Mesoscale studies in the Indian Ocean using altimeter ...

3 downloads 0 Views 2MB Size Report
periods of 40-50 days which are probably related to .... at 8 days intervals, ERS-l on a 0.5° 0.5° spatial ..... Luther, M. E., and J. J. o'Brian, A model of the sea.
Mesoscale studies in the Indian Ocean using altimeter data Vibeke E. Jensen, Pa ul Samuel, Ola M. Johannessen* Nansen Environmental and Remo te Sensing Center, Edvard Griegs vei 3a,N-5037 Solheimsviken, Norway; Phone +47-55 29 72 88, Fax +47-55 20 00 50, [email protected]

1989; Bruce et al., 1994; Schott et al., 1994; Evensen and van Leeuwen, 1995; Shetye et al., 1996]. The South West (SW) monsoon is normally observed over the Arabian Sea from May to August and the North East (NE) monsoon from October to January [ Indian Meteorological Society, 1994]. The surface circulation in the Indian Ocean is schematicly represented in Figure 1 for the two mon­ soon periods. During the NE monsoon there is a strong southward flowing current off the northeast African coast called the Somali Current. During the SW monsoon the westward flowing Southern Equatorial Current (SEC) feeds the Somali Current (SC) which then flows northward with speeds up to 200 cm/s [ Pickard and Emery, 1990],and turning off­ shore at 2°_3° S. The SC develops two anticyclonic gyres, the Great Whirl (GW) (centered at about 8° N) and the Southern Gyre (SG) (located in the area 0°-5° N, 53° E) [Perigaud and Minster, 1988; Moli­ nari et al., 1990; and Schott et al., 1994]. Significant mesoscale variability has also been ob­ served in the northern Indian Ocean and a number of eddies have been examined by previous investi­ gators. Satellite altimetry is particularly suited for studying mesoscale variability and has been used in the Southern Indian Ocean by investigators such as Wakker et al. (1990), Snaith and Robinson (1996) and Grundling (1995). Evensen and van Leeuwen (1995) assimilated Geosat altimeter data in an ocean circulation monsoon to study the shedding of Agul­ has rings. Further altimeter studies in the northern In­ dian Ocean have been done by Perigaud and Min­ ster (1988) who studied the variability of the So­ mali Current using Seasat altimeter data. Peri­ gaud and Delecluse [1992; 1993] studied the an­ nual/interannual variability of currents in the area using Geosat altimeter data. Bruce et al. (1994) studied the formation of a large anticyclonic eddy in the eastern Arabian Sea during the NE monsoon us­ ing Geosat altimeter data,and Greiner and Perigaud (1993) assimilated Geosat sea level variations into a nonlinear reduced gravity shallow-water model. However,so far the extensive data set from ERS-1 and Topex/Poseidon altimeters have not been used

ABSTRACT The oceanic variability in response to the monsoon system in the northern Indian Ocean was for the first time studied using the three altimeters Geosat,ERS1 and Topex/Poseidon. The Root Mean Square, computed from the sea surface height anomalies in­ dicates values of 15-17 cm in highly energetic ar­ eas such as the Somali Current, and 13-15 cm in the western part of the Bay of Bengal. As the most prominent variations occur,and because most in situ observations are from these areas the investigation was focused on these two areas. Complex Principal Component analysis was ap­ plied to the gridded SSH anomalies and the results indicate that the annual signal is strongest when the monsoon period is at its peak, and the semi-annual time series seems to be related to the onset and decay phase of the SW and NE monsoon. CPCA analysis of the Topex/Poseidon data indicates characteristic periods of 40-50 days which are probably related to earlier observations in the study area of mesoscale variability generated by the horizontal shear in the velocity field. Rossby waves are identified from the semi annual mode, and they are observed to have a westward propagating speed of 13 km/d at the equa­ tor, with decreasing speed away from the equator. Mesoscale eddies with diameters between 2501000 km and maximum amplitudes up to 45 cm are observed during the monsoon periods. The propa­ gation and life time of eddies in the study area are in good agreement with those reported in previous investigations using in situ data and numerical mod­ els. 1 INTRODUCTION The circulation patterns in the Indian Ocean,forced by the reversing southwest (May to August) and northeast (October to January) monsoonal winds, represent a unique oceanographic problem that has been the subject of a number of investigations [e.g. Cox,1976; Anderson and Rowlands,1976; Rao et al., * Also at Geophysical Institute, Norway.

