Continental Shelf Research 31 (2011) 2012–2017

Contents lists available at SciVerse ScienceDirect

Continental Shelf Research

0278-43

doi:10.1

n Corr

E-m

journal homepage: www.elsevier.com/locate/csr

Research papers

Coastal and oceanic SST variability along the western Iberian Peninsula

F. Santos n, M. Gomez Gesteira, M. deCastro

EPphysLab (Environmental Physics Laboratory), Universidade de Vigo, Spain

a r t i c l e i n f o

Article history:

Received 30 December 2010

Received in revised form

28 July 2011

Accepted 10 October 2011Available online 28 October 2011

Keywords:

Sea surface temperature

Atlantic multidecadal oscillation

Upwelling

Thermohaline circulation

43/$ - see front matter & 2011 Elsevier Ltd. A

016/j.csr.2011.10.005

esponding author.

ail address: fsantos@uvigo.es (F. Santos).

a b s t r a c t

The inter-annual variability of the sea surface temperature (SST) was analyzed along the western

Iberian Peninsula in the region ranging from 9.5 1W to 21.5 1W and from 37.5 1N to 42.5 1N with a

spatial resolution of 11�11 from 1900 to 2008. Both coastal and oceanic SST showed an overall increase

with warming and cooling cycles similar to those observed in the North Atlantic region and in previous

regional studies. In addition, the evolution of coastal and ocean water has been observed to be different.

In general, ocean water is more affected by the different warming–cooling cycles than coastal water. In

spite of coast and ocean are highly influenced by global changes affecting the whole North Atlantic

region, near shore SST has been observed to be correlated with local wind regime, which is itself a

manifestation of the Eastern Atlantic (EA) teleconnection pattern.

& 2011 Elsevier Ltd. All rights reserved.

1. Introduction

The temperature of the sea surface water (SST) is a funda-mental parameter in the ocean-atmosphere heat exchange andhence in the climatic regulation. In addition, SST is influenced byclimatic, meteorological, hydrodynamic and topographic para-meters. During the last century, a great effort has been devotedto develop reliable SST series with global coverage, first by meansof measurements from voluntary observation ships, drifters andmoored buoys (Brohan et al., 2006; Smith et al., 2008) and then bymeans of satellite-derived data. Great efforts were also devoted tocorrect uncertainties in the SST data due to several factors as:changes in the ship routes after the opening of Panama and SuezCanals, sampling sparseness during the world wars, differences inwater collection and more recently, uncertainties due to thepresence of aerosols and clouds, which can cause a cool bias,and to the fact that satellite instruments record skin temperatureinstead of near-surface temperature (for a complete understand-ing of the different bias and the methods to correct them see:Kushnir, 1994; Folland and Parker, 1995; Kaplan et al., 1998;Smith and Reynolds, 2002, 2003, 2004, 2005; Worley et al., 2005;Kent and Berry, 2005; Kent and Challenor, 2006; Kent and Taylor,2006; Brohan et al., 2006; Smith et al., 2008).

Numerous studies have tried to quantify trends in SST, whichhave shown to be extremely dependent on spatial and temporalscales being possible to observe opposite trends when consider-ing different periods of time (Parker et al., 1994; Smith et al.,1994; Casey and Cornillon, 2001). Despite these differences, most

ll rights reserved.

of the studies carried out during the last decade concluded that aconsiderable global warming in SST has occurred over the lastcentury no matter the considered data set (Folland et al., 1984;Folland et al., 1992; Parker et al., 1994; Nicholls et al., 1996; Caseyand Cornillon, 2001). In addition, similar trends were observed inwind (Caires et al., 2003; Gillet and Thompson, 2003; Cheltonet al., 2004), cloud coverage (Wiley et al., 2002; Roderick andFarquhar, 2002) and humidity (Flohn et al., 1990).

Global warming is far from being uniform in time and space. Onthe one hand, global warming is not spatially uniformly distributedall over the world’s oceans since there are some regions where thewarming is faster or slower than the global average (Levitus et al.,2000; Palttridge and Woodruff, 1981). In particular, the AtlanticOcean contributes most to the increase of the heat content (Neremet al., 1999; Levitus et al., 2000; Strong et al., 2000). On the otherhand, according to the Intergovernmental Panel on Climate Change(2007), global SST time series show distinct warming-coolingperiods during the last century. In particular, some authors(Garcia-Soto et al., 2002; deCastro et al., 2009; Gomez-Gesteiraet al., 2011) have pointed out the existence of warming-coolingperiods at several locations in the North Atlantic, which reflect thechanges exhibited by the North Atlantic Ocean. Nevertheless, as faras know, the different response of coastal and ocean water toglobal warming-cooling cycles had not been considered in previousresearch.

