum to maximum of the seasonal signal which is of the order of 0 2 m month In New Caledonia 22 S 165 E where seasonal variations are amongst the lowest in the SWTP Figure 8b and interannual variability is high Figure
64. ien et la mise en place d observatoires du milieu oc anique sont n cessaires pour comprendre et aboutir terme une pr vision op rationnelle d ENSO en particulier au voisinage des TOM du Pacifique directement affect s par le ph nom ne Dans ce cadre la proposition financ e par le FIDES en 2000 2001 visait am liorer l observation et la compr hension de la variabilit climatique l chelle r gionale en priorit l chelle d ENSO mais galement d autres chelles de temps Quatre types de travaux ont t effectu s e Le maintien d un thermosalinographe mesurant la temp rature et la salinit dans Je lagon de l le de Wallis e la mise en place d un thermosalinographe sur un navire de commerce faisant une route r guli re dans le Pacifique tropical sud ouest avec transmission des donn es en temps r el e la mise disposition via Internet des donn es r colt es e la valorisation scientifique et la publication de r sultats li s l exploitation des donn es r colt es Ces quatre types de travaux sont d crits tour tour 2 TRAVAUX EFFECTUES 2 1 Mise en place d un thermosalinographe dans le lagon de l le de Wallis Un thermosalinographe de type Seabird SBE 21 a t install pr s de la passe de Futumanimi 13 13 334 176 15 094 E sur l le de Wallis le 22 08 1998 par 9 m de profondeur Figure 1 Cet appareil aurait t enlev sans le soutien financier
65. im 7 929 948 1994 Rodbell D T G O Seltzer D M Anderson M B Abbott D B Enfield and J H Newman An CORREGE ET AL STRONGER ENSO EVENTS IN A MID HOLOCENE MASSIVE CORAL 15 000 year record of El Ni o driven alluviation in southwestern Ecuador Science 283 516 520 1999 Runyon R P A Haber DJ Pitenger and K A Coleman Fundamentals of Behavioral Statistics McGraw Hill New York 1996 Stoll H M and D P Schrag Effects of Quaternary sea level cycles on strontium in seawater Geochim Cosmochim Acta 62 7 1107 1118 1998 Trenberth K E and T J Hoar The 1990 1995 El Ni o Southern Oscillation event Longest on re cord Geophys Res Lett 23 57 60 1996 Trenberth K E and T J Hoar El Ni o and climate change Geophys Res Lett 24 3057 3060 1997 Trenberth K E and J W Hurrell Decadal atmos phere ocean variations in the Pacific Clim Dyn 9 303 319 1994 Weaver A J Extratropical subduction and decadal modulation of El Ni o Geophys Rex Let 26 743 746 1999 Wyrtki K and J Wenzel Possible gyre uyre inter action in the Pacific Ocean Nature 309 538 540 1984 Zhang R H L M Rothstein and A J Busalacchi Origin of upper ocean wanning and El Ni o change on decadal scales in the tropical Pacific Ocean Nature 391 879 883 1998 Zhang Y JM Wallace and D S Battisu ENSO like interdecadal variability 1900 93 J Clim 10 1004 1020 1997 Zielins
66. ime function denotes C a EOF 1 51 0 Rainfall Seasonal time function 4 1 01 76 01 78 01 80 01 82 01 84 01 86 01 88 01 90 01 92 01 94 01 96 01 98 01 00 b EOF 1 51 0 Rainfall Spatial pattem 10 S p Solom 148 18 S 22 S D o aled 160 E 170E 180 W 170 W 160 W 150 W 140 W c EOF 1 52 5 Rainfall Interannual time function 0 3 3 0 2 1 DD dd ores MER IEA Poho ee ern een 2 OA be ee ee es Nob pe ONE FA MS eee RS D nee ann Lire ae N l 0 aPC E E E E eet es A RC FES Sete Phe sles N Ba 7 SNS 1 0 z E EEE aad seit wth S T Neds a N oe rice D ER eee ed ee 10 160 E 170 E 180 W 170 W 160 W 150 W 140 W Figure 8 Same as Figure 6 for the precipitation The units are so defined that the product between the spatial pattern and the time function denotes m month a Sea surface salinity 17 S b Sea surface Salinity 175 E Jan00 Jan00 Jan98 Jan98 Jan96 Jan96 Jang4 Jan94 Jan92 Jan92 Jango Jan90 i Jangs f 2 Jan88 Jan86 Jan86 Jan84 Jan84 Jan82 Jan82 Jan80 Jan80 Jan78 Jan78 Jan76 Jan76 170 E 180E 170 W 160 W 150 W 12 S 14 S 16 S 18 S 20 S 22S c Sea surface temperature C 17 S d Sea surface temperature C 175 E Jan00 Jan00 Jan98 Jan98 Jan96 Jang6 Jan94 1 30 Jan94 30 Jan92 Jan92 Jan90 3 Jan90 oe Jan88 Tas ee Jan88 28 Jang6 ye Jange 27 Jan84 2