University of Bergen,

1

dard repeat track analysis to obtain SSH anomalies [Cheney et al., 1991; Samuel, 1993]. For each data point, Root Mean Square (RMS) of the SSH were computed. The SSH anomalies are also interpolated to a regular space-time grid using a simple exponential weighting scheme [Samuel, 1993]. Geosat data are gridded on a 1.0° � 1.0° spatial grid at 8 days intervals, ERS-l on a 0.5° � 0.5° spatial grid at 15 days time intervals and Topex/Poseidon on a 2.0° � 2.0° spatial grid at 5 days intervals. These grid spacings and intervals were choses in order to properly present the spatial and temporal sampling characteristics of the different satellite missions. The spatial/temporal modes of the data sets were extracted using Complex Principal Compo­ nent Analysis (CPCA) on the gridded data sets. CPCA is a variant of Principal Component Anal­ ysis which is used to decompose the variability in the data sets into orthogonal modes, and is designed specifically to handle propagating features [Horel, 1984; Preisendorfer, 1988]. CPCA is done only for Topex/Poseidon and not for ERS-l since the dura­ tion is too short (one and a half years).

(a) SW monsoon (August)

(b) NE monsoon (January) Figure 1: Surface circulation during the SW and NE monsoon. Modified from Molinari. et al. (1990). So­ mali Current (SC), Great Whirl (GW), Indian Mon­ soon Current (IMC), Equatorial Jet (EJ), Southern Equatorial Current (SEC), Northern Equatorial Cur­ rent (NEC), Southern Equatorial Counter Current (SECC), East African Counter Current (EACC), Coastal Current (CC).

3 SPATIAL AND TEMPORAL VARIABILITY 3.1 Root Mean Square variability Root Mean Square (RMS) of the SSH provides an indication of the distribution of mesoscale variability in the study area. Over most of the study area the RMS variability is low, less than 7 cm (Figure 2). High RMS variability is observed in highly energetic areas such as the Somali Current area (15-17 cm), and along the western part of the Bay of Bengal (1315 cm). High variability is also observed along 10° S at 45°-50° E and at 75°-82° E (11-13 cm) and in the eastern part of the Bay of Bengal (9-11 cm). The high RMS values are about half of those ob­ served in highly energetic areas such as the Gulf Stream or the Agulhas, where similar analysis us­ ing altimeter data have shown RMS values up to 30 cm [Le Traon and Rouquet, 1990] and 32 cm [Wakker et al., 1990J. In the Nordic Sea, Samuel et al. (1994) observed values with a maximum of 15 cm. There is a close relationship between the areas of greatest SSH variability and the major currents, the Somali Current, East Indian Coastal Current [McCreary et al., 1996; Shankar et al., 1996J and Southern Equatorial Current, and higher variability is most prevalent in the western parts of the oceans. As an overall assessment the distribution of RMS height in the Indian Ocean seems to be consistent for the three data sets. In general the ERS-l shows more extended areas with high variability (9-11 cm), which is particularly seen over the Chagos Laccadive Ridge (north-south ridge, extending between 73°-

jointly in the Indian Ocean circulation. Therefore the main objective of our investigation was to gain an improved understanding of the oceanic variability in response to the monsoon system in the northern Indian Ocean using altimeter data. The altimeter data used and the method of pro­ cessing are described in Section 2. In Section 3 the details of the SSH variability field from the altimeters is examined, and the temporal and spatial character­ istics obtained from Complex Principal Component Analysis (CPCA) are presented. In Section 4 results pertaining to the propagation and life time of ed­ dies are presented, and in Section 5 a discusion and conclusion is given.

2 ALTIMETER DATA PROCESSING The altimeter data used in the study are from the Geosat (17 days repeat period, 01.11.86 - 01.09.89), the ERS-l (35 days repeat period 01.04.92 31.12.93) and the Topex/Poseidon (10 days repeat period, 01.01.93 - 10.08.95) missions. The first 10 cycles of Topex/Poseidon data are not used in this study, because of attitude problems with the satellite [AVISO, 1992]. The data were processed using stan-