Regional differences in the warming rate could be explained interms of local and remote forcing factors (Cole et al., 2000; Lemosand Pires, 2004; Ginzburg et al., 2004; Santos et al., 2005; Gomez-Gesteira et al., 2008; deCastro et al., 2008). Among the remotefactors, the Thermohaline Circulation (THC) highly influences SSTfeatures in the North Atlantic region carrying warm water fromthe tropics to northern latitudes. This circulation has been often

Fig. 1. Study area. SST points (þ) are placed on a 11�11 grid from 9.5 1W to

21.5 1W and from 42.5 1N to 37.5 1N. Circles (0) represent coastal (9.5 1W) an

oceanic (17.5 1W) reference SST points. Crosses represent the location of wind data

(located at 10.0 1W, 37.5 1N, 40.0 1N and 42.5 1N).

F. Santos et al. / Continental Shelf Research 31 (2011) 2012–2017 2013

stated as the main reason why Western Europe is so temperatecompared to the same latitude in Eastern America. THC can beanalyzed in terms of the Atlantic Multidecadal Oscillation index(AMO). AMO is a coherent pattern of multidecadal variability inSST centered on the North Atlantic Ocean with a cycle rangingfrom 35 to 80 years depending on the author (Delworth et al.,1993; Timmermann et al., 1998; Kerr, 2000; Dima and Lohmann,2007). Following Trenberth and Shea (2006), the magnitude of theAMO signal is modest; the range is less than 0.4 1C. The AMO hasbeen linked with the variability in Northeast Brazilian rainfall(Folland et al., 2001), North American climate (Sutton andHodson, 2005) and U.S rainfall and river flows (Enfield et al.,2001). In addition, the AMO also affects the number of Atlantichurricanes and the tropical storms (Goldenberng et al., 2001;Trenberth and Shea, 2006). Delworth and Mann (2000), suggesteda link between the AMO and the variability of the THC as themean THC transports sufficient heat northward (Ganachaud andWunsch, 2000) to warm the Northern Hemisphere by severaldegrees (Vellinga and Wood, 2002). More recently, (Knight et al.,2005), by means of a 1400 year simulation with the HADCM3climate model (Gordon et al., 2000), were able to simulate theobserved AMO pattern and amplitude from measurements datingback to the nineteenth century. The results imply that the AMO isa genuine quasi-periodic cycle of internal climate variabilitypersisting for many centuries, and is related to variability in theoceanic THC.

Upwelling forcing on SST is possibly the most importantoceanographic feature in the so called EBUEs (Eastern BoundaryUpwelling Ecosystems) since it involves the replacement ofwarmer surface water by cooler subsurface water. Even, accord-ing to some authors, changes in the thermal gradient betweenland and ocean can be responsible of changes observed inupwelling intensity (Bakun, 1990; Mendelssohn and Schwing,2002; McGregor et al., 2007).

The western coast of the Iberian Peninsula (37 1N to 43 1N) maybe regarded as the northern boundary of influence of a broaderupwelling system (Eastern North Atlantic Upwelling System), thatacts all along the northwest coast of Africa and the Atlantic coast ofthe Iberian Peninsula (Nykjaer and Van Camp, 1994; Santos et al.,2005; Alvarez et al., 2008a). These previous studies have shownthat upwelling is mainly a seasonal event that occurs with higherprobability from April to September.

The aim of the present study is to describe the differences inSST evolution during the last century at coastal and oceaniclocations along western Iberian Peninsula. The SST variability willbe analyzed in terms of upwelling and THC intensity.

2. Data and processing

SST was obtained from the UK Meteorological office, HadleyCenter HadISST1.1-Global sea-Ice coverage and SST (http://badc.nerc.ac.uk/data/hadisst) (Rayner et al., 2003). Data are availablefrom 1870 to nowadays, with monthly periodicity on a 11�11grid with global coverage. In the present study, 78 data points infront of the western Iberian Peninsula (WIP) coast were consid-ered from 1900 to 2008. These points range from 9 1W to 21.5 1Wand from 37.5 1N to 42.5 1N, (Fig. 1). Monthly SST data wereseasonally and annually averaged.

The SST difference between coast and ocean was calculated as:

DSST¼ SSTocean�SSTcoast ð1Þ

Twelve points were considered for this purpose. Points locatedat 9.5 1W (17.5 1W) are representative of coastal (oceanic) condi-tions (Fig. 1, circles). Differences were calculated between each

pair of points located at the same latitude and then meridionalyaveraged.

The AMO index, which characterizes the large-scale pattern ofmultidecadal variability of SST, was calculated as the SST anomalyaveraged for the North Atlantic region. This index has beentraditionally calculated as the average of SST anomaly for theNorth Atlantic north of the equator (Enfield et al., 2001). Forpractical purposes the grid used in the present study covers from7.5 1W to 75.5 1W and from 0 to 59.5 1N. The SST northern limitwas kept at 59.5 1N to avoid problems with sea ice changes.

