Indian summer monsoon variability during the holocene as recorded in sediments of the Arabian Sea: Timing and implications

Indian monsoon precipitation fluctuated significantly during the Holocene and a reliable reconstruction of the timing of the events and their implications is of great benefit to our understanding of the effect and response of low latitude climate systems to the forcing factors. We have carried out high-resolution terrigenous proxy studies on a laminated sediment core from the Oxygen Minimum Zone of the eastern Arabian Sea margin to reconstruct the summer monsoon-controlled precipitation changes during the Holocene. The temporal variation in the terrigenous proxy indicators of this core, in combination with other high-quality cores from the Arabian Sea, suggests several abrupt events in monsoon precipitation throughout the Holocene. The early Holocene monsoon intensification occurred in two abrupt steps at 9500 and 9100 years BP and weakened gradually thereafter, starting at 8500 years BP. A weakening in precipitation recorded at ∼7000 years BP, synchronous with similar conditions in India. One of the most significant weak monsoon periods recorded in our studies lies between 6000 and 5500 years BP. Spectral analysis of the precipitation records reveals statistically significant periodicities at 2200, 1350, 950, 750, 470, 320, 220, 156, 126, 113, 104 and 92 years. Most of these millennial-to-centennial cycles exist in various monsoon records as well as the tree ring Δ¹⁴C data and/or other solar proxy records. We suggest that throughout the Holocene, externally, small changes in solar activity controlled the Indian monsoon to a large extent, whereas internally, non-solar causes could have influenced the amplitude of decadal-to-centennial oscillations.     

Surface layer temperature inversion in the Arabian Sea during winter

Surface layer temperature inversion in the south eastern Arabian Sea, during winter has been studied using Bathythermograph data collected from 1132 stations. It is found that the inversion in this area is a stable seasonal feature and the occurrence is limited to the coastal waters. The inversion layer is found to have thickness varying from 10 to 80 meters and gradient of 0.0–1.2°C. The causative factor for the inversion is identified to be the winter-time surface-advection of cold less saline Bay of Bengal water over the warm saline Arabian Sea water along the west coast of India. Finally, the possible forcing mechanism for such an advection was examined using a hydrographic section and wind observations along the west coast of India.     

Waves in the nearshore waters of northern Arabian Sea during the summer monsoon

Waves at 15 m water depth in the northern Arabian Sea are measured during the summer monsoon for a period of 45 days and the characteristics are described. The significant wave height varied from 1.1 to 4.5 m with an average value of 2.5 m. 75% of the wave height at the measurement location is due to the swells arriving from the south-west and the remaining is due to the seas from south-west to north-west. Wave age of the measured data indicates that the waves in the nearshore waters of northern Arabian Sea during the summer monsoon are swells with young sea.     

Observation and modelling of the variability of the Arabian Sea upper layers during the development of the SW monsoon in 1988 and 1989

Experimental data on the vertical structure of theT, S temporal variations in the upper 1000 m layer of the Arabian Sea during the inter-monsoon period from April to June are reported. Variations in the upper mixed layer (UML) thickness and temperature distributed horizontally are considered. The local UML model was used to compute these variations from the atmospheric influence preset from observations.Translated by Mikhail M. Trufanov.     

Living planktonic foraminifera during the late summer monsoon period in the Arabian Sea

During the late summer monsoon living planktonic foraminifera were collected in the southeastern Arabian Sea between 3°N and 15°N by using six vertical plankton tows. Sixteen species of planktonic foraminifera were identified. Among them, Globigerinoides ruber and Globigerinoides sacculifer are the most abundant species, while the ecologically most important species Globigerina bulloides is very rare. The low abundance of G. bulloides can be explained by the warming of the surface water in combination with deepening of the mixed layer, since this species preferentially dwells in nutrient-rich upwelling waters. The population density of planktonic foraminifera ranges between 31 and 185 specimens per 10⁻³ m³. The low absolute numbers of planktonic foraminifera are similar to the numbers which were reported before from the non-upwelling areas in the Arabian Sea. The low absolute numbers and the collected foraminiferal assemblages are therefore highly indicative of the Arabian Sea non-upwelling areas. Particularly significant are the low absolute and relative numbers of the non-spinose species Globorotalia menardii and Neogloboquadrina dutertrei. The absence of these species indicate the relatively low nutrient levels in this area at the tail end of the summer monsoon period.     