67. ji la fois avec des mesures directes et en utilisant des donn es coralliennes Lebec et al 2000 e D une tude sur la variabilit de la temp rature de surface Vanuatu dans laquelle la p riode actuelle est compar e la p node du mid holoc ne Corr ge et al 2000 e D une tude sur la variabilit de la temp rature de surface en Nouvelle Cal donie dans laquelle la p riode actuelle est compar e la p riode du petit age glaciaire Corr ge et al 2001 Ces quatre articles sont reproduits en Annexe REFERENCES Alory G et T Delcroix 1999 Climatic variability in the vicinity of Wallis Futuna and Samoa islands 13 S 15 S 180 170 W Oceano Acta 22 249 263 Correge T T Delcroix J Recy W Beck G Cabioch et F Le Comec 2000 Evidence for stronger El Ni o Southern Oscillation ENSO events in a mid Holocene massive coral Paleoceanography 15 465 470 Correge T T Quinn T Delcroix F Le Cornec J Recy and G Cabioch 2001 Litte Ice Age sea surface temperature variability in the southwest tropical Pacific Geophys Res Let 28 3477 3480 Delcroix T et C H nin 1989 Mechanisms of subsurface thermal structure and sea surface thermohaline variabilities in the southwestern tropical Pacific during 1979 85 J Mar Res 47 777 812 Delcroix T C H nin V Porte et P Arkin 1996 Precipitation and sea surface salinity in the tropical Pacific Deep Sea Res 43 1123 1141 Del
68. ki G A P A Mayewski L D Meeker S Whitlow M S Twickler M Morrison D A Meese AJ Gow and R B Alley Record of vol canisin since 7000 B C from the GISP green land ice core and nnplicanons for the volcano climate system Science 264 948 952 1994 W Beck NSF Arizona AMS Facility Depart ment of Physics University of Arizona Tucson AZ 85721 G Cabioch T Corr ge T Delcroix and J R cy IRD BP A5 Noumea New Caledonia correge noumea ird nc F Le Cornec Laboratoire des Formations Su perficielles IRD 31 Avenue Varagnat 93143 Bondy cedex France received May 18 1999 revised March 13 2000 accepted March 27 2000 GEOPHYSICAL RESEARCH LETTERS VOL 27 NO 23 PAGES 3897 3900 DECEMBER 1 2000 A coral O record of ENSO driven sea surface salinity variability in Fiji south western tropical Pacific Nolwenn Le Bec Anne Juillet Leclerc Laboratoire des Sciences du Climat et de l Environnement CEA CNRS Gif sur Yvette France Thierry Corr ge Institut de Recherche pour le D veloppement Noum a New Caledonia Dominique Blamart Laboratoire des Sciences du Climat et de l Environnement CEA CNRS Gif sur Y vette France Thierry Delcroix Institut de Recherche pour le D veloppement Noum a New Caledonia Abstract The role of salinity in the dynamics and thermodynamics of El Ni o Southern Oscillation ENSO events is increasingly being investigated However instrument
69. l component of pseudo windstress and the SOI not shown This close relationship was to be expected since the SOI is a measure of the variability of the sea level atmospheric pressure gradient between Tahiti and Darwin and so and so is tightly related to the meridional geostrophic component of the wind It is therefore not surprising that the zonal component of the Ekman drift ue t pfh varies in phase with the SOI westward eastward anomalies of Ekman velocities occurred during El Ni o La Nifia events One difficulty in estimating the magnitude of the Ekman velocity comes from our inability to quantitatively scale the Ekman depth A and its low frequency variability Using a realistic 15 h 30 m at 17 S results in a westward current anomaly smaller than 3 cm s at the beginning of 1998 and an eastward anomaly lower than 2 cm s in January 1999 and in January 2000 These currents thus reinforce the zonal component of the geostrophic velocity although they are of a smaller order of magnitude Based on XBT data Delcroix and H nin 1989 showed that during the 1982 83 El Ni o event isotherms rose in the northern part of the SWTP through Ekman pumping in relation with the northward displacement of the SPCZ This rise was accompanied by a southward shift of the center of the large scale anticyclonic gyre reinforcing the westward flowing SEC in the SWTP region This southward shift during the 1982 83 El Ni o event was also evident in sea l
70. limate oscillation with impacts on salmon production Bull Am Meteorol Soc 78 1069 1079 1997 Meyers G H Phillips N Smith and J Sprintall Space and time scales for optimal interpolation of temperature tropical Pacific ocean Progr Oceanogr 28 189 218 1991 Ortlieb L The documentary historical record of El Ni o events in Peru An update of the Quinn record sixtennth through nineteenth centuries in El Ni o and the Southern Oscillation Variability Global and Regional Impacts H Diaz amp V Markgraf Eds Cambridge University Press 207 295 2000 Philander S G El Ni o La Ni a and the Southern Oscillation Academic Press NY 1990 Philander S G A review of tropical ocean atmosphere interactions Tellus 51 A B 71 90 1999 Quinn TM T J Crowley F W Taylor C H nin P Joannot and Y Join A multi century stable isotope record from a New Caledonia coral interannual and decadal sea surface temperature variability in the southwest Pacific since 1657 A D Paleoceanography 13 412 426 1998 Stahle D W R D D Arrigo P J Krusic M K Cleaveland E R Cook R J Allan J E Cole R B Dunbar M D Therrell D A Gay M D Moore M A Stokes B T Burns J Villanueva Diaz and L G Thompson Experimental dendroclimatic reconstruction of the southern oscillation Bull Am Meteor Soc 79 2137 2152 1998 Timmermann A J Oberhuber A Bacher M Esch M Latif and E Roeckner Increased