Wind data were obtained from the National Center of Atmo-spheric Research/National Center for Environmental Prediction(NCEP/NCAR) (http://www.esrl.noaa.gov/psd/data/reanalysis/reanalysis.shtml). Reanalysis Data Archive are available from 1948onwards, with a global coverage and a spatial resolution of2.5�2.51. In this study, points located in front of the WIP at10 1W and 42.5 1N, 40 1N and 37.5 1N, respectively, were consid-ered from 1948 to 2008 (Fig. 1, crosses). Time series obtained atthe different points showed to be well correlated (R40.8) allow-ing a meridional average. The upwelling index (UI) can be definedas minus the zonal component of Ekman transport. Note that theshore line is macroscopically perpendicular to the Equator alongthe WIP:

UI¼�Qx ¼�ty

rwfð2Þ

where

ty ¼ raCdðW2xþW2

y Þ1=2Wy ð3Þ

being ty the meridional wind stress, W the wind speed nearsurface, rw the sea water density (1025 kg m�3), Cd a dimension-less drag coefficient, (1.4�10�3), ra the air density (1.22 kg m�3)and f is the Coriolis parameter defined as twice the verticalcomponent of the Earth’s angular velocity, O, about the localvertical given by f ¼ 2OsinðyÞ at latitude y. Finally, x subscriptcorresponds to the zonal component and the y subscript to themeridional one. Negative (positive) ty values result in positive(negative) UI values, which correspond to upwelling favorable(unfavorable) conditions.

According to previous research (deCastro et al., 2008), theEastern Atlantic mode (EA) shows a significant negative correla-tion with upwelling along the entire western coast of the IberianPeninsula. The EA pattern consists of a north–south dipole thatspans the entire North Atlantic Ocean with centers near 55 1N, 20

F. Santos et al. / Continental Shelf Research 31 (2011) 2012–20172014

to 35 1W and 25 to 35 1N, 0 to 10 1W. This pattern, which is thesecond prominent mode of low-frequency variability overthe North Atlantic, is structurally similar to the NAO, but withthe anomaly centers displaced south-eastward to the approxi-mate nodal lines of the NAO pattern. The North Atlantic Oscilla-tion pattern (NAO), which is the first prominent mode over theNorth Atlantic, just has influence on the upwelling index between38 and 41 1N. Thus, EA will be the only mode considered in thepresent study. The teleconnection indices were obtained from theClimate Prediction Center (CPC) at the NCEP (www.cpc.noaa.gov)at monthly time scales from 1950 to 2009. Rotated principalcomponent analysis (RPCA) was used to identify the NorthernHemisphere teleconnection patterns and indices (Barnston andLivezey, 1987). This procedure isolates the primary teleconnec-tion patterns for all months and allows time series of the patternsto be constructed.

3. Results and discussion

The spatial and temporal distribution of SST along the studyarea is depicted in Fig. 2a. A running average of 75 years wasused to smooth high frequency variations in temperature. A SSTgradient on the order of 1 1C can be observed from coast(�9.5 1W) to ocean, with lower values near coast. In addition,several warming cooling cycles can be observed at all longitudesduring the period 1900–2008. Maximum values were observed

Fig. 2. (a) Spatial and temporal distribution of SST along the study area for the

period 1900–2008. (b) Inter-annual variation of the meridional average of coastal

(solid line) and oceanic (dashed line) SST (1C). A running average of 75 years was

considered in both frames to highlight warming–cooling periods.

around 1950 and at present and minimum values around 1900and 1975. The inter-annual evolution of the meridional average ofcoastal (at 9.5 1W) and oceanic (at 17.5 1W) SST are represented inFig. 2b. A running average of 75 years was also considered.Macroscopically, there is an overall increase of SST at bothlongitudes during this period (approximately 1.1 1C both atcoastal and ocean locations). This increase is comparable to theone obtained measured at adjacent areas (Garcia-Soto et al., 2002;deCastro et al., 2009; Gomez-Gesteira et al., 2011), and also to thevalues provided by the IPCC for the whole Atlantic Basin (IPCC,2007). In addition, coastal water (solid line) is cooler than theoceanic one (dashed line) as mentioned above. According toprevious research, this temperature gradient is mainly due tocoastal upwelling in spring–summer (Nykjaer and Van Camp,1994; Santos et al., 2005; Alvarez et al., 2008a) and to watercooling developed in shallow waters at the end of autumn due tonet heat loss from surface (Fiuza, 1983; desChamps et al., 1984).As mentioned above, three different warming-cooling periods canbe observed in both signals, a warming period from 1920 to 1950,a cooling period from 1950 to 1974 and another warming periodfrom 1974 to 2008. In the first warming period, oceanic (coastal)SST shows an increase of 0.26 (0.18) 1C per decade and in thesecond one shows an increase of 0.30 (0.28) 1C per decade. Thelast warming period is more intense than the previous one bothnear coast and in the ocean. In addition, the cooling period showsa decrease of �0.33 (�0.21) 1C per decade. Similar warming-cooling cycles had been observed at regional seas in the NorthAtlantic area as described by Garcia-Soto et al. (2002) in the Celticshelf, deCastro et al. (2009) in the Bay of Biscay and Gomez-Gesteira et al. (2011) in the area close to the Northwestern cornerof the Iberian Peninsula. It is worth noting that both cooling andwarming tend to be more intense at open ocean locations thannear coast. This effect is better depicted by means of the inter-annual variability of the meridional average of the DSST (1C)described in Eq. (1) (Fig. 3). A running average of 75 years wasalso applied to smooth the signal. The DSST parameter showsthree different periods that were highlighted with a straight line:two increasing periods (from 1925 to 1956 and from 1987 to2008) and one decreasing period (from 1956 to 1987). Theincrease and decrease rates for the different periods are shownin Table 1. Note that the intervals where the thermal gradientocean-coast is more (less) intense do not exactly coincide with