2

major portion of the variability will be contained in the first modes, and therefore only these modes need to be explained. In our case, the first two modes of the Geosat and Topex/Poseidon data set accounted for 28.3 % and 41.0 % respectively, and these will be dis­ cussed. Figure 3 shows the phase and the amplitude time series of the first two modes for Geosat and Topex/Poseidon. Dominant modes for both the data sets have a clear annual period as seen in the phase plots Figure 3a and 3c. The phase plot for the Geosat first mode shows a clear annual signal while for Topex/Poseidon there is some leakage from shorter periods. Both the data sets indicate that the annual signal has max­ ima around January and July, with small variations from year to year, possibly related to variability in the onset of the monsoon. The amplitude time series (Figure 3b and 3d) indi­ cate when the annual signal is strongest. For Geosat the strongest is January 1989 and also reluctivly strong in January 1987 and 1988, May 1988 and 1989. The amplitude time series for Topex/Poseidon mode one (Figure 3d) suggests that this mode actu­ ally contains two signals, the annual signal and a signal with a period of approximately two months. As mentioned earlier, this is also evident in the cor­ responding phase plot (Figure 3c). The fact that the two signals are combined in one mode suggests that they are correlated. Several investigators have ob­ served oscillations with a 40-60 days period in the western Indian Ocean using in sit'lL data and numer­ ical models [Mysak and Mertz, 1984; Schott et al., 1988; Swallow et al., 1988; Woodberry et al., 1989] and explain these oscillations as caused by horizontal shear (barotropic instability) which generate eddies at a period of 40-50 days. This signal is possibly not resolved by the 17 days repeat period of Geosat, explaining the absence from Figure 3a and 3b. The phase time series of the second mode for Geosat (Figure 3e) show a clear semi-annual period with maxima in May, August and November, Febru­ ary. This mode therefore appears to be related to the onset and decay phase of the SW and NE mon­ soons. For the Topex/Poseidon phase no clear semi annual period is observed (Figure 3g). There is a strong indication of a half yearly period from May'93 to November'93, but for the rest of the sampling pe­ riod, another phase with a period of approximately 2 months similar to that above, overrides the semi­ annual signal. The amplitude (Figure 3h) show that the period when the semi-annual phase is observed is weeker than the 2 months period observed. In order to examine the significance of the peaks and lows in the temporal characteristics, and to de­ scribe the varying physical attributes of propagating features in traveling wave fields, a truncated data set

'ON

'ON

'0'

(a) Geosat 20N

'ON

'0'

(b) ERS-l 20N

'ON

,os

(c) Topex/Poseidon Figure 2: RMS Sea Surface height anomaly distribu­ tion. Units in em. 78° E). Other areas with high RMS distribution are around 78° E, 10° S where the Southern Equatorial Current is found and around 47° E, 10° S which is in the area of the East African Counter Current . Geosat indicates values of 11-13 cm in these areas, while the other two altimeters show values of 7-9 cm. In the eastern part of the Bay of Bengal values of 1315 cm are shown for all three altimeters. 3.2 CPCA Complex principle Component Analysis (CPCA) de­ composes the data set into a number of orthogonal modes, each of which are represented by a complex time series and complex spatial eigenvector. The time series indicate the temporal evolution of the spatial averaged variability, while the spatial pat­ terns (eigenvectors) indicate the distribution of tem­ poraly averaged variability. The modes are ordered according to the associated eigenvalues which give an indication of the fraction of the total variability that is explained by the respective modes. Usually 3

( a)

Phase, Geosat mode 1

( e)

Phase T / P mode 1

(�) Amp.

(f )

Geosat mode 1

(c)

Phase, Geosat mode 1

...-

Amp. T / P mode 1

( g ) Phase T /P mode 2

( d) Amp.

(h)

Geosat model

Amp. T / P mode 2

Figure 3: Temporal characteristics for first and second Geosat and Topex / Poseidon SSH CPCA mode. The amplitude is in cm, and the phase is from -7r to +7r. may be reconstructed using the first two modes. The results will be discussed in section 3.3.

low water simulation. Perigaud and Delecluse, 1993 observed Rossby waves which takes from 4 months to cross the basin at equator, and have a speed up to 12 km / d along 15° S. Although Topex / Poseidon accounts for more variability than the two Geosat modes, the propagating waves are easier observed using Geosat altimeter data, mainly due to a better spatial sampling. The same features where also observed from the Topex / Poseidon data ( not shown here ) with the same characteristics.

3.3 Large scale features. A reconstruction using the annual and semi-annual mode for Geosat and Topex / Poseidon was done to describe the varying physical attributes of propagat­ ing features. Only the anomaly fields during the NE monsoon in 1988 for Geosat ( Figure 4 ) are shown, since the propagation is most easily seen here. The annual mode shows a feature, ( A ) which is ob­ served in the Somali Current area ( Figure 4 ) during the 1987 / 88 NE monsoon from the Geosat data. No clear evidence of direction of propagation is observed. In the Bay of Bengal a rotational movement ( B ) is observed, but it is not possible to determine the di­ rection, but we assume that this is an anticyclonic gyre due to analysis of results from a multi-layer, adiabatic model, driven by climatological monthly mean winds winds [Potemra et ai., 1991] and ship drift data [Defant, 1961;Cutler and Swallow, 1984] which have observed a large anticyclonic flow in the Bay of Bengal. For the semi-annual mode it appears that the main phenomena is Rossby waves. Our investigation iden­ tifies Rossby waves along 7° N ( C ) ,along equator ( D ) and between 5° S and the equator ( E ) . The westward propagating waves have a speed up to 6 km / d along 7° N ( C ) , up to 13 km / d along equator ( D ) and 7 km / d between 5° S and the equator ( E ) , indicating a decrease in speed away from the equator. Our re­ sults show generally lower speeds than Perigaud and Delecluse, 1993 interannual sea level studies, using Geosat data based on crossover analysis and shal-