Zooplankton herbivory in the Arabian Sea during and after the SW monsoon, 1994

Microzooplankton herbivory in the Arabian Sea was measured using dilution experiments towards the end of the SW monsoon in September and during the intermonsoon to NE monsoon period in November–December 1994. Microzooplankton grazing resulted in a turnover of phytoplankton stocks that ranged from 11 to 49% per day. This was equivalent to grazing fluxes of between 1 and 17 mg C m^-3 d^-1. Depth-integrated microzooplankton herbivory ranged between 161 and 415 mg C m^-2 d^-1 during the SW monsoon cruise, and between 110 and 407 mg C m^-2 d^-1 during the intermonsoon period. Microzooplankton grazed between 4 and 60% of daily primary production, with higher percentages found during the intermonsoon season. Phytoplankton growth rates during the SW monsoon ranged from 0.3 to 1.8 d^-1, with lower values in upwelling waters and higher values in downwelling and oligotrophic areas. During the intermonsoon period, phytoplankton growth was more uniform across the basin and averaged 0.68±0.15 d^-1. Microzooplankton abundance in experimental samples varied between 2800 and 16 162 cells l^-1, equivalent to a biomass of between 1.1 and 7.2 mg C m^-3. The mean cell carbon content of microzooplankton was similar in both periods and ranged from 0.33 to 0.55 ng C cell^-1. Microzooplankton were smallest in downwelling waters and largest in oligotrophic waters. Average clearance rates in those taxa that took up fluorescently-labelled algae ranged from 0.2 to 14 μl ind^-1 hr^-1. Average mesozooplankton grazing rates, derived from biomass data, varied from 19 to 92 mg C m^-2 d^-1; these rates accounted for removal of between 4 and 12% of the daily primary production. Mesozooplankton herbivory was most pronounced in upwelling and downwelling waters and reduced in stratified oligotrophic waters during the SW monsoon period. Microzooplankton herbivory was greater than the average mesozooplankton herbivory at all stations, during both the SW monsoon and intermonsoon periods.     

Phytoplankton community structure in the Arabian Sea during and after the SW monsoon, 1994

Diatoms, dinoflagellates, coccolithophores, nanoflagellates, picophytoplankton and procaryote algae (Synechococcus spp. and prochlorophytes) were quantified by microscopy and flow cytometry, and their biomass determined, at 12 stations along a 1600 km transect across the Arabian Sea at the end of the SW monsoon in September, and during the inter-monsoon period of November/December 1994. The transect spanned contrasting oceanic conditions that varied from seasonally eutrophic, upwelling waters through mesotrophic, downwelling waters to permanently oligotrophic, stratified waters. The overall diversity of diatoms, dinoflagellates and coccolithophores along the transect was not significantly different between the SW monsoon and inter-monsoon. However, diatoms showed greatest diversity during the SW monsoon and coccolithophores were most diverse during the inter-monsoon. Integrated phytoplankton standing stocks during the SW monsoon ranged from 3 to 9 g C m^-2 in the upwelling eutrophic waters, from 3 to 5 g C m^-2 in downwelling waters, and from 1 to 2 g C m^-2 in oligotrophic waters. Similar phytoplankton standing stocks were found in oligotrophic waters during the inter-monsoon, but were ca. 40% lower compared to the SW monsoon in the more physically dynamic waters. Phytoplankton abundance and biomass was dominated by procaryote taxa. Synechococcus spp. were abundant (often >10⁸ cells l^-1) during both the SW monsoon and inter-monsoon, where the nitrate concentration was ⩾0.1 μ mol l^-1, and often dominated the phytoplankton standing stocks. Prochlorophytes were restricted to oligotrophic stratified waters during the SW monsoon period but were found at all stations along the transect during the inter-monsoon, dominating the phytoplankton standing stocks (>40%) in the oligotrophic region during this period. Of the nano- and micro-phytoplankton, only diatoms contributed significantly to phytoplankton standing stocks, and then only in near-shore upwelling waters during the SW monsoon. There were significant changes in the temporal composition of the phytoplankton community. In nearshore waters a mixed community of diatoms and Synechococcus spp. dominated during the SW monsoon. This gave way to a community dominated by Synechococcus spp. in the intermonsoon. In the downwelling zone, a Synechococcus spp. dominated community was replaced by a mixed procaryote community of Synechococcus spp. and prochlorophytes. In the oligotrophic stratified waters, the mix of procaryote algae was replaced by one dominated by prochlorophytes alone.     