Fig. 3. Inter-annual evolution of the meridional average of DSST (1C) from 1900 to

2008. DSST was defined as ocean SST minus coastal SST as described in Eq. (1). A

running average of 75 years was considered to smooth the signal. Straight lines

were used to highlight the increasing and decreasing periods of DSST.

Table 1

Annual DSST (1C per decade) from 1900 to 2008.

DSST was meridionally averaged along the WIP. All

values have a significance level of 99% in the t-test.

Period Slope

1925–1956 �0.13

1956–1987 0.16

1987–2008 �0.1

Fig. 4. (a) Evolution of AMO index (gray solid line) compared with the mean SST

anomaly (SSTa) (black solid line) calculated at the SST points depicted in Fig. 1.

Dashed lines correspond to SSTa72s (SSTa). Once again, a running average of 75

years was considered. (b) Correlation coefficient between the annual AMO index

and the annual meridional average of SST at different longitudes calculated for the

period 1900–2008. Both signals were filtered with a running average (75 years)

prior to the calculation of correlation coefficients. Correlation is significant at 99%

for all longitudes.

Fig. 5. Correlation coefficients were calculated between seasonal (April–Septem-

ber) meridional wind stress (ty) and AMJJAS SST at different longitudes for the

period 1948–2008. Once again, signals were filtered with a running average (75

years) prior to the calculation of correlation coefficients. Only the three first

locations (until 12 1W) are statistically significant at 90%.

F. Santos et al. / Continental Shelf Research 31 (2011) 2012–2017 2015

the warming (cooling) cycles shown in Fig. 2b. Actually, they aredelayed in about 5–10 years. Although this effect can be partiallydue to the use of a running average of 75 years, the different timescales of the involved phenomena, global changes in temperatureand coastal upwelling, can also play a non negligible role.

As we mentioned above, AMO index is a good indicator of SSTchanges in the North Atlantic mainly related to THC. Fig. 4a showsthe inter-annual variability of AMO (solid gray line) and SSTanomaly SSTa (black solid line) averaged at the 78 locationsshown in Fig. 1. The dashed lines correspond to SSTa72s (SSTa).

Both signals, which were previously, smoothed by means of a 75year running average, show a similar pattern with comparablewarming-cooling cycles. Overall, the AMO warming for the period1900–2008 is on the order of 0.8 1C, which is slightly lower thanobserved along the WIP (1.1 1C as mentioned above). The appar-ent similarity between AMO and SSTa can be shown to bedependent on longitude (Fig. 4b). In spite of the correlation issignificant at 99% for all longitudes, the correlation coefficientdecreases rapidly from values close to 0.96 at open oceanlocations (21.5 1W) to 0.88 near coast (9.5 1W). The decrease ofthe correlation coefficient is observed to be more sharply nearcoast, which seems to indicate the existence of a local mechanismaffecting SST, apart from the global warming-cooling of the oceandue to changes in the THC. As we mentioned above, changes inwind patterns can be related to changes in SST. According toprevious research carried out along the WIP (Nykjaer and VanCamp, 1994; Alvarez et al., 2008a; Gomez-Gesteira et al., 2006;Alvarez et al., 2008b) there is clear prevalence of northerly(southerly) wind along the dry (wet) season. During the wetseason (ONDJFM) wind tends to drive air from southern latitudes,which is warmer than local air. Thus, the meriodional componentof wind stress (ty) tends to be positively correlated with SST bothnear coast and at open sea locations. Actually, the correlationcoefficient between ty calculated at 10 1W and SST for everylongitude from 9.5 1W to 21.5 1W is approximately equal to 0.75(significant at 95%) with a negligible dependence on longitude. Onthe contrary, during the dry season (AMJJAS), northerly windstend to generate coastal upwelling. Fig. 5 shows the correlationcoefficient between ty calculated at 10 1W and SST for longitudesranging from 9.5 1W to 21.5 1W for the period 1949–2008.Significant correlation (R�0.55) at a 90% level in the t-test is onlyobserved near coast with a sharp decrease seaward, reachingvalues below 0.3 for longitudes over 15.5 1W. ty and near shoreSST are positively correlated since when ty becomes morenegative (more intense northerly winds) coastal SST decreasesdue to the enhancement of coastal upwelling, which replacessurface water with cooler water. A similar conclusion wasattained by Schwing and Mendelssohn (1997) who comparedSST and wind stress in the California Current System.

