Taxes & Accounting

Seasonal and interannual variability of chlorophyll-ain the Gulf of Oman compared to the open Arabian Sea regions

Seasonal and interannual variability of chlorophyll-ain the Gulf of Oman compared to the open Arabian Sea regions
of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Deep-Sea Research II 53 (2006) 1548–1559 Seasonal and inter-annual variability of chlorophyll- a concentration in the Mauritanian upwelling: Observation of ananomalous event during 1998–1999 Yaswant Pradhan a,c,  , Samantha J. Lavender a,c ,Nick J. Hardman-Mountford b,c , James Aiken b,c a SEOES, University of Plymouth, Plymouth PL4 8AA, UK  b Plymouth Marine Laboratory, Plymouth PL1 3DH, UK  c Centre for observation of Air–Sea Interactions and fluXes (CASIX), UK  Received 2 August 2005; received in revised form 9 February 2006; accepted 12 May 2006Available online 10 August 2006 Abstract Monthly chlorophyll- a  (Chl- a ) concentrations derived from SeaWiFS data for 1997–2005 and chlorophyll measurementsfrom the Atlantic Meridional Transect for 1995–2001 have been analysed to describe seasonal and inter-annual variabilityof surface Chl- a  in the Mauritanian upwelling. There was a moderate to strong correspondence between the seasonal cyclesof surface Chl- a  and the seasonal cycles of ocean physical and meteorological fields (such as sea-surface temperature, sea-surface height, and prevailing wind), with a noticeable exception in 1998 that corresponded to a strong anomalous Chl- a event (  250% increase) in the Mauritanian upwelling. Alongshore wind-stress and wind-stress curl were found to be themost significant factors controlling the variability of Chl- a  (jointly explaining more than 50% of total variance). Thebiological response to the alongshore wind-stress was immediate, but it lagged the wind-stress curl by 1–2 months (eachexplaining more than 40% of the total Chl- a  variability). These observations also demonstrate a link, hitherto unreported,between the Pacific El-Nin ˜o Southern Oscillation (ENSO) and anomalous Chl- a  field in the Mauritanian upwelling. Themultivariate ENSO index was shown to account for a significant part of the variability of the autumn–winter Chl- a anomaly ( r ¼ 0.52,  p o 0.01). A cold event, following an intense El Nin ˜o in the Pacific during summer, was found tomirror the intensity of wind forcing and phytoplankton concentration in the Mauritanian upwelling a few months later.Therefore, ENSO-related changes in the local atmospheric fields are considered as the preferred candidates for explainingthe observed biological changes in the Mauritanian upwelling during 1998–1999. r 2006 Elsevier Ltd. All rights reserved. Keywords:  Mauritanian upwelling; Chlorophyll- a ; ENSO; Seasonal variations; Time-series; AMT 1. Introduction Upwelling ecosystems associated with Easternboundary currents (EBC) are characterised byhigh phytoplankton production (Carr, 2001) andhigh levels of spatial and temporal variability ARTICLE IN PRESS$-see front matter r 2006 Elsevier Ltd. All rights reserved.doi:10.1016/j.dsr2.2006.05.016  Corresponding author. SEOES, University of Plymouth,Plymouth PL4 8AA, UK. Tel.: +441752232435;fax: +441752232406. E-mail address: Pradhan).  (Hill et al., 1998). Over seasonal and inter-annualtime scales, this variability may have significantimplications for carbon-cycling in the oceans andthe sustainability of commercially important fishpopulations.The Atlantic Meridional Transect (AMT) pro-gramme sampled the surface ocean between the UKand Falkland Islands twice yearly (Aiken and Bale,2000). The Mauritanian upwelling (Fig. 1), 10–25 1 Nand between 30 1 W to the northwest African coast, istraversed by these AMT tracks. Persistent north-easterly trade winds drive upwelling along thenorthwest African coast and, with the  CanaryCurrent  (Zhou et al., 2000) (Fig. 1), determine the phytoplankton distribution in the Mauritanianupwelling system. The  Canary Current  is charac-terised by strong seasonal variability and a sharpcontrast in the biology (phytoplankton blooms)between northern and southern waters (Sætersdalet al., 1999). The study area also covers the GuineaDome (centred near 22 1 W and 12 1 N) in the south(Fig. 1) that results from the interaction of theeastward-flowing  North Equatorial Counter Current and the westward-flowing  North Equatorial Current (Yamagata and Iizuka, 1995).The surface chlorophyll seasonal variability in thefour major EBC systems was described by Thomaset al. (2001b; 2004), using Coastal Zone ColourScanner (CZCS) and Sea viewing Wide Field-of-view Sensor (SeaWiFS) data. They reported that thepeak high and low concentrations in the  CanaryCurrent  were located at 10–20 1 N and   35 1 N,respectively. Signorini et al. (1999) described thespatio-temporal variability of surface chlorophyll(observed by SeaWiFS) in the tropical Atlanticduring the 1997–1998 El Nin ˜o–La Nin ˜a transition.They observed anomalous changes in physical andbiological properties throughout the eastern tropicalAtlantic, but inter-annual anomalies in the Maur-itanian upwelling were not discussed.The inter-relationships between Pacific El-Nin ˜oSouthern Oscillation (ENSO) and local anomaliesof hydro-meteorological fields in different parts of the globe have been studied extensively in recentyears. In a recent study, Larkin and Harrison (2005)showed the influence of ENSO on global seasonaltemperature and precipitation. ENSO-related hy-drographic and phytoplankton anomalies have beenreported extensively for Pacific EBCs (Chavez et al.,1999; Thomas et al., 2001a; Thomas et al., 2004). The relationship between boreal winter El-Nin ˜o sea-surface temperature (SST) anomaly and borealspring tropical SST gradient was investigated byHuang et al. (2005). An ENSO-related coastal SSTanomaly in the eastern Atlantic was also linked tothe variability in annual catches of sardine (Roy andReason, 2001). Most of these reports relate ENSOto SST or precipitation anomalies. However, acomprehensive study of the seasonal and inter-annual variability of coupled bio-physical processesin the Mauritanian upwelling (including the anoma-lies during the ENSO period) has not beenpublished to date.In this paper, the seasonal and inter-annualvariability of surface chlorophyll- a  (Chl- a ) in theMauritanian upwelling was investigated usingmore than 8 years of Earth observation (EO)data combined with in situ observations from theAMT cruises. The major goals of this paper are(i) to quantify relationships between observedChl- a  and local wind forcing (alongshore wind-stress ( t  y ) and wind-stress curl ( r t )) and (ii) toinvestigate links between the 1997–1998 ENSOand the anomalies observed in the Mauritanianupwelling. ARTICLE IN PRESS -30  ° -20    -10    0    West Longitude 0    10    20    30    40    50       N  o  r   t   h   L  a   t   i   t  u   d  e   Atlantic Meridional Transectsin the Northern Hemisphere AfricaUK AMT-02AMT-03AMT-04AMT-05AMT-06bAMT-06AMT-07AMT-08AMT-09AMT-10AMT-11AMT-01 Chl- a  [mg/m 3 ]  AzoresCanary NEC  NECC    6050403020105. Fig. 1. AMT (1995–2001) in the northern hemisphere (shown ascoloured lines) overlaid on the 8-year (1997–2005) meanSeaWiFS Chl- a  map. The arrows represent major current systems(see text). The Mauritanian upwelling