4

MESOSCALE EDDIES

Closer inspection of the gridded SSH anomaly data set showed evidence of several mesoscale eddies. In this section some of the observed eddies will be ex­ amined. Figure 5 shows the location and rotation of the most persistent eddies observed from the three data sets. We observed eddies to be more domi­ nant during the monsoon periods, and the two areas, the Somali Current area in the Arabian Sea ( 30°61° E, 0°-20° N ) and the western part of the Bay of Bengal ( 75°-95° E, 0°-20° N ) examined, indicate eddies with a maximum amplitude up to 45 cm in July / August and December / January, with a spatial scale ranging between 250-1000 km. The typical life cycle of these eddies is illustrated by examining con­ secutive time steps of the gridded Geosat,ERS-1 and Topex / Poseidon data in the area of the Somali Cur­ rent. The growth of an anticyclonic eddy E1,located ap­ proximately at 8.5° N, 53° E, is observed from June to September 1989 from the data sets. This eddy is most intense during the first half of August,with am4

"

• ...... .

__

A

.

�!!!!-



.. I·

...

.. .

�.

�.

�.

... --

... -

( a ) November 19, 1987

... --

( b ) December 1, 1987

_. t· .. I· •

_. . . I· •

.

0

.. ..

�...

... -

( c ) December 13, 1987

_. 1- " I· •

0

.. . .

( d ) December 25, 1987

.. I·

.. .

.. ..

� ...

... --



( e ) January 6, 1988

� ...

... -





( f ) January 18, 1988

Figure 4: SSB anomaly field reconstructed from the first and second Geosat CPCA mode during the NE monsoon. plitudes of 30-35 cm ( some of the individual passes over this eddy show amplitudes up to 45 cm, before interpolation ) , with a spatial extent up to 750 km. This compares well with the in situ data from [Fis­ cher et al., 1996] . This eddy propagates northeast­ ward with a speed up to 17 km / d. The anticyclonic eddy E2, located approximately at 13.5° N, 56.5° E, is also observed from June to September 1989, and is more intense at the end of August, with amplitudes of 20-25 cm, and a spatial extent ranging from 250550 km. This eddy propagated southeastward with a speed up to 7 km / d, and in September it seems to coalence with El. The third eddy observed is the cyclonic eddy E3, located at approximately 9° N, 510 E, with maxima of 25-30 cm at the end of June and during the first half of August. This cold wedge occurs as a result of the two anticyclonic eddies E1 and E2 and defines the so called northern wall [Tsai et at., 1992; Luther and o'Brian, 1985] . No clear ev­ idence of propagation is seen until August where a slight northward propagation, with a speed up to 17 km / d is observed. The life time of these eddies is observed to be more than three months. Eddy E1 is assumed to be the eddy called the Great Whirl [Peri­ gaud and Minster, 1988; Fischer et al., 1996] and E2 the northern Socotra eddy [Ali, 1990] . The anticyclonic eddy E4 and the cyclonic eddy E6 are observed during the NE monsoon in the same position as E1 and E3. Geosat show that E4 is most intense ( 30-35 cm ) during the first half of December in 1988, while the ERS-1 and Topex / Poseidon data show that this eddy is most intense during the be­ ginning of October 1993 ( 25-30 cm ) and 1994 ( 16-20 cm ) respectively. Geosat show a southward propaga­ tion up to 6 km / d, while there is no clear evidence

of propagation from the other two data sets. Geosat show that the cyclonic eddy E6 is most intense at the end of December 1993 ( 25-30 cm ) . The ERS-1 data show that this eddy is most intense during October (25-30 cm ) and in November 1994 ( 16-20 cm ) for Topex / Poseidon. All three show a southward prop­ agation up to 6 km / d. In the Bay of Bengal the anticyclonic eddies ElO, Ell and a cyclonic eddy E13 were observed during the SW monsoon. These are most intense during the months May and June, with maxima of 20-25 cm. Ell has a southward propagation with a speed up to 7 km / d, E13 has a westward propagation with a speed up to 20 km / d, and the northward propagating (11 km / d ) anticyclonic eddy E12, was observed only from the Geosat data. During the NE monsoon an anticyclonic north­ ward propagating eddy E14 with a speed up to 27 km / d, and a cyclonic westward propagating eddy E17, with a speed up to 10 km / d is observed. These eddies are most intense during the second half of De­ cember 1988, and 1994. 5 DISCUSSION AND CONCLUSION For the first time SSB data from three altimeter mis­ sions are used to study mesoscale variability in the Arabian Sea and the Bay of Bengal. The correlation between the three altimeters is very good. The Root Mean Square of the SSB anomalies indicates where the high energetic areas are. This formed the basis for the further investigation. Using CPCA analy­ sis we were able to study the temporal and spatial characteristics of the monsoon periods. The annual period based on Geosat data indicate maxima in Jan­ uary and June, indicating the time when the mon5

Southern Gyre, which together with the Great Whirl forms the well-known two gyre system in the Somali Current, was the only major eddy not observed in this investigation. Inspite of few investigations in especially the Bay of Bengal investigating the sea surface variability,our results confirms and strengthen previous results us­ ing other techniques [Potemra et al., 1991; Shetye et al., 1991;Shetye et al., 1996]. REFERENCES Figure 5: Overview of the location and rotation for eddies observed in the Indian Ocean using altimetry data.

Ali, M., Estimation of the horizontal velocity of the Scocotra eddy and some observations of sea surface thermal features using INSAT-IS, Int. J. Rem. Sensing, 11,41-47,1990. Anderson,D. L. T.,and P. B. Rowlands, The Somali Current response to the monsoon : The relative importance of local and remote forcing, J. Ma­ rine. Res., 34, 395-417,1976. AVISO, AVISO User Handbook : Merged

soon is most intense. The semi-annual period indi­ cate maxima in May,August and November, Febru­ ary,and appears to be related to the onset and decay phase of the SW and NE monsoons. Topex/Poseidon indicate oscillations caused by horizontal shear which generate eddies at a period of 40-50 days [Mysak and Mertz,1984; Schott et al., 1988; Swallow et al., 1988; Woodberry et al., 1989]. Westward propagating fea­ tures, possibly Rossby waves are also identified from the CPCA, along equator with speeds of 13 km/d. Along 4°_10° N speeds up to 6 km/d are observed, and between 5° S and equator they travel with a speed of 7 km/d. This indicate a decrease in speed away from equator, which compare well with Peri­ gaud and Delecluse (1992) results. Mesoscale eddies are located,and the propagation, spatial range, amplitude and life time of these are observed during the monsoon periods. These results compare well with previous investigators who have used altimeter measurements [Ali, 1990; Perigaud and Minster [1988; 1992; 1993]; Greiner and Peri­ gaud, 1993]. The dimensions of the eddies and the horizontal movement are within the expected range of the observations and model simulations. The cor­ relation between the three altimeter data sets used in this investigation is also very good. Small varia­ tions occur, due to the satellites different temporal and spatial sampling and due to different errors and corrections. The Great Whirl,the Socotra eddy and a cyclonic eddy are identified in the Arabian Sea dur­ ing the SW monsoon. During the NE monsoon an anticyclonic eddy, E4 located in the same place as the Great Whirl was observed, and another Socotra eddy. A cyclonic eddy was also observed north of the E4. A total of three anticyclonic eddies and one cyclonic eddy are observed in the Bay of Bengal,and their propagation speed and direction is investigated. The propagation of eddies during the SW monsoon seems to be eastward and throughout the months of the NE monsoon the eddies have a westward prop­ agation, following the atmospheric circulation. The

TOPEX/POSEIDON Products, AVI-NT-02-101C, AVISO/CNES, Toulouse, France, first ed.,

1992. Bruce, J. G., D. R. Johnson, and J. C. Kindle, Ev­ idence for eddy formation in the eastern Arabian Sea during the northeast monsoon, J. Geophys. Res., 99,7651-7664,1994. Cheney, R. E., N. S. Doyle, B. C. Douglas, R. W. Agreen, 1. Miller, E. 1. Timmerman, and D. C. McAdoo, The complete Geosat altimeter GDR handbook, National Ocean Service, NOAA, Rockville, MD, 1991. Cox, M., Equatorially trapped waves and the gen­ eration of the Somali Current, Deep-Sea Res., 23, 1139-1152,1976. Cutler, A., and J. Swallow, Surface currents of the Indian Ocean:Compiled from historical data archieved by the Meterologial Office,Bracknell, UK, Rep. 187, Inst. of Oceanography, Wormley, England, 1984. Defant,A., Water bodies and stationary current con­ ditions at boundary surfaces, JPO, 1, 451-475, 1961. Evensen, G., and P. J. van Leeuwen, Assimilation of Geosat Altimeter Data for the Agulhas Cur­ rent using the Ensemble Kalman Filter with a Quasi-Geostrophic Model, submitted to Monthly Weather review, 1995. Fischer, J., F. Schott, and L. Stramma, Currents and transports of the Great Whirl-Socotra Gyre system during the summer monsoon,August 1993, JGR, 101, 3573-3587,1996. Greiner,E.,and C. Perigaud, Assimilation of Geosat Altimetric Data in a Nonlinear Reduced-Gravity Model of the Indian Ocean. Part 1: Adjoint Ap­ proach and Model-Data Consistency, J. Phys. 6