Distribution of biogenic sulphur compounds during and just after the southwest monsoon in the Arabian Sea

The Arabian Sea is characterised by strong seasonal oscillations of biological productivity generated by its monsoonal climate. The southwest monsoon causes reversal in the surface circulation of the Arabian Sea, which generates a seasonal upwelling of nutrient-rich waters along the coast of Oman. Concentrations of biogenic sulphur compounds were measured on a transect from the eutrophic waters off the coast of Oman to the oligotrophic waters of the open Arabian Sea, during the UK NERC Arabesque cruise 27 August–4 October 1994. The concentrations of dimethylsulphide (DMS), dimethylsulphoxide (DMSO) and dimethylsulphoniopropionate (DMSP) were found to be elevated in the eutrophic area due to enhanced biological production. However, this increase in DMS, DMSO and DMSP concentration was not observed until after the southwest monsoon had relaxed, and appeared to correspond to increased concentrations of hexanoyloxyfucoxanthin, an indicator of prymnesiophytes. DMSO concentrations were correlated with those of DMS and DMSP in the near surface waters of the Arabian Sea. Additionally, DMSO appeared to be ubiquitous throughout the water column, being easily detectable in deep waters, which suggests that DMSO may act as a sink for DMS in the world’s oceans.     

Bacterioplankton activity in the surface waters of the Arabian Sea during and after the 1994 SW monsoon

Bacterial biomass and production were measured on two cruises to the northwestern Arabian Sea in 1994; the first cruise took place towards the end of the SW monsoon in September, and the second cruise during the inter-monsoon period in November and December. Although phytoplankton production was significantly higher during the monsoon, bacterial numbers showed little difference. Bacteria were most abundant in the euphotic zone and highest bacterial numbers were measured during the monsoon period in the Gulf of Oman and the shelf waters off southern Oman; in these regions, numbers ranged from 0.9 to 1.6×10⁹ bacteria l^-1. On both cruises, bacteria were less abundant in the euphotic zone of the central Arabian Sea and typically ca 0.8×10⁹ cells l^-1 were present. The majority of bacteria (80–95%) were small cocci that were larger (median diameter 0.40 μm) during the monsoon period than the inter-monsoon, when the cells had a diameter of 0.36 μm; there was no comparable change in cell dimensions of bacteria present as rods. Bacterial production was measured by the incorporation of ³H-thymidine and ³H-leucine. On both cruises, uptake rates were highest on the Omani shelf and decreased offshore. In the central Arabian Sea, thymidine incorporation rates were similar in the monsoon and inter-monsoon periods, but higher rates of leucine incorporation were measured during the monsoon period. Bacterial production was a relatively small proportion of phytoplankton production in both periods sampled; bacterial production was equivalent to between 10 and 30% of the daily primary production in the Arabian Sea.     

阿拉伯海东南海域盐度收支的季节变化

采用SODA海洋同化产品的月平均资料,本文分析了阿拉伯海东南海域表层盐度的季节变化特征,发现局地海面淡水通量不能解释盐度的变化。两个典型区域的表层海水盐度收支分析表明,海洋的平流输送是造成阿拉伯海东南海域盐度冬季降低、夏季升高的主要原因,而淡水通量仅在夏季印度西侧沿岸区域造成盐度降低。冬季,东北季风环流将孟加拉湾北部的低盐水沿同纬度输送到阿拉伯海,然后向北输送,使表层海水盐度降低;夏季,西南季风环流把阿拉伯海西北部的高盐水向南、向东输送,使阿拉伯海东南海域盐度升高。受地理位置因素的影响,阿拉伯海东南海域表层盐度的变化冬季明显强于夏季。     

The influence of the south-west monsoon upon the nutrient biogeochemistry of the Arabian Sea