71. llons d eau de mer se sont av r es totalement irr alistes Ces tentatives doivent tre reconduites 40 N 20 N 0 20 S 40 S 120 E 160 E 160 W Figure 2 Route type en noir du M S Kyowa Hibiscus sur lequel est install un thermosalinographe Les cercles repr sentent les thermosalinographes mouill s Wallis et en Nouvelle Cal donie 2 3 Mise a disposition via Internet des donn es r colt es L installation d crite pr c demment vient en r alit compl ter un r seau de thermosalinographes install s sur des navires marchands par notre laboratoire depuis le d but des ann es 1990 Ce type de mesures existe sur plusieurs lignes de navigation actuellement sur 4 lignes Pacifique ouest et 3 lignes tour du monde L avantage de cette nouvelle installation financ e par le FIDES pour le Pacifique tropical sud ouest incluant les TOM frangais est de permettre un chantillonnage beaucoup plus fin la fois dans le temps et dans l espace de cette r gion Laboratoire d oc anographie Leo du Contre IRO de Noun s Nouvolta Cai donia s CLIMA Qy g Eg D bo AN Bacirique TRI Three Decades of In Situ Sea Surface Salinity Measurements in the Tropical Pacific Ocean 1969 Thiry DELCROIx 1999 Chistian HE A P ane MS BI Ouvid VARILLON Stagg iagnd IRD key 20 Figure 3 Couverture du CD ROM rassemblant les mesures de salinit de surface des ann es 1969 2000 dans le Pacifique tropi
72. ltimetry only enables us to compute the anomalies of geostrophic currents and we do not know the mean current in the period under study Secondly because it is difficult to obtain a reliable evaporation precipitation budget at interannual time scale Thirdly because we do not know the depth of the salinity mixed layer a term which appears in the salt conservation equation We realize that to be conclusive this analysis should be more quantitative This could be achieved with models adequately reproducing our observations as far as we know this has not yet be done for the SPCZ region Acknowledgement The 1976 2000 bucket and TSG data set represents the combined effort of many IRD colleagues involved in the ship of opportunity program and particularly owes much to Luc Foucher Jean Marc Thily and David Varillon for the recent measurements The satellite derived geostrophic current anomalies were processed by Frangois Masia at IRD Noum a We are also indebted to all our colleagues who gave us free access to their data set through the Web We would like to thank Jo l 17 Picaut two anonymous reviewer and the JGR editor and co editor for their constructive comments on the manuscript This work was funded by the Institut de Recherche pour le D veloppement IRD and the Programme National d Etude de la Dynamique du Climat PNEDC References Alory G and T Delcroix Climatic variability in the vicinity of Wallis Futuna and Samoa isl
73. mperature The first EOF on SST represents 45 of the variance 21 for EOF 2 Correlation between the time function and the filtered SOI curve is Ro 0 51 with no lag and Ryy 0 53 with a 2 month lag Figure 7c SST is colder respectively warmer during El Nifio respectively La Nifia events which is consistent with Delcroix and Lenormand s 1997 analysis for the New Caledonia sector 22 S 165 E The amplitude of variation due to inter annual variability 0 75 C maximum in 1996 and 1999 in the Vanuatu region is less pronounced than the amplitude of the seasonal variations Though weak this ENSO signal is still detectable in the paleorecord of coral drilled in Vanuatu Corr ge et al 2000 Maximum interannual variability occurred at the southern edge of the mean position of the warm pool in between Vanuatu and the Fiji islands Figure 7d In spite of a good phasing agreement Figure 7c there is no obvious link between the intensity of the SOI and the amplitude of the SST anomaly Indeed the very strong 1982 83 El Nifio event has a much weaker SST signature than the 11 smaller 1986 87 event The greatest anomalies occurred during the 1996 La Ni a although the SOI is only slightly positive and during the 1999 00 La Ni a 4 1 c Precipitation The first EOF on precipitation represents 53 of the variance 19 for the EOF 2 There is a clear correlation between the time function and the SOI R5 0 80 with no lag Rymax 0 87 wit