Oceanogr., 24, 1783-1804,1993. Grundling, M. L., Tracking eddies in the south­ east Atlantic and southwest Indian Oceans with Topex / poseidon, J. Geophys. Res., 100, 2497724986, 1995. Horel, J., Complex Principal Component Analysis: Theory and examples, J. Climate and Appl. Me­ teor., 23, 1660-1673,1984. Indian Meteorological Society, A., Calender 1994, Calender, 1994. Le Traon, P., and M. Rouquet, Spatial Scales of Mesoscale Variability in the North Atlantic as De­ duced from Geosat Data, J. Geophys. Res., 95, 20267-20285,1990. Luther,M. E.,and J. J. o'Brian, A model of the sea­ sonal circualtion in the Arabian Sea forced by ob­ served winds, Prog. Oceanog., 14, 353-385, 1985. McCreary, J., W. Han, D. Shankar, and S. Shetye, Dynamics of the East India coastal Current 2. Nu­ merical solutions, JGR, 101, 13993-14010,1996. Molinari, R L., D. Olson, and G. Reverdin, Surface current distributions in the tropical India Ocean derived from compilations of surface bouy trajec­ tories, J. Geophys. Res., 95, 7217-7238,1990. Mysak, L., and G. Mertz, A 40-60 day oscillation in the source region of the Somali Current during 1976, JGR, 89, 711-715,1984. Perigaud, C., and P. Delecluse, Annual Sea Level Variations in the Southern Tropical Indian Ocean from Geosat and Shallow-Water Simula­ tions, J. Geophys. Res., 97, 20169-20178,1992. Perigaud, C., and P. Delecluse, Interannual sea level variations in the tropical Indian Ocean from Geosat and shallow water simulations, J. Phys. Oceanogr., 23, 1916-1934,1993. Perigaud, C., and J. Minster, Variability of the So­ mali Current as Observed from Seasat Altimetry, J. Phys. Oceanogr., 18, 25-38,1988. Pickard, G. L., and W. J. Emery, Descriptive Phys­ ical Oceanography, an Introduction, Pergamon Press, 1990. Potemra, J. T., M. E. Luther, and J. L. O'Brian, The seasonal circulation of the upper ocean in the Bay of Bengal, J. Geophys. Res., 96,12667-12683, 1991. Preisendorfer, R, Principal Component Analysis in Meteorology and Oceanography, Elsevier, New York,1988. Rao, R R, R L. Molinari, and J. F. Festa, Evo­ lution of the climatological near-surface thermal structure of the tropical Indian Ocean. 1. Descrip­ tion of mean monthly mixed layer depth, and sea surface temperature, surface current, and surface meteorological fields, J. Geophys. Res., 94 ( C8 ) , 10801-10815,1989. Samuel, P., Applications of Satellite Altimeter Data

Atlantic Ocean and the Nordic Seas, Ph.D. thesis,

Univ. of Bergen / NERSC, Bergen,Norway,1993. Samuel, P., J. A. Johannessen, and O. M. Johan­ nessen, A study on the inflow of Atlantic water to the GIN Sea using GEOSAT atimeter data, in Nansen Centennial Volume on the Role of the Po­ lar Oceans in Shaping the Global Environment,