Variations in the nutrient concentrations were studied during two cruises to the Arabian Sea. The situation towards the end of the southwest monsoon season (September/October 1994) was compared with the inter-monsoonal season during November and December 1994. Underway surface transects showed the influence of an upwelling system during the first cruise with deep, colder, nutrient-rich water being advected into the surface mixed layer. During the southwesterly monsoon there was an area of coastal Ekman upwelling, bringing colder water (24.2°C) into the surface waters of the coastal margin. Further offshore at about 350 km there was an area of Ekman upwelling, as a result of wind-stress curl, north of the Findlater Jet axis; this area also had cooler surface water (24.6°C). Further offshore (>1000 km) the average surface temperatures increased to >27°C. These waters were oligotrophic with no evidence of the upwelling effects observed further inshore. In the upwelling regions nutrient concentrations in the close inshore coastal zone were elevated (NO₃=18 μmol l^-1, PO₄=1.48 μmol l^-1); higher concentrations also were measured at the region of offshore upwelling off the shelf, with a maximum nitrate concentration of 12.5 μmol l^-1 and a maximum phosphate concentration of 1.2 μmol l^-1. Nitrate and phosphate concentrations decreased with increasing distance offshore to the oligotrophic waters beyond 1400 km, where typical nitrate concentrations were 35.0 nmol l^-1 (0.035 μmol l^-1) in the surface mixed layer. A CTD section from the coastal shelf, to 1650 km offshore to the oligotrophic waters, clearly showed that during the monsoon season, upwelling is one of the major influences upon the nutrient concentrations in the surface waters of the Arabian Sea off the coast of Oman. Productivity of the water column was enhanced to a distance of over 800 km offshore. During the intermonsoon period a stable surface mixed layer was established, with a well-defined thermocline and nitracline. Surface temperature was between 26.8 and 27.4°C for the entire transect from the coast to 1650 km offshore. Nitrate concentrations were typically between 2.0 and 0.4 μmol l^-1 for the transect, to about 1200 km where the waters became oligotrophic, and nitrate concentrations were then typically 8–12 nmol l^-1. Ammonia concentrations for the oligotrophic waters were typically 130 nmol l^-1, and are reported for the first time in the Indian Ocean. The nitrogen/phosphorus (N/P) ratios suggest that phytoplankton production was potentially nitrogen-limited in all the surface waters of the Arabian Sea, with the greatest nitrogen limitation during the intermonsoon period.     

A cold pool south of Indo-Sri Lanka channel and its intrusion into the Southeastern Arabian Sea during winter

During winter, south of the Indo-Sri Lanka Channel (ISLC), the observed sea-surface temperature (SST) distribution shows a distinct mini-cold pool (MCP) with relatively cooler waters (SST<28 °C). All the available satellite and in-situ measurements are utilized to characterize and explain the mechanisms that govern the evolution of the observed MCP. During December–January, the northeasterly surface winds blow through the ISLC manifesting a patch of strong winds in the south with peak intensity of about 10 m/s, enhance surface turbulent heat losses and drive near-surface vertical mixing resulting in the observed cooling. The vertical temperature profiles in this region also show cooling and deepening of the near-surface isothermal layer from November to January. This cooling occurs episodically on an intra-seasonal time scale with a typical periodicity of 8–15 days and is stronger when the surface winds intensify, surface net heat losses are larger and the near-surface circulation is more pronounced. The cooling episodes varied in number, intensity, duration and spatial extent in each winter during 1998–2006. The cooler surface waters from this MCP flow initially southwestward and are then topographically steered northwestward by the Maldives Island Chain. The resultant near-surface circulation also appears to strengthen the amplitude of the near-surface thermal inversions observed in the SouthEastern Arabian Sea (SEAS).     

Seasonal variability of sonic layer depth in the Central Arabian Sea

The seasonal variability of sonic layer depth (SLD) in the central Arabian Sea (CAS) (0 to 25°N and 62-66°E) was studied using the temperature and salinity (T/S) profiles from Argo floats for the years 2002–2006. The atmospheric forcing responsible for the observed changes was explored using the meteorological data from NCEP/NCAR and Quickscat winds. SLD was obtained from sound velocity profiles computed from T/S data. Net heat flux and wind forcing regulated SLD in the CAS. Up-welling and down-welling (Ekman dynamics) associated with the Findlater Jet controlled SLD during the summer monsoon. While in winter monsoon, cooling and convective mixing regulated SLD in the study region. Weak winds, high insolation and positive net heat flux lead to the formation of thin, warm and stratified sonic layer during pre and post summer monsoon periods, respectively.     