The time evolution of upwelling and DSST signals during thedry season can be observed in Fig. 6. The anomaly of both signalsrelative to the period 1948–2008 was previously calculated to

Fig. 6. Time evolution of seasonal (April–September) UI (full bars), EA index

(empty bars) and DSST (dark line) anomalies relatives to the period 1948–2008.

The signals were filtered by means of a running average (75 years). In addition,

the UI anomaly was multiplied by a factor 0.75/(max(UI)�min(UI)) to allow a

better visual comparison between signals. UI showed to be correlated with DSST

(R¼0.67, significant at 90%) and negatively correlated with EA (R¼�0.62,

significant at 90%).

F. Santos et al. / Continental Shelf Research 31 (2011) 2012–20172016

allow visual comparison. In addition, the high frequency wasfiltered by means of a running average (75 years). Macroscopi-cally, periods with UI values (full bars) over the mean tend tocoincide with periods with DSST over the mean (black line). Notethat high DSST values correspond to large temperature differencesbetween ocean and coastal water. Actually, the correlation coeffi-cient between both signals (R¼0.67) is significant at 90%. Accord-ing to deCastro et al. (2008), changes in upwelling intensity arerelated to changes in the teleconnection patterns, in particular inEA pattern. The empty bars in Fig. 6 represent the time evolution ofEA, which showed to be negatively correlated with UI (R¼�0.62,significant at 90%). Thus the observed changes in UI are not anisolated coastal fact, but they can be related to changes in EApattern, which explain most of the upwelling variability.

Schwing and Mendelssohn (1997) found upwelling intensifi-cation coinciding with warmer SST in the California CurrentSystem. There the long-term SST trend masks the cooling effectdue to upwelling increase. Remarkable differences have beenfound between that manuscript and the present analysis. First,DSST was used in order to remove the global component from thesignal. Thus, the warming-cooling cycles affecting the NorthAtlantic are considered to affect in a similar way to both coastaland ocean locations. Then, the temperature gradient (DSST) isassumed to be mainly due to local forcing. Second, the observedtrends do not indicate increase in seasonal upwelling as sug-gested by Bakun (1990) who hypothesized the existence of nearcoast SST cooling via increased coastal upwelling generated bythe different warming rates over land and sea (Jones et al., 2001;Sutton et al., 2007). No evidence of such a mechanism was foundin the present study. Actually, most of the studies carried outduring the last decades have shown a decreasing trend inupwelling intensity. Thus, Lemos and Pires (2004) found evidenceof a progressive weakening of the upwelling regime for the period1941–2000, Alvarez et al. (2008a) did not observe a clear trend atmonthly scale, although, in average, upwelling tended to decreasefor the period 1967–2006. A remarkable decrease in upwellingindex was also described by Perez et al. (2010) at the north-western corner of the Iberian Peninsula for the period 1965–2007.The same authors (Pardo et al., 2011) extended their analysis tothe period 1948–2009 with similar conclusions. Gomez-Gesteira

et al. (2011) found a significant decrease in upwelling index forthe region Galicia–North Portugal over the period 1975–2008.Only Santos et al. (2011) hypothesized the existence of differentperiodicities in upwelling evolution without a clear increasing ordecreasing trend.

Finally, it is unclear if the correlation between SST and windstress is due to a causal physical process or simply they covarywith other processes at a global scale, since the observedcommonality in trends does not necessarily imply the existenceof an underlying dynamical link as pointed out by Mendelssohnand Schwing (2002). Possibly, a nonlinear interaction betweenchanges in ocean temperature and atmospheric patterns is behindthe observed evolution of SST and UI, although new researchshould be conducted to elucidate the underlying mechanism.

4. Concluding remarks

Different warming-cooling cycles in SST had been observed inthe North Atlantic Region during the last century, although thedifferent response of ocean and coastal waters had not beendescribed so far. New findings about the differences in coastal andocean SST trends along the western Iberian Peninsula can bewithdrawn from the present research:

–

Ocean and coastal SST evolve at different rates. Actually, thedifference in SST (ocean minus coast) tends to increase(decrease) during warming (cooling) periods.

–

Both coastal and oceanic SST is highly correlated with theAtlantic Multidecadal Oscillation (AMO), although the correla-tion was observed to decrease coastward.

–

Near coast SST is also correlated with meridional wind stress(ty) during the dry season. This ty characterizes coastalupwelling, which is the main feature of the western IberianPeninsula coast during spring and summer.

Acknowledgements

This work is supported by the Xunta de Galicia under theproject 10PXIB 383169PR.

References

Alvarez, I., Gomez-Gesteira, M., deCastro, M., Dias, J.M., 2008a. Spatiotemporalevolution of upwelling regime along the western coast of the IberianPeninsula. Journal of Geophysical Research, 113. doi:10.1029/2008JC004744.

Alvarez, I., Gomez-Gesteira, M., deCastro, M., Novoa, E.M., 2008b. Ekman transportalong the Galician Coast (NW, Spain) calculated from QuikSCAT winds. Journalof Marine Systems 72, 101–115.