74. nd C H nin Mecanisms of subsurface thermal structure and sea surface thermohaline variabilities in the south western tropical Pacific during 1979 1985 J Mar Res 47 777 812 1989 Delcroix T Observed surface oceanic and atmospheric variability in the tropical Pacific at seasonal and ENSO timescales a tentative overview J Geophys Res 103 18611 18633 1998 Delcroix T C H nin V Porte and P Arkin Precipitation and sea surface salinity in the tropical Pacific Ocean Deep Sea Res 1 43 1123 1141 1996 Dunbar R B G M Wellington M W Colgan and P W Glynn Eastern Pacific sea surface t mperature since 1600 A D the 6 O record of climate variability in Galapagos corals Paleoceanography 9 2 291 315 1994 Epstein S R Buchsbaum H Lowenstam and H C Urey Revised carbonate water isotopic temperature scale Bull Geol Soc Am 62 417 425 1953 Jouzel J G L Russel R J Suozzo R D Koster J W C White and W S Broecker Simulations of the HDO and H O atmospheric cycles using the NASA GISS General Circulation Model the seasonal cycle for present day conditions J Geophys Res 92 14739 14760 1987 Klein R A W Tudhope C P Chilcott J P tzold A Z M Fine A E Fallick and Y Loya Evaluating southern Red Sea corals as a proxy record for the Asian monsoon Earth Planet Sci Let 148 381 394 1997 Linsley B K R B Dunbar G M Wellington and D A Mucciarone A coral base
75. nvironnement climatique et ses modifications pass es pr sentes et futures Les deux grands traits caract ristiques de la r gion sont l existence de la Zone de Convergence du Pacifique Sud SPCZ cf Vincent 1994 et du grand tourbillon anticyclonique qui recouvre l ensemble du Pacifique sud La SPCZ est associ e des pr cipitations tr s importantes de 2 6 m an e g Delcroix et al 1996 et le grand tourbillon une circulation oc anique vers l est au sud d environ 15 S et vers l ouest au nord de cette latitude En p riode EI Ni o la SPCZ se d place vers l quateur et l axe du grand tourbillon se d place de quelques centaines de kilom tres vers le sud Wyrtki et Wenzel 1984 Delcroix et H nin 1989 Des d placements inverses sont observ s en p riode La Ni a Les d placements m ridiens de la SPCZ et du grand tourbillon entra nent de fortes anomalies atmosph riques et oc aniques en particulier ils modifient de mani re fondamentale la pluviom trie r gionale la salinit le niveau de la mer et la circulation oc anique avec renverse possible des courants zonaux Nous savons maintenant sans toutefois en comprendre pleinement les m canismes que les anomalies ENSO observ es dans les TOM du Pacifique tropical sud ouest peuvent tre de signes contraires aux anomalies de la bande quatoriale Delcroix et Lenormand 1997 Delcroix 1998 Alory et Delcroix 1999 Nicet et Delcroix 2000 Le maint
76. ocessing 2 1 SSS and SST measurements In situ SSS and SST measurements collected through a VOS program initiated by IRD Institut de Recherche pour le D veloppement formerly ORSTOM in 1969 constitute the main source of data for this study Before 1992 meteorological buckets were employed to collect salinity samples and measure SST along regular shipping routes Bucket measurements were generally made every 30 5 60 nautical miles 50 100 km at about every 6 hours From 1992 these merchant ships were equipped with ThermoSalinoGraphs TSG H nin and Grelet 1996 which provide one measurement every 15 seconds Before being entered in the data base a median filter is applied to TSG time series to give one value every 5 minutes In the present analysis a median filter was applied to the 5 minutes time series to give one value representing one hour measurements H nin and Grelet 1996 deduced the quality of the bucket and TSG measurements through comparison with simultaneous Conductivity Temperature Depth CTD measurements Bucket data are less accurate than TSG data and the bucket minus CTD difference is much more variable than the TSG minus CTD difference Taking CTD data as a reference we added the following values obtained by H nin and Grelet 1996 to the measurements bucket salinity 0 1 bucket temperature 0 15 TSG salinity 0 02 and TSG temperature 0 2 About 115 000 SSS and SST 41 bucket 59 TSG 0 02