Nitrogen dynamics during the Arabian Sea Northeast Monsoon

This investigation focused on the weaker and less well understood of the two Arabian Sea monsoonal wind phases, the NE Monsoon, which persists for 3–4 months in the October to February period. Historically, this period has been characterized as a time of very low nutrient availability and low biological production. As part of the US JGOFS Arabian Sea Process Study, 17 stations were sampled on a cruise in January 1995 (late NE Monsoon) and, 15 stations were sampled on a cruise in November 1995 (early NE Monsoon). Only the southern most stations (10° and 12°N) and one shallow coastal station were as nutrient-depleted as had been expected from the few relevant prior studies in this region. Experiments were conducted to ascertain the relative importance of different nitrogenous nutrients and the sufficiency of local regeneration processes in supplying nitrogenous nutrients utilized in primary production. Except for the southern oligotrophic stations, the euphotic zone concentrations of NO₃⁻ were typically 5–10-fold greater than those of NO₂⁻ and NH₄⁺. There was considerable variation (20–40-fold) in nutrient concentration both within and between the two sections on each cruise. All nitrogenous nutrients were more abundant (2–4-fold) later in the NE Monsoon. Strong vertical gradients in euphotic zone NH₄⁺ concentration, with higher concentrations at depth, were common. This was in contrast to the nearly uniform euphotic zone concentrations for both NO₃⁻ and NO₂⁻. Half-saturation constants for uptake were higher for NO₃⁻ (1.7 μmol kg⁻¹ (s.d.=0.88, n=8)) than for NH₄⁺ (0.47 μmol kg⁻¹ (s.d.=0.33, n=5)). Evidence for the suppressing effect of NH₄⁺ on NO₃⁻ uptake was widespread, although not as severe as has been noted for some other regions. Both the degree of sensitivity of NO₃⁻ uptake to NH₄⁺ concentration and the half-saturation constant for NO₃⁻ uptake were correlated with ambient NO₃⁻ concentration. The combined effect of high affinity for low concentrations of NH₄⁺ and the effect of NH₄⁺ concentration on NO₃⁻ uptake resulted in similarly low f-ratios, 0.15 (s.d.=0.07, n=15) and 0.13 (s.d.=0.08, n=17), for early and late observations in the NE Monsoon, respectively. Stations with high f-ratios had the lowest euphotic zone NH₄⁺ concentrations, and these stations were either very near shore or far from shore in the most oligotrophic waters. At several stations, particularly early in the NE Monsoon, the utilization rates for NO₂⁻ were equal to or greater than 50% the utilization rates for NO₃⁻. When converted with a Redfield C : N value of 6.7, the total N uptake rates measured in this study were commensurate with measurements of C productivity. While nutrient concentrations at some stations approached levels low enough to limit phytoplankton growth, light was shown to be very important in regulating N uptake at all stations in this study. Diel periodicity was observed for uptake of all nitrogenous nutrients at all stations. The amplitude of this periodicity was positively correlated with nutrient concentration. The strongest of these relationships occurred with NO₃⁻. Ammonium concentration strongly influenced the vertical profiles for NO₃⁻ uptake as well as for NH₄⁺ uptake. Both NO₂⁻ and NH₄⁺ were regenerated within the euphotic zone at rates comparable to rates of uptake of these nutrients, and thus maintenance of mixed layer concentrations did not require diffusive or advective fluxes from other sources. Observed turnover times for NH₄⁺ were typically less than one day. Rapid turnover and the strong light regulation of NH₄⁺ uptake allowed the development and maintenance of vertical structure in NH₄⁺ concentration within the euphotic zone. In spite of the strong positive effect of light on NO₂⁻ uptake and its strong negative effect on NO₂⁻ production, the combined effects of much longer turnover times for this nutrient and mixed layer dynamics resulted in nearly uniform NO₂⁻ concentrations within the euphotic zone. Responses of the NE Monsoon planktonic community to light and nutrients, in conjunction with mixed layer dynamics, allowed for efficient recycling of N within the mixed layer. As the NE Monsoon evolved and the mixed layer deepened convectively, NO₂⁻ and NO₃⁻ concentrations increased correspondingly with the entrainment of deeper water. Planktonic N productivity increased 2-fold, but without a significant change the new vs. recycled N proportionality. Consequently, NO₃⁻ turnover time increased from about 1 month to greater than 3 months. This reflected the overriding importance of recycling processes in supplying nitrogenous nutrients for primary production throughout the duration of the NE Monsoon. As a result, NO₃⁻ supplied to the euphotic zone during the NE Monsoon is, for the most part, conserved for utilization during the subsequent intermonsoon period.     