Bakun, A., 1990. Global climate change and intensification of coastal upwelling.Science 247, 198–201. doi:10.1126/science.247.4939.198.

Barnston, A.G., Livezey, R.E., 1987. Classification, seasonality and persistence oflow-frequency atmospheric circulation patterns. Monthly Weather Review115, 1083–1126.

Brohan, P., Kennedy, J.J., Harris, I., Tett, S.F.B., Jones, P.D., 2006. Uncertaintyestimates in regional and global observed temperature changes: a new dataset from 1850. Journal of Geophysical Research, 111.

Caires, S., Sterl, A., Komen, G., Swail, V., 2003. Global Wave Climatology Atlas.Royal Netherlands Meteorological Institute (KNMI) /http://www.knmi.nl/waveatlasS.

Casey, K.S., Cornillon, P., 2001. Global and regional sea-surface temperature trends.Journal of Climate 14, 3801–3818.

Chelton, D.B., Schlax, M.G., Freilich, M.H., Milliff, R.F., 2004. Satellite measurementsreveal persistent small-scale features in ocean winds. Science 303, 978–998.

Cole, J.E., Dunbar, R.B., McClanahan, T.R., Muthiga, N.A., 2000. Tropical Pacificforcing of decadal SST variability in the western Indian Ocean over the pasttwo centuries. Science 287, 617–619.

deCastro, M., Gomez-Gesteira, M., Lorenzo, M.N., Alvarez, I., Crespo, A.J.C., 2008.Influence of atmospheric modes on coastal upwelling along the western coastof the Iberian Peninsula, 1985 to 2005. Climate Research 36, 169–179.

F. Santos et al. / Continental Shelf Research 31 (2011) 2012–2017 2017

deCastro, M., Gomez-Gesteira, M., Alvarez, I., Gesteira, J.L.G., 2009. Presentwarming within the context of cooling–warming cycles observed since 1854in the Bay of Biscay. Continental Shelf Research 29, 1053–1059.

Delworth, T.L., Manabe, S., Stouffer, R.J., 1993. Interdecadal variability of thethermohaline circulation in a couple ocean-atmosphere model. Journal ofClimate 6, 1993–2001.

Delworth, T.L., Mann, M.E., 2000. Observed and simulated multidecadal variabilityin the Northen Hemisphere. Climate Dynamics 16, 661–676.

desChamps, P.Y., Frouin, R., Crepon, M., 1984. Sea surface temperatures of thecoastal ones of France observed by the HCMM satellite. Journal of GeophysicalResearch 89, 8123–8149.

Dima, M., Lohmann, G., 2007. A hemispheric mechanism for the Atlantic Multi-decadal Oscillation. Journal of Climate 20, 2706–2719.

Enfield, D.B., Mestas-Nunez, A.M., Trimble, P.J., 2001. The Atlantic multidecadaloscillation and its relation to rainfall and river flows in the continental U.S.Geophysical Research Letters 28 (10), 2077–2080.

Fiuza, A.F.G., 1983. Upwelling patterns off Portugal. In: Suess, E., Thiede, J. (Eds.),Coastal Upwelling. Its Sediment Records (Part A), Plenum, New York, pp. 85–98.

Flohn, H., Kapala, A., Knoche, H.R., Machel, H., 1990. Recent changes of the tropicalwater and energy budget and of midlatitude circulations. Climate Dynamics 4,237–252.

Folland, C., Parker, D., Kates, F., 1984. Worldwide marine temperature fluctuations1856–1981. Nature 310, 670–673.

Folland, C., Nicholls, N., Nyenzi, B., Parker, B., Vinnikov, K., 1992. Observed climatevariability and change. In: Houghton, J., Jenkins, G., Ephraums, J. (Eds.), ClimateChange 1992: The Supplementary Report to the IPCC Scientific Assessment,Cambridge University Press, pp. 135–170.

Folland, C., Parker, D., 1995. Correction of instrumental biases in historical seasurface temperature data. Quarterly Journal of the Royal MeteorologicalSociety 121, 319–367.

Folland, C., Colman, A.W., Rowell, D.P., Davey, M.K., 2001. Predictability of north-east Brazil rainfall and real-time forecast skill, 1987–1998. Journal of Climate14, 1937–1958.

Garcia-Soto, C., Pingree, R.D., Valdes, L., 2002. Navidad Development in theSouthern Bay of Biscay: climate change and Swoddy structure from remotesensing and in situ measurements. Journal of Geophysical Research 107 (C8),3118. doi:10.1029/2001JC001012.

Ganachaud, A., Wunsch, C., 2000. Improved estimates of global ocean circulation,heat transport and mixing from hydrographic data. Nature 408, 453–457.

Gillet, N.P., Thompson, D.W.J., 2003. Simulation of recent southern hemisphereclimate change. Science 302, 273–275.

Ginzburg, A.I., Kostianoy, A.G., Sheremet, N.A., 2004. Seasonal and interanualvariability of the Black Sea surface temperature as revealed from satellite data(1982–2000). Journal of Marine Systems 52, 33–50.

Goldenberng, S.B., Landsea, C.W., Mestas-Nunez, A.M., Gray, W.M., 2001. Therecent increase in Atlantic hurricane activity: causes and implications. Science293, 474–479.

Gomez-Gesteira, M., Moreira, C., Alvarez, I., deCastro, M., 2006. Ekman transportalong the Galician coast (NW, Spain) calculated from forecasted winds. Journalof Geophysical Research 111, C10005. doi:10.1029/2005JC003331.0.

Gomez-Gesteira, M., deCastro, M., Alvarez, I., Gomez-Gesteira, J.L., 2008. CoastalSea surface temperature warming trend along the continental part of AtlanticArc (1985–2005). Journal of Geophysical Research 113, C04010. doi:10.1029/2007JC004315.

Gomez-Gesteira, M., Gimeno, L., deCastro, M., Lorenzo, M.N., Alvarez, I., Nieto, R.,Taboada, J.J., Crespo, A.J.C., Ramos, A.M., Iglesias, I., Gomez- Gesteira, J.L., Santo,F.E., Barriopedro, D., Trigo, I.F., 2011. The state of climate in North–West Iberia.Climate Research. doi:10.3354/cr00967.

Gordon, C., Cooper, C., Senior, C.A., Banks, H., Gregory, J.M., Johns, T.C., Mitchell,J.F.B., Wood, R.A., 2000. The simulation of SST, sea ice extents and ocean heattransports in a version of the Hadley Centre coupled model without fluxadjustments. Climate Dynamics 16, 147–168.

Intergovernmental Panel on Climate Change, 2007. Climate Change 2007: ThePhysical Science Basis. Contribution of Working Group 1 to the FourthAssessment Report of the Intergovernmental Panel on Climate Change CambridgeUniversity Press, Cambridge, UK.

Jones, P.D., Osborn, T.J., Briffa, K.R., Folland, C.K., Horton, E.B., Alexander, L.V.,Parker, D.E., Rayner, N.A., 2001. Adjusting for sampling density in grid box landand ocean surface temperature time series. Journal of Geophysical Research106, 3371–3380.

Kaplan, A., Cane, M.A., Kushnir, Y., Clement, A.C., Blumenthal, M.B., Rajagopalan, B.,1998. Analyses of global sea surface temperature 1850–1991. Journal ofGeophysical Research 103, 18567–18589.

Kent, E., Berry, D.I., 2005. Quantifying random measurement errors in voluntaryobserving ships’ meteorological observations. International Journal of Clima-tology 25, 843–856.

Kent, E., Challenor, P.G., 2006. Toward estimating climatic trends in SST. Part II:random errors. Journal of Atmospheric and Oceanic Technology 23, 476–486.

Kent, E., Taylor, P.K., 2006. Toward estimating climatic trends in SST. Part I: methodsof measurements. Journal of Atmospheric and Oceanic Technology 23, 464–475.

Kerr, R.A., 2000. A North Atlantic climate pacemaker for the centuries. Science 288(5473), 1984–1986.

Knight, J.R., Allan, R.J., Folland, C.K., Vellinga, M., Mann, M.E., 2005. A signature ofpersistent natural thermohaline circulation cycles in observed climate. Geo-physical Research Letters 32, L20708.

Kushnir, Y., 1994. Interdecadal Variations in North Atlantic Sea Surface Tempera-ture and Associated Atmospheric Conditions American Meteorological Society(pp. 141–157).

Lemos, R.T., Pires, H.O., 2004. The upwelling regime off the west Portuguese coast,1941–2000. International Journal of Climatology 24, 511–524. doi:10.1002/joc.1009.

Levitus, S., Antonov, J.I., Timothy, P.B., Stephens, C., 2000. Warming of the WorldOcean. Science 287, 2225–2229.

McGregor, H.V., Dima, M., Fischer, H.W., Mulitza, S., 2007. Rapid 20th-centuryincrease in coastal upwelling off northwest Africa. Science 315, 637–663.doi:10.1126/science.1134839.

Mendelssohn, R., Schwing, F.B., 2002. Common and uncommon trends in SST andwind stress in the California and Peru–Chile current Systems. Progress inOceanography 53, 141–162.

Nerem, R.S., Chambers, D.P., Leuliette, E.W., Mitchum, G.T., Giese, B.S., 1999.Variations in global mean sea level associated with the 1997–1998 ENSOevent: implications for measuring long-term sea level change. GeophysicalResearch Letters 26, 3005–3008.

Nicholls, N., Gruza, G.V., Jouzel, J., Karl, T., Ogallo, L., Parker, D., 1996. Observedclimate variability and change. In: Houghton, J., Jenkins, G., Ephraums, J. (Eds.),Climate Change 1995: The Science of Climate Change, Cambridge UniversityPress, pp. 133–192.