zone (10–25 1 N, 30 1 W tocoast) is defined by the box and the Guinea Dome (12 1 N, 22 1 W)is shown by the dashed oval. Y. Pradhan et al. / Deep-Sea Research II 53 (2006) 1548–1559  1549  2. Data and methods  2.1. Earth observation dataSeaWiFS Chl-a : Level-3 mapped monthly meanSeaWiFS global 9-km Chl- a  data (reprocessingversion 5.1) were downloaded from National Aero-nautics and Space Administration (NASA) Ocean-color website ( The Chl- a  maps were generated using thefourth generation Maximum Band Ratio algorithm,OC4v4, that yields a reasonably good correlation( r 2 ¼ 0.892) and small error (RMS ¼ 0.222) on aglobal scale that includes samples from all watertypes (O’Reilly et al., 2000). Generally, the satellite-derived Chl- a  measurement represents the meanvalue for the surface layer derived from theupwelling radiance from the surface to first attenua-tion depth, heavily weighted towards the surface. ERS-2 and QuikSCAT vector wind  : Weeklygridded European Remote-Sensing Satellite-2(ERS2) and NASA-QuikSCAT (QS) mean windfields (MWF) at 0.5 1  resolution were obtainedthrough an anonymous FTP ( A comparisonbetween ERS2 and QS wind speeds showed goodagreement, but a systematic bias was found over theMauritanian upwelling showing an under-estima-tion of ERS2 wind speed compared to QS windspeed. Based on the concurrent (August 1999– December 2000) datasets from both sensors, thedifference was adjusted to best fit the QS wind; itwas assumed that QS performed better than ERS2since the mean estimation error (provided with theMWF product) for QS wind (0.25ms  1 ) is sig-nificantly smaller than for ERS2 wind (0.55ms  1 ).A detailed discussion is beyond the scope of thispaper, but the correction involved the application of a linear scaling to the zonal and meridionalcomponents of ERS2 wind speed as u ¼ 1 : 107 u E2 þ 0 : 4395 and v ¼ 1 : 065 v E2  0 : 6332,  ð 1 Þ where  u  and  v  are the corrected across-shore (zonal)and alongshore (meridional) wind speed compo-nents, and  u E2  and  v E2  are the respective ERS2 windspeed components. The wind-stress ( t ) was calcu-lated using variable drag coefficients ( C  D ) (Yellandand Taylor, 1996) t ¼ r a C  D j W  j W  , (2)where  r a   1 : 3kgm  3 (average air density) and  W  is the wind speed. C  D  ¼  0 : 29 þ 3 : 1 W   þ  7 : 7 W  2   10  3 for3 p W  p 6ms  1   C  D  ¼  0 : 6 þ 0 : 07 W  ð Þ 10  3 for6 p W  p 26ms  1    or  C  D  ¼ 10  4 ;  otherwise : The 9-year (1997–2005) monthly wind parameter(wind speed, wind-stress and wind-stress curl) time-series was constructed using both ERS2 and QSafter applying this correction (Eq. (1)) to the ERS-2data. Additional EO and climatological data : TheReynolds Optimally Interpolated SST dataset (Rey-nolds and Smith, 1994) is based upon a combinationof advanced very high-resolution radiometer satel-lite and in situ data 1 , available from the NASAPhysical Oceanography Distributed Active ArchiveCenter (PO.DAAC Product  ] 119). The 1/3 1  merged(TOPEX/Poseidon, Jason and ERS1/2) altimetersea-level anomaly (SLA) data were produced andprovided by the enhanced ocean data assimilationand climate prediction project (EVK2-CT2001-00117), and are available via an anonymous FTP( multivariate ENSO index (MEI) datafor the 1997–2005 period, available from the NOAAClimate Diagnostics Centre (, were used to verify thelarge-scale atmospheric modulation on the phyto-plankton variability in the study area. All MEIvalues are normalised for each bimonthly period sothat the 44 values (from 1950 to 1993) have zeromean and unit variance (Wolter, 2006). The MEIwas used since it integrates more information thanthe Southern Oscillation Index or various SSTindices; it reflects the nature of the coupledocean–atmosphere system better than either com-ponent and it is less vulnerable to occasional dataglitches in the monthly update cycles (Wolter, 2006).  