Seasonal and spatial variability in N2O distribution in the Arabian Sea

Extensive measurements of nitrous oxide (N₂O) were made in the central and eastern Arabian Sea during the northeast monsoon (February–March), intermonsoon (April–May) and southwest monsoon (July–August) seasons. The latitudinal and longitudinal variations, along with seasonal changes with respect to winter convection and coastal upwelling, are clearly discernible. Vertical profiles collected down to 1000 m show that the Arabian Sea water column is supersaturated with N₂O at all depths. However, N₂O consumption at intermediate depths, coincident with the oxygen minimum and associated with sediment–water interfaces, and in the denitrifying zone, coincident with NO^-₂ secondary maxima, are also apparent. The N₂O concentration varies from ∼10 nM near the surface to about 80 nM in the secondary peak region (≈800 m). Interrelationships with chemical parameters suggest nitrification to be the main process for the production of N₂O in the oceanic water. Plots of apparent oxygen utilization vs production of N₂O indicate a consistent linear relationship for AOU between 0 and 200 μM.     

Seasonal and annual variability of vertically migrating scattering layers in the northern Arabian Sea

A 30-month time series of mean volume backscattering strength (MVBS) data obtained from moored acoustic Doppler current profilers (ADCPs) is used to analyze the evolution of vertically migrating scattering layers and their seasonal and annual variability in the Arabian and Oman Seas. Substantial diel vertical migration (DVM) is observed almost every day at all three mooring sites. Two daytime layers (Layers D1 and D2) and one nighttime layer (Layer E1) are typically present. The greatest biomass is observed near the surface during the night in Layer E1 and at depth between 250 and 450 m during the daytime in Layer D2. All layers are deepest during the spring inter-monsoon and shallowest during the summer/fall southwest monsoon (SWM). Seasonal modulation of the D2 biomass change is evident in our high-resolution data. The lowest biomass in D2 is measured in the early summer (May or June) followed by a rapid biomass increase during the SWM (June–November) until the biomass reaches a maximum at the end of the SWM season. Short-period oscillations in D2 biomass are often seen with periods ranging from days to one month. Occasionally, a lower nighttime layer E2 is formed between 180 and 270 m, mostly near the time of full moons. The upper daytime layer D1 is centered at 200 m and densely concentrated. It is only formed during the winter northeast monsoon (NEM) and the spring inter-monsoon. The influence of physical processes on layer distribution is also investigated. Interestingly, the two daytime layers are found to be formed at the two boundaries of the Persian Gulf outflow water (PGW) and follow the seasonal depth change of the PGW. The timing of the DVM and the formation, persistence, decay and reformation of the deep scattering layers seem to be governed by light, both solar and lunar. The scattering strength, the layer depth and the layer thickness are likewise closely related to the Moon phase at night. Cloud coverage, the isotherm and the isohaline also appear to affect the distribution and depth of the scattering layers. The continuous multiple-year acoustic data from ADCPs allow us, for the first time, to study the seasonal and annual variations of scattering layers in this region.     

Observed intra-seasonal to interannual variability of the upper ocean thermal structure in the southeastern Arabian Sea during 2002–2008

The southeastern Arabian Sea (SEAS), located in the Indian Ocean warm pool, is a key-region of the regional climate system. It is suspected to play an important role in the dynamics of the Asian summer monsoon system. The present study reports the salient features derived from a newly harvested observational dataset consisting of repeated fortnightly XBT transects in the SEAS over the period 2002–2008. The fortnightly resolution of such a multi-year record duration is unprecedented in this part of the world ocean and provides a unique opportunity to examine the observed variability of the near-surface thermal structure over a wide spectrum, from intra-seasonal to interannual timescales. We find that most of the variability is trapped in the thermocline, taking the form of upwelling and downwelling motions of the thermal stratification. The seasonal variations are consistent with past studies and confirm the role of the monsoonal wind forcing through linear baroclinic waves (coastally-trapped Kelvin and planetary Rossby waves). Sub-seasonal variability takes the form of anomalous events lasting a few weeks to a few months and occurs at two preferred timescales: in the 30–110 day band, within the frequency domain of the Madden–Julian oscillation and in the 120–180 day band. While this sub-seasonal variability appears fairly barotropic in the offshore region, the sign of the anomaly in the upper thermocline is opposite to that in its lower part on many occasions along the coast. Our dataset also reveals relatively large interannual temperature variations of about 1 °C from 50 to 200 m depth that reflect a considerable year-to-year variability of the magnitude of both upwelling and downwelling events. This study clearly demonstrates the necessity for sustained long-term temperature measurements in the SEAS.