77. on a 60 year record of coral derived monthly SST from the south western tropical Pacific during the early eighteenth century This period lies in the heart of the Little Ice Age LIA 1400 1850 A D an era of documented cooler climate in mid latitudes Bradley and Jones 1993 and in the tropical Atlantic Ocean Keigwin 1996 Winter et al 2000 which precedes the industrial revolution and associated increase in greenhouse gases Our aim is twofold 1 Search for a possible cooling of surface waters associated with the LIA in the southwestern tropical Pacific Ocean and 2 If any evaluate how such a cooling affected climate variability at interannual i e ENSO to interdecadal time scale Institut de Recherche pour le D veloppement BP A5 98848 Noum a New Caledonia College of Marine Science University of South Florida 140 Seventh Avenue South MSL 119 St Petersburg Florida 33701 USA Institut de Recherche pour le D veloppement 32 Avenue Varagnat 93143 Bondy cedex France Copyright 2001 by the American Geophysical Union Paper number 2001GL013216 0094 8276 01 2001GL013216 05 00 New Caledonia Delcroix and Lenormand Florence Le Cornec Methods and Data The coral we used is a portion of a large colony of Porites cf lutea collected near Am d e Lighthouse New Caledonia 22 30 S 166 30 E subsequently AL Coral slabs were sampled and trace elements analyses were performed according to proce
78. ow 25 C south of 22 S Seasonal and interannual variability is apparent in the monthly time series of SSS averaged over the SWTP region figure 5 During non ENSO years for example during the 1979 1982 or 8 1984 1986 periods SSS is minimum in March around 35 1 and maximum in September around 35 4 The interannual variability of SSS is clearly linked to El Nifio or La Ni a as a significant divergence from the mean 35 29 is observed following extreme values of the SOI curve SSS increases during El Ni o 1976 77 1982 83 1987 1991 92 1993 1997 98 and decreases during Le Ni a 1988 89 1986 1998 99 00 The greatest divergences occurred during the 1982 83 El Ni o when SSS reached 35 85 and during the 1988 99 00 La Nifias when SSS decreased to 34 75 In order to quantify the relative amplitude of the seasonal and interannual signals and to obtain a synthetic view of seasonal and inter annual variability an EOF analysis is performed on SSS SST and rainfall fields For each parameter the high frequency signal is separated from the interannual one in the following manner we first smooth the original time series with a 25 month Hanning filter Blackman and Tuckey 1958 to get variations at periods gt 12 months then subtract the original time series from the filtered ones to obtain variations at periods lt 12 months The EOF analysis is applied separately to each variable on the high frequency variations to focus on s
79. re positively correlated R 0 80 at zero month lag and indi cate that ENSO can account for 64 of the interannual vari ance of SST During the warm phase of ENSO EI Ni o SST tends to be colder in Vanuatu consistent with stronger than average westward flow sea level drops see Delcroix 1998 plate Sa and so do the thermocline shoals resulting in higher seasonal variability in SST at this latitude During the cold phase of ENSO La Ni a the deeper thermocline results in a weak seasonal SST amplitude Although quite high the correlation between the SOI and the SSTA in Vanuatu clearly highlights the complexity of the coupled atmosphere ocean system and the nonlinear re sponse of SST to changes in the SOI Spectral analysis of the monthly SST time series using the Analyseries program Paillard et al 1996 indicate that significant peaks are present in the 2 4 year and 6 5 7 year bands which are the classical ENSO period Enfield and Cid 1991 Sea surface salinity SSS results not shown anomalies also correlate well with the SOI reinforcing our confidence that Vanuatu is a pertinent area to document ENSO variability through time 4 Mid Holocene Coral Record The mid Holocene SST record from Vanuatu also strongly exhibits ENSO like periodicity This paleo SST record starts in 4175 15 B P and ends in 4128 15 B P Figure 3 The raw paleo SST were resampled at monthly intervals assuming CORREGE ET AL STRONGER ENSO EV
80. record derived from Sr Ca and U Ca analyses In each sample both trace element ratios were converted to SST following the equations presented in Corr ge et al 2000 and then averaged The whole series was then resampled at monthly intervals e Interannual thin line and interdecadal thick line monthly SST anomalies with respect to the 60 year period average SST Anomalies were calculated by applying respectively a 25 month for interannual anomalies and a 169 month for interdecadal anomalies Hanning filter to the monthly SST resulting in reconstructed SST which would be too cold The coral trace element data indeed indicate that the mean SST from 1701 to 1761 was 22 C as opposed to 21 6 C given by the 5 O This represents a cooling of 1 4 C with respect to the last 30 years The paleo record from AL also indicates a strong interdecadal SST modulation with the lowest SST between 1720 and 1740 being 2 C colder than present day values Fig 1 This drop in SST is consistent with the 1 C cooling derived from planktonic foraminifers for the LIA in the Sargasso Sea Keigwin 1996 and with the 2 to 3 C cooling measured in Caribbean corals of similar age Winter et al 2000 CORREGE ET AL LITTLE ICE AGE SEA SURFACE TEMPERATURE VARIABILITY Although part of the cooling described by the long 80 record can be attributed to a change in the isotopic composition of surface water it can be nonetheless concluded that the so