Nykjaer, L., Van Camp, L., 1994. Seasonal and interannual variability of coastalupwelling along northwest Africa and Portugal from 1981 to 1991. JournalGeophysical Research 99 (C7), 14197–14207.

Palttridge, G., Woodruff, S., 1981. Changes in global surface temperature from1880 to 1977 derived from historical records of sea surface temperature.Monthly Weather Review 109, 2427–2434.

Pardo, P.C., Padın, X.A., Gilcoto, M., Farina-Busto, L., Perez, F.F., 2011. Evolution ofupwelling systems coupled to the long term variability of sea surfacetemperature and Ekman transport. Climate Research. doi:10.3354/cr00989.

Parker, D.E., Jones, P., Folland, C., Bevan, A., 1994. Interdecadal changes of surfacetemperature since the late nineteenth century. Journal of GeophysicalResearch 99, 14339–14373.

Perez, F.F., Padin, X.A., Pazos, Y., Gilcoto, M., Cabanas, M., Pado, P.C., Doval, M.D., Farina-Bustos, L., 2010. Plankton response to weakening of the Iberian coastal upwelling.Global Change Biology 16, 1258–1267. doi:10.1111/j.1365-2486.2009.02125.x.

Rayner, N.A., Parker, D.E., Horton, E.B., Folland, C.K., Alexander, L.V., Rowell, D.P.,Kent, E.C., Kaplan, A., 2003. Global analyses of sea surface temperature, sea ice,and night marine air temperature since the late nineteenth century. JournalGeophysical Research, 108.

Roderick, R., Farquhar, G.D., 2002. The cause of decreased pan evaporation over thepast 50 years. Science 298, 1410–1411.

Santos, A.M.P., Kazmin, A.S., Peliz, A., 2005. Decadal changes in the Canaryupwelling system as revealed by satellite observations: their impact onproductivity. Journal of Marine Research 63, 359–379.

Santos, F., Gomez-Gesteira, M., de Castro, M., Alvarez, I., 2011. Upwellingvariability along the western coast of the Iberian Peninsula: can trends beaccurately calculated? Climate Research. doi:10.3354/cr00972.

Schwing, F.B., Mendelssohn, R., 1997. Increased coastal upwelling in the CaliforniaCurrent System. Journal Geophysical Research 102B, 3421–3438.

Smith, T.M., Reynolds, R.W., Ropelewski, C.F., 1994. Optimal averaging of seasonalsea-surface temperatures and associated confidence intervals (1860–1989).Journal of Climate 7, 949–964.

Smith, T.M., Reynolds, R.W., 2002. Bias corrections for historic sea-surfacetemperatures based on marine air temperatures. Journal of Climate 15, 73–87.

Smith, T.M., Reynolds, R.W., 2003. Extended reconstruction of global sea-surfacetemperatures based on COADS data (1854–1997). Journal of Climate 16,1495–1510.

Smith, T.M., Reynolds, R.W., 2004. Improved extended reconstruction of SST(1854–1997). Journal of Climate 17, 2466–2477.

Smith, T.M., Reynolds, R.W., 2005. A global merged land–air–sea-surface tem-perature reconstruction based on historical observations (1880–1997). Journalof Climate 18, 2021–2036.

Smith, T.M., Reynolds, R.W., Peterson, T.C., Lawrimore, J., 2008. Improvements toNOAAs historical merged land–ocean surface temperature analysis (1880–2006). Journal of Climate 21, 2283–2296.

Strong, A.E., Kearns, E.J., Gjovig, K.K., 2000. Sea surface temperature signals fromsatellite—an update. Geophysical Research Letters 27, 1667–1670.

Sutton, R.T., Hodson, D.L.R., 2005. Atlantic Ocean forcing of North American andEuropean summer climate. Science 290, 2133–2137.

Sutton, R.T., Dong, B., Gregory, J.M., 2007. Land/sea warming ratio in response toclimate change: IPCC AR4 model results and comparison with observations.Geophysical Research Letters 34, L02701. doi:10.1029/2006GL028164.

Timmermann, A., Latif, M., Voss, R., Grotzner, A., 1998. Northern hemispheric inter-decadal variability: a coupled air-sea mode. Journal of Climate 11, 1906–1931.

Trenberth, K.E., Shea, D.J., 2006. Atlantic hurricanes and natural variability in 2005.Geophysical Research letters 33, L12704.

Vellinga, M., Wood, R.A., 2002. Global climatic impacts of a collapse of the Atlanticthermohaline circulation. Climate Change 54, 251–267.

Wiley, D.P., Frey, R., Zhang, H., Menzel, W.P., 2002. Extending HIRS High CouldTrends with NODIS. NASA /http://modis.gsfc.nasa.gov/sci_team/meetings/200207/presentations/menzel.pdfS.

Worley, S.J., Woodruff, S.D., Reynolds, R.W., Lubker, S.J., Lott, N., 2005. ICOADSrelease 2.1 data and products. International Journal of Climatology 25,823–842.