edited by Johannessen, O. M., J. E. Overland, and R Muench,number 85 in Geophysical Mono­ graph, American Geophysical Union, 1994. Schott, F., M. Fieux, J. Kindle, J. Swallow, and R Zantopp, The boundary current east and north of Madagascar,2 , Direct measurements and model comparisons, JGR, 93,4963-4974, 1988. Schott, F., J.Reppin, and J.Fischer, Currents and transports of the Monsoon Current south of Sri Lanka, J. Geophys. Res., 99, 25127-25141,1994. Shankar, D., J. McCreary, W. Han, and S. Shetye, Dynamics of the East India Coastal Current 1. An­ alytic solutions forced by interior Ekman pump­ ing and local alongshore winds, JGR, 101, 1397513991,1996. Shetye, S., A. Gouveia, D. Shankar, S. Shenoi, P. Vinayachandran, D. Sundar, G. Michael, and G. Nampoothiri, Hydrography and circulation in the western Bay of Bengal during the Northeast mon­ soon, JGR, 101, 114011-14025,1996. Shetye, S. R, S. S. C. Shenoi, A. D. Gouveia, G. S. Michael, D. Sundar, and G. Nmpoothiri, Wind driven coastal upwelling along the western bound­ ary of the Bay of bengal during the southwest monsoon, Continental Shelf Research, 11, 13971408, 1991. Snaith, H. M., and 1. S. Robinson, A study of cur­ rents south of Africa using Geosat satellite altime­ try, JGR, 101, 18141-18154,1996. Swallow, J., M. Fieux, and F. Schott, The bound­ ary current east and north of Madagascar,1 , Geostrophic currents and transports, JGR, 93, 4951-4962,1988. Tsai, P. T., J. J. O'Briab, and M. E. Luther, The 26-day oscillation observed in the satellite sea sur­ face temperature measurements in the equatorial western Indian Ocean, J. Geophys. Res., 97,96059618, 1992. Wakker, K., R Zandbergen, M. Naeije, and B. Am­ brosius, Geosat Atimeter Data Analysis for the Oceans Around South Africa, J. Geophys. Res., 95, 2991-3006,1990. Woodberry, K. E., M. E. Luther, and J. J. O'Brien, The Wind-Driven Seasonal Circulation in the Southern Tropical Indian Ocean, J. Geophys. Res., 94, 17985-18002,1989.

for Studies on the Water Exchange between the

7

Ali, M., Estimation of the horizontal velocity of the Scocotra eddy and some observations of sea surface thermal features using INSAT-IS, Int. J. Rem. Sensing, 11,41-47, 1990. Anderson,D. L. T.,and P. B. Rowlands, The Somali Current response to the monsoon : The relative importance of local and remote forcing, J. Ma­ rine. Res., 34, 395-417,1976. AVISO, AVISO User Handbook : Merged

duced from Geosat Data, J. Geophys. Res., 95, 20267-20285,1990. Luther,M. E.,and J. J. o'Brian, A model of the sea­ sonal circualtion in the Arabian Sea forced by ob­ served winds, Prog. Oceanog., 14, 353-385, 1985. McCreary, J., W. Han, D. Shankar, and S. Shetye, Dynamics of the East India coastal Current 2. Nu­ merical solutions, JGR, 1 01, 13993-14010,1996. Molinari, R. L., D. Olson,and G. Reverdin, Surface current distributions in the tropical India Ocean derived from compilations of surface bouy trajec­ tories, J. Geophys. Res., 95, 7217-7238,1990. Mysak, L., and G. Mertz, A 40-60 day oscillation in the source region of the Somali Current during 1976, JGR, 89, 711-715,1984. Perigaud, C., and P. Delecluse, Annual Sea Level Variations in the Southern Tropical Indian Ocean from Geosat and Shallow-Water Simula­ tions, J. Geophys. Res., 97, 20169-20178, 1992. Perigaud, C., and P. Delecluse, Interannual sea level variations in the tropical Indian Ocean from Geosat and shallow water simulations, J. Phys. Oceanogr., 23, 1916-1934,1993. Perigaud, C., and J. Minster, Variability of the So­ mali Current as Observed from Seasat Altimetry, 1. Phys. Oceanogr., 18, 25-38, 1988. Pickard,G. L., and W. J. Emery, Descriptive Phys­ ical Oceanography, an Introduction, Pergamon Press, 1990. Potemra, J. T., M. E. Luther, and J. L. O'Brian, The seasonal circulation of the upper ocean in the Bay of Bengal, J. Geophys. Res., 96,12667-12683, 1991. Preisendorfer, R., Principal Component Analysis in Meteorology and Oceanography, Elsevier, New York, 1988. Rao, R. R., R. L. Molinari, and J. F. Festa, Evo­ lution of the climatological near-surface thermal structure of the tropical Indian Ocean. 1. Descrip­ tion of mean monthly mixed layer depth, and sea surface temperature, surface current, and surface meteorological fields, J. Geophys. Res., 94 ( C8 ) , 10801-10815,1989. Samuel, P., Applications of Satellite Altimeter Data

-

T OPEX/POSEIDON Products, AVI-NT-02-101C, AVISO/CNES, Toulouse, France, first ed.,