81. rth of 14 S It must be noted that this spatial distribution reflects a difference in temporal distribution While time series are almost complete in the center of the region there are practically no data after 1992 east of 180 W south of 22 S and north 13 S making any extrapolation dubious Therefore the bins east of 180 W south of 22 S and north 13 S were excluded from the analysis 2 2 Additional data Rainfall data are derived from the analysis of Xie and Arkin 1997 They result from the merging of different data sources rain gauges a number of satellite estimates and forecasts from the NCEP National Centers for Environmental Prediction re analysis The data are available monthly on a 2 5 latitude by 2 5 longitude spatial grid from 1979 to 2000 To match the SST SSS grid size we first interpolated the rainfall data every 2 of latitude and then averaged the data over 10 of longitude Two sets of data relative to satellite derived surface current anomalies are available GEOSAT from November 1986 to February 1989 Picaut et al 1990 Delcroix et al 1994 and TOPEX Poseidon from October 1992 to October 2000 Delcroix et al 2000 Anomalies of zonal and meridional geostrophic velocity are computed every 5 days on a 0 5 latitude by 5 longitude grid For GEOSAT the anomalies are computed relatively to the 1986 1989 period while the TOPEX Poseidon anomalies are computed relatively to the 1993 1995 period Comp
82. s B P all dates given 15 years see Figure 3b would require volcanic erup tions so large that their signature would certainly be found elsewhere around the globe for example the Pinatubo erup tion in 1991 caused sea surface temperature anomalies SSTA of only 0 5 C in the Western Pacific Warm Pool Gagan and Chivas 1995 In particular sulfate peaks would be present in the polar ice core records The detailed record of sulfate concentration in the Greenland Ice Sheet Project GISP 2 ice core Zielinski et al 1994 only docu ments one peak at 4157 calendar years B P which could correlate with either the 4166415 or the 4149 15 B P cool ing events The other cold snaps however cannot be ex plained by volcanism By analogy to the modern instrumental record it could then be argued that the fossil SST record documents a succes sion of long lasting La Ni a like to average i e SOI 0 conditions interrrupted by strong El Ni o like events It is likely that during the mid Holocene the southward north ward shift of the large scale anticyclonic gyre center during El Ni o La Ni a resulted in a shoaling deepening of the thermocline and a decrease increase of the annual amplitude in SST as seen today Still the El Ni o of 1982 1983 and 1986 1987 caused SST anomalies filtered data of the order of 0 5 C at Vanuatu Cooling anomalies of 1 C or more like those seen in the fossil record imply large scale oceanic
83. seasonal and ENSO timescales a tentative overview J Geophys Res 103 18 611 18 633 1998 Delcroix T and C H nin Mechanisms of subsur face thermal structure and sea surface thermohaline variabilities in the southwestern tropical Pacific during 1979 85 J Mar Res 47 777 812 1989 Dunbar R B G M Wellington M W Colgan and P W Glynn Eastern Pacific sea surface tempera ture since 1600 A D the 8 O record of climate variability in Galapagos corals Paleoceanogra phy 9 291 315 1994 Enfield D and L Cid Low frequency changes in Ei Ni o Southem Oscillation J Clim 4 1137 1146 199 Gagan M K and A R Chivas Oxygen isotopes in western Australian coral reveal Pinatubo aerosol in the western Pacific Warm Res Lett 22 9 1069 1072 induced cooling Pool Geophys 1995 Ghil M and R Vautard Interdecadal oscillations and the wanning trend in global temperature time series Nature 350 324 327 1991 Gu D and G H Philander Internal climate fluctua tions that depend on exchanges between the trop ics and extratropics Science 275 805 807 1997 Guilderson T P and D P Schrag Abrupt shift in subsurface temperatures in the tropical Pacific as sociated with changes in El Ni o Science 281 240 243 1998 Harrison D E and N K Larkin Darwin sea level pressure 1876 1996 Evidence for climate change Geophys Res Let 24 1779 1782 1997 Holbrook N J and N L
84. serve displacements of surface isotherms This is the case in plate 1d where 13 there is a tendency towards southward respectively northward shifts of the warm pool SST gt 28 C during La Ni a respectively El Ni o events The time evolution of the north south movements of the precipitation field also corresponds closely to the SOI the tongue of maximum precipitation yellow to red in Plate 1f extends southward northward during La Nifia El Nifio events in relation to the SPCZ southwestward northeastward displacements An exception occurred at the end of 1989 and the beginning of 1990 with a precipitation shortage from 11 S to 23 S although there is no minimum in the SOI that might indicate an El Nifio event plate 1f The time longitude precipitation plot does not display any clear front plate Le Precipitation is minimum in the east and in the west of the SWTP In the center of the region precipitation variability is also linked to the SOI Precipitation is minimum in 1983 1987 1992 and 1998 and maximum in 1981 82 1984 85 1989 and from 1993 to 1996 To sum up there are southward incursions of the warm pool in the SWTP during La Ni a events Deep atmospheric convection is closely linked to SST above 28 C as is the position of the SPCZ see also Vincent 19941 It is thus not surprising that rainfall increases in the SWTP during La Nifia events This excess of precipitation is qualitatively consistent with interannual
85. t and maximum precipitation tongues The importance of the SPCZ has been brought to light since the advent of satellite imagery and the use of Outgoing Longwave Radiation OLR data as a proxy for deep tropical convection and global circulation The annual cycle of the SPCZ is thus characterized by high deep convection activity during the austral winter and low deep convection activity during the summer Meehl 1987 Vincent 1994 The SPCZ shifts north and east south and west of its average position during an El Ni o La Nifia event Pazan and Meyers 1982 Vincent 1994 Few previous studies investigated the links between El Ni o and La Ni a and the interannual variability of oceanic and atmospheric parameters in the South Western Tropical Pacific SWTP defined here as the region between 24 S 10 S and 160 E 140 W Figure 1 Delcroix and H nin 1989 analyzing Voluntary Observing Ship VOS data between 1979 and 1985 stressed the influence of the 1982 1983 El Nifio event on SST SSS and subsurface thermal structure variability They particularly noted that SSS increased by 1 in early 1983 mainly in response to a rainfall shortage due to a northward migration of the SPCZ Delcroix and H nin 1991 studying the 1972 1988 SSS variability along 4 main shipping routes found that during the warm phases of ENSO SSS is fresher than average west of 150 W within 8 S 8 N and saltier than average poleward of 8 latitude i e in the
86. uthwestern Pacific in the vicinity of New Caledonia was cooler during part of the LIA The spatial extension of this cooling is difficult to assert but at present low frequency SST changes at any point in the southwest Pacific can be extrapolated with confidence over about 4 6 of latitude and 10 15 of longitude Meyers et al 1991 Coral records from the Great Barrier reef and from Panama also show cooler and or drier conditions prior to 1850 but other corals from the central and eastern south Pacific record no obvious trend see Gagan et al 2000 for a synthesis The coral record from AL also delivers valuable information about interannual ENSO to interdecadal variability in pre industrial times The fact that three of the four coolest episodes on record namely 1720 1728 and 1747 48 see Fig D are rated as strong to very strong El Ni o years in documentary evidence from South America Ortlieb 2000 provides additional proofs of the robustness of our coral record The coldest year on record in the studied period 1744 x D Q a a SOI index 1650 1700 1750 1800 1850 1900 1950 2000 year Figure 2 a Times series of Am d e Lighthouse oxygen isotopes Quinn et al 1998 b regional SST from 1903 1993 GISST2 2 data from the GOSTA plus CD ROM and reconstructed AL SST from 1701 1761 e reconstructed PDO index Biondi et al 2001 d reconstructed northern hemisphere winter DJF SOI index Stahle et al
87. variation in the SSS front position Besides since Delcroix and H nin 1989 have shown that during the 1982 83 El Nifio the South Equatorial Current was stronger than usual zonal advection could also possibly account for the interannual displacement of the SSS front 4 2 6 Potential role of zonal and meridional advection Reverdin et al 1994 estimated mean surface current from buoy drifts and current meter records between January 1987 and April 1992 for the tropical Pacific In the SWTP at 17 S the mean zonal component of the surface current is westward east of 175 E and eastward west of 175 E Thus there is a tendency for zonal convergence of the surface currents to strengthen the SSS front One means of obtaining information on the surface currents at inter annual time scale is to use altimeter data which allow computation of geostrophic current anomalies Two major contrasting periods were sampled by the altimeters the 1987 1989 El Nifio La Ni a with GEOSAT and the 1997 14 2000 EI Nifio La Ni a with TOPEX Poseidon Anomalies of surface geostrophic zonal currents and SSS field are displayed along 17 S in plate 2 for these two periods The velocity anomalies are generally within 5 cms though they exceed these values for a few months in 1987 and in 1998 1999 During the 1987 88 El Ni o a westward anomaly gt 5 cms is present from February 1987 to July 1987 between 170 E and 160 W During the second half of the
88. vents 1976 77 1982 83 1986 87 1991 92 1993 1997 98 the SWTP is saltier than during La Ni a events 1988 89 1996 1998 99 00 In early 1983 during the 1982 83 El Ni o the increase in salinity was 0 5 in the Samoa Fiji region In 1989 during the 1988 89 La Ni a event the decrease in salinity was 10 0 62 in the same region Thus in the SWTP ENSO signals have twice the impact of seasonal signals on SSS The agreement between the SOI and SSS is less convincing during the 1993 1995 period Figure 6c While the SOI stayed negative from 1992 to the end of 1995 SSS returned to mean values around mid 1993 and was fresher than average from the beginning of 1994 to mid 1996 rather as if the SWTP were not influenced by the quasi permanent unusual El Ni o conditions that lasted from 1993 to 1995 at the equator Trenberth and Hoar 1996 Latif et al 1997 The strong 1997 98 Et Nifio event gave rise to the smallest positive SSS anomaly of the observed El Nifios in the time series and the greatest negative anomalies of the period of study were observed during the 1998 99 La Ni a event Due to the filtering technique the 1999 00 La Nifia does not appear on figure 6c but the monthly time series of SSS averaged over the SWTP region indicates that the negative SSS anomaly in 2000 is as great as in 1999 Figure 5 No clear explanation was found to account for the second EOF which retains 19 of the total variance 4 1 b Sea surface te
89. veraged over 10 of longitude to fit the SSS and SST grid size 3 Mean fields and seasonal variability 3 1 Mean fields Mean SST and SSS distributions over the Pacific Ocean from 1979 to 1992 are summarized in Delcroix 1998 figure 2 of this paper being reproduced here Figure 1 zoom on the SWTP region is presented in Figure 4 for the 1976 2000 period These maps are very similar to those presented in Delcroix and H nin 1989 computed for the years 1979 81 1984 85 but excluding the 1982 1983 El Ni o period This region is characterized by a horizontal SSS gradient between the south eastward oriented tongue of fresh water featuring the warm pool in the north west part of the study area and the westward oriented tongue of high salinity water advected by the southern branch of the South Equatorial Current from the central south Pacific Figure 4a The low salinity tongue lies under the SPCZ and minimum salinity lt 34 7 is found where precipitation is maximum at 10 S 175 E gt 0 25 m month Figure 4c Precipitation is low on either side of the SPCZ in the southwest corner close to New Caledonia and in the east in French Polynesia where salinity is highest The SST field presents quasi zonally oriented isotherms with waters warmer than 28 C north of 16 S The 28 C isotherm is generally used as an arbitrary limit defining the area of the warm pool South of 16 S the meridional gradient of SST increases and SST falls bel

Contribution à l`étude de la variabilité climatique de l`Océan

Contents

Download Pdf Manuals

Related Search

Related Contents

Contribution &agrave; l`&eacute;tude de la variabilit&eacute; climatique de l`Oc&eacute;an

Contents

Download Pdf Manuals

Related Search

Related Contents

Contribution à l`étude de la variabilité climatique de l`Océan