2.2. In situ data To obtain the vertical distribution of Chl- a , datafrom fluorometers attached to the AMT conducti-vity–temperature–depth (CTD) frame were used. ARTICLE IN PRESS 1 and the references therein Y. Pradhan et al. / Deep-Sea Research II 53 (2006) 1548–1559 1550  The fluorometer data were calibrated at the BritishOceanographic Data Centre (BODC) using theflurometric Chl- a  measurements made on CTDbottle samples. CTD bottle samples were alsoanalysed to estimate the total Chl- a  (Tchl- a )concentration using the high-performance liquidchromatography (HPLC) technique (Aiken et al.,1998); Tchl- a  is the sum of chlorophyllide- a , Chl- a epimer, Chl- a  allomer, monovinyl Chl- a  and divinylChl- a  (Mueller et al., 2003). Tchl- a  values were thenaveraged over the top 15m in the Mauritanianupwelling region.All the HPLC data collected during AMT cruiseswere obtained from the SeaWiFS Bio-opticalArchive and Storage System (SeaBASS) publicaccess database ( Available nutrient data(NO 3 +NO 2  concentration by colorometric auto-analysis method), particulate organic nitrogen(PON) and CTD profile data were collected onAMT cruises and provided by BODC.  2.3. EO data analysis For all latitude–time cross-sections described inthis paper, the monthly (and reconstructedmonthly) EO data were pre-interpolated to a1 1  1 1  grid space using a kriging method (Stein,1999). Monthly anomalies were then computed asdepartures from 8-year (1998–2005) mean monthlyvalues. Before constructing the spatially averagedtime-series, each anomaly data cube (longitude,latitude and time) was standardised over time sothat the data in each grid point, in the entiremonthly time-series (1997–2005), have an average of ‘0’ and a standard deviation of ‘1’. The strength anddirection of the relationship between surface Chl- a (prediction) variability and the forcing fields (ex-planatory) were evaluated using Spearman’s rankcorrelation coefficient as it provides a credible resultwhen the sample size is relatively small (Altman,1991). The combined effect of all independentvariables (only strength) also was determinedthrough multiple correlation analysis (Huberty,2003). 3. Results The meridional cross-section (zonally averaged)of SeaWiFS estimated Chl- a  concentration in theMauritanian upwelling, between 10 and 20 1 N,showed a strong seasonal cycle dominated by aspring peak (Fig. 2A). Elevated concentrations( 4 1.0mgm  3 ) were present from January–Mayand typically peaked (up to   8.0mgm  3 ) inFebruary–March in the south and April–Mayfurther north. Around 20 1 N, elevated Chl- a  con-centrations persisted throughout the year; it isaround this latitude that the main branch of the  Canary Current  leaves the African coast. Thisseasonal cycle of Chl- a  was consistent throughoutthe time series, with the exception of 1998 whenan anomalous second peak of greater magnitudewas observed from September to December. Atime–latitude plot of Chl- a  anomalies (Fig. 2B)showed the magnitude of this anomalous event( 4 2.5mgm  3 ).To validate the observed Chl- a  anomaly, latitu-de–depth sections from AMT cruises were examined(Fig. 3). These revealed a 2–3 times enhancement of Chl- a  concentration between 10 and 22 1 N, duringspring and autumn 1998 as compared with otheryears. In contrast, the surface signature as estimatedby satellite showed only autumn 1998 values to be ARTICLE IN PRESS 1998199920002001 2002 