1992. Bruce, J. G., D. R. Johnson, and J. C. Kindle, Ev­ idence for eddy formation in the eastern Arabian Sea during the northeast monsoon, J. Geophys. Res., 99,7651-7664,1994. Cheney, R. E., N. S. Doyle, B. C. Douglas, R. W. Agreen, L. Miller, E. L. Timmerman, and D. C. McAdoo, The complete Geosat altimeter GDR National Ocean Service, NOAA, handbook, Rockville, MD, 1991. Cox, M., Equatorially trapped waves and the gen­ eration of the Somali Current, Deep-Sea Res., 23, 1139-1152,1976. Cutler, A., and J. Swallow, Surface currents of the Indian Ocean:Compiled from historical data archieved by the Meterologial Office,Bracknell, UK, Rep. 187, Inst. of Oceanography, Wormley, England, 1984. Defant,A., Water bodies and stationary current con­ ditions at boundary surfaces, JPO, 1, 451-475, 1961. Evensen, G., and P. J. van Leeuwen, Assimilation of Geosat Altimeter Data for the Agulhas Cur­ rent using the Ensemble Kalman Filter with a Quasi-Geostrophic Model, submitted to Monthly Weather review, 1995. Fischer, J., F. Schott, and L. Stramma, Currents and transports of the Great Whirl-Socotra Gyre system during the summer monsoon,August 1993, J G R, 1 01, 3573-3587, 1996. Greiner,E.,and C. Perigaud, Assimilation of Geosat Altimetric Data in a Nonlinear Reduced-Gravity Model of the Indian Ocean. Part 1: Adjoint Ap­ proach and Model-Data Consistency, J. Phys. Oceanogr., 24, 1783-1804,1993. Grundling, M. L., Tracking eddies in the south­ east Atlantic and southwest Indian Oceans with Topex/poseidon, J. Geophys. Res., 100, 2497724986,1995. Horel, J., Complex Principal Component Analysis: Theory and examples, J. Climate and Appl. Me­ teor., 23, 1660-1673,1984. Indian Meteorological Society, A., Calender 1994, Calender, 1994. Le Traon, P., and M. Rouquet, Spatial Scales of Mesoscale Variability in the North Atlantic as De-

for Studies on the Water Exchange between the Atlantic Ocean and the Nordic Seas, Ph.D. thesis,

Univ. of Bergen/NERSC, Bergen, Norway, 1993. Samuel, P., J. A. Johannessen, and O. M. Johan­ nessen, A study on the inflow of Atlantic water to the GIN Sea using GEOSAT atimeter data, in Nansen Centennial Volume on the Role of the Po­ lar Oceans in Shaping the Global Environment,

edited by Johannessen, O. M., J. E. Overland, and R. Muench,number 85 in Geophysical Mono­ graph, American Geophysical Union, 1994. Schott, F., M. Fieux, J. Kindle, J. Swallow, and R. Zantopp, The boundary current east and north of 8

Madagascar,2 , Direct measurements and model comparisons, JGR, 93, 4963-4974,1988. Schott, F., J.Reppin, and J.Fischer, Currents and transports of the Monsoon Current south of Sri Lanka, J. Geophys. Res., 99, 25127-25141,1994. Shankar, D., J. McCreary, W. Han, and S. Shetye, Dynamics of the East India Coastal Current 1. An­ alytic solutions forced by interior Ekman pump­ ing and local alongshore winds, JGR, 101, 1397513991,1996. Shetye, S., A. Gouveia, D. Shankar, S. Shenoi, P. Vinayachandran, D. Sundar, G. Michael, and G. Nampoothiri, Hydrography and circulation in the western Bay of Bengal during the Northeast mon­ soon, JGR, 101, 114011-14025,1996. Shetye, S. R., S. S. C. Shenoi, A. D. Gouveia, G. S. Michael, D. Sundar, and G. Nmpoothiri, Wind driven coastal upwelling along the western bound­ ary of the Bay of bengal during the southwest monsoon, Continental Shelf Research, 11, 13971408, 1991. Snaith, H. M., and 1. S. Robinson, A study of cur­ rents south of Africa using Geosat satellite altime­ try, JGR, 101, 18141'-18154,1996. Swallow, J., M. Fieux, and F. Schott, The bound­ ary current east and north of Madagascar,1 , Geostrophic currents and transports, JGR, 93, 4951-4962,1988. Tsai, P. T., J. J. O'Briab, and M. E. Luther, The 26-day oscillation observed in the satellite sea sur­ face temperature measurements in the equatorial western Indian Ocean, J. Geophys. Res., 97,96059618, 1992. Wakker, K., R. Zandbergen, M. Naeije, and B. Am­ brosius, Geosat Atimeter Data Analysis for the Oceans Around South Africa, J. Geophys. Res., 95, 2991-3006,1990. Woodberry,K. E., M. E. Luther, and J. J. O'Brien, The Wind-Driven Seasonal Circulation in the Southern Tropical Indian Ocean, J. Geophys. Res., 94, 17985-18002,1989.

9