200320042005  19981999200020012002 200320042005  19981999200020012002 200320042005 SeaWiFS monthly mean Chlorophyll-a [mg m   -3   ] 10    15    20    25    10    15    20    25    10    15    20    25    0.050.505.00 N   10.0 SeaWiFS monthly Chlorophyll-a anomaly [mg m  -3   ]  -5-2.502.55 Year ERS2_QS monthly Windspeed anomaly [m s   -1  ]  -1-0.500.51 (A)(B)(C) Fig. 2. Zonally averaged (30 1 W–NW African coast) time–-latitude plots of (A) monthly mean SeaWiFS Chl- a  (values 4 1.0mgm  3 are shown as bold contour), (B) SeaWiFS monthlyChl- a  anomalies from mean monthly Chl- a  values for the 8-yearperiod (1998–2005) and (C) same as B, but for the reconstructedERS2_QS wind speed anomaly (see methods). Y. Pradhan et al. / Deep-Sea Research II 53 (2006) 1548–1559  1551  anomalously high (Fig. 2B). Even though thespace–time sampling differences between the twomethods are obvious, both showed major anomaliesin 1998. The surface (0–15m) in situ samplesshowed that, during 1998, elevated Tchl- a  (fromHPLC) coincided with 2–3 times rise in nutrient(NO 3 +NO 2 ) and PON levels (Fig. 4). The inter-annual variability of Chl- a , nutrients and PONwere very similar ( r (Chl–  a , nutrient) ¼ 0.7,  p o 0.1; r (Chl–  a , PON) ¼ 0.6,  p ¼ 0.1). Particulate organicmatter such as carbon (POC) and nitrogen (PON)comprise mostly faecal pellets and other detritalmaterial (arising mainly from zooplankton grazingof phytoplankton), higher values occurring withhigher phytoplankton biomass.To investigate the relationship between theseasonal cycle of surface Chl- a  and putative physicaldrivers (SST, wind-stress, alongshore wind-stress,wind-stress curl and SLA), a cross-correla-tion analysis was performed on the mean fieldsfor each variable. The strongest correlation wasseen with meridional wind-stress ( t  y ) without a timelag (Table 1). All the other physical variablesshowed modest correlations ( r 4 0.6) with Chl- a concentration (Table 1); the period of analysiswas from 1999 to 2005 so that 1998 was notincluded. When 1998 was included, the seasonalcorrelations were substantially reduced ( r o 0.5) forall variables.To further investigate forcing of this anomalousevent, time–latitude plots of anomalies in physicalvariables were examined. Wind speed anomalies(Fig. 2C) exceeding 0.5ms  1 were observedthroughout the whole region, from 10 to 25 1 N,during autumn–winter 1998. However, other anom-alous wind events did not correspond to anomalous ARTICLE IN PRESS autumn spring Year  1996  No data 5N 10N 15N 20N 25N     1   9   9   6   (   S  e  p  -   O  c   t   ) 50m 100m  1997  No data 5N 10N 15N 20N 25N     1   9   9   7   (   S  e  p  -   O  c   t   ) 50m 100m 5N 10N 15N 20N 25N  50m 100m     1   9   9   8   (   A  p  r  -   M  a  y   ) Chl- a  [mg m -3 ] 0.04 0.09 0. 5 1 5 1998  5N 10N 15N 20N 25N  50m 100m     1   9   9   8   (   M  a  y  -   J  u  n   ) 5N 10N 15N 20N 25N     1   9   9   8   (   S  e  p  -   O  c   t   ) 50m 100m  1999  5N 10N 15N 20N 25N  50m 100m     1   9   9   9   (   M  a  y  -   J  u  n   ) No data 2000  5N 10N 15N 20N 25N  50m 100m     2   0   0   0   (   A  p  r  -   M  a  y   ) 5N 10N 15N 20N 25N     2   0   0   0   (   S  e  p   ) 50m 100m  Fig. 3. Vertical structure (latitude–depth) of seasonal (spring and autumn) Chl- a  (CTD fluorometer profiler) concentration observedduring AMT campaigns (1996–2000). A nearest neighbourhood interpolation method was implemented to fill the data gaps betweensampling stations (shown as vertical lines) in the latitudinal direction. Y. Pradhan et al. / Deep-Sea Research II 53 (2006) 1548–1559 1552
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks