or collective redistirbutionor collective of this by of any article portion Th THE INDONESIAN SEAS published in been has is article Oceanography
INDONESIAN THROUGHFLOW , Volume 4, a quarterly journal of Th 18, Number TRANSPORT VARIABILITY ESTIMATED or other means reposting, photocopy machine, is only w permitted FROM
Satellite Society. Copyright e Oceanography 2005 by Th
AltimetryBY JAMES T. POTEMRA ith the approval of Th e Oceanography Society. All rights reserved. Permission reserved. Society. All rights is e Oceanography gran e Oceanography Society. Send all correspondence to: Society. Send e Oceanography [email protected] or Th
Figure 1. Th e Pacifi c to Indian Ocean exchange, known as the Indonesian throughfl ow (ITF), has been sampled with a repeat expendable bathythermograph (XBT) hydrographic section from Australia to Indonesia (IX-1, red stars). Sea-level estimates from the TOPEX/Poseidon (T/P) satel- lite (ground tracks given with the black lines) are used to estimate ITF transport. Th e major straits through which the ITF passes are indicated on the map. ted in teaching to copy thisand research. for use Repu article e Oceanography Society, PO Box 1931, Rockville, MD 20849-1931, USA. 1931, Rockville, Box PO Society, e Oceanography blication, systemmatic reproduction,
98 Oceanography Vol. 18, No. 4, Dec. 2005 The relatively intense boundary cur- ment of sea level that may be used to prise the Indonesian seas (Figure 1). rents of the western equatorial Pacifi c, index ITF fl ow. Velocities through some of the narrow as well as the western Pacifi c warm pool, Wyrtki (1987) was the fi rst to suggest straits, Lombok Strait for example, can have long been recognized as key com- that ITF transport could be estimated be quite extreme, reaching more than 1 ponents in the global climate system. from sea-level differences between the m s-1 over a shallow sill at the southern More recently, the eastern Indian Ocean Pacifi c and Indian Oceans. The reason- side (Hautala et al., 2001). In addition, has been identifi ed as a potential source ing was based on the hypothesis that the tides in the Indonesian seas are large and for climate variability over a larger area large-scale pressure difference between can sometimes contribute to swift fl ows (Saji and Yamagata, 2003). These systems the two basins provided the forcing for in the straits. Upper-ocean instruments are not isolated, however, and Pacifi c- the ITF. Wyrtki used sea-level varia- on a bottom-anchored mooring in the to-Indian Ocean exchange occurs via tions from coastal tide-gauge stations at Makassar Strait, for example, were driven the Indonesian throughfl ow (ITF) (see Davao, Mindanao Island (Philippines), more than 100 m vertically from moor- Gordon, this issue). Understanding this and at Darwin, Australia. The latter was ing blow-over due to semi-diurnal tides exchange, and the ability to estimate it, to be representative of Indian Ocean within the strait (Gordon et al., 1999). are therefore essential for understand- conditions. Wyrtki pointed out, however, Although fl ow through these individ- ing the global climate system. Directly that a more suitable location for sea-level ual straits is of interest, the net exchange measuring this fl ow has been challeng- measurements would have been along between the Pacifi c and Indian Oceans ing. This paper outlines a method to es- the Indonesian archipelago, but data is also important, as it represents a key timate the large-scale oceanic transport were not available at that time. Clarke component in the heat and freshwater in the ITF by using satellite-measured and Liu (1994) further explained that the budgets of the two ocean basins. The sea level from the TOPEX/Poseidon (T/P) sea-level patterns associated with the an- variability of ITF transport, as it affects altimeter. The satellite measurements do nual cycle and those associated with in- upper-ocean heat content, may also play not directly measure transport, but they terannual variations were quite different a role in climate variability in the region. comprise a long, almost global, measure- at these locations; the simple sea-level The forcing, therefore, as well as the difference between Davao and Darwin potential effects of ITF transport vari- cannot adequately represent both time ability, is large scale. To estimate ITF scales. Today, however, satellite estimates variations, simultaneous large-scale of sea level can be used to produce a measurements are needed. Satellite mea- more robust index, while long-term di- surements of sea level provide one such rect measurements of ITF transport have measurement. Sea level from the T/P not yet become available. altimeter provides nearly simultane- The ITF is diffi cult to measure for ous, global coverage and may be used a variety of reasons. As Pacifi c waters to produce an estimate of net ITF ex- move from the western equatorial Pa- change. The challenge is to determine cifi c to the eastern Indian Ocean, they the relationship between sea level and travel through a complex array of nar- ITF transport. In this study a numerical row straits and deep basins that com- model is used.
To estimate ITF variations, simultaneous large-scale measurements are needed.
Oceanography Vol. 18, No. 4, Dec. 2005 99 Potemra et al. (1997) (P97 model) altimeter data to estimate ITF transport. multivariate sequential data assimilation showed that sea level in certain key re- First, the variability of the ITF transport scheme (Carton et al., 2000) that updates gions, as measured by T/P, could be com- is described in order to judiciously select the ocean model with observations of bined to produce an estimate of net ITF sea-level locations that will best refl ect temperature and salinity every ten days. transport. A simple, linear model was ITF forcing. Next, the model results are It should be noted that satellite altimeter constructed as follows: described, and fi nally the index is ap- data were not assimilated into the model plied to the satellite data. run used in this study. T V ITF(t) = Σ αi ηi(t), [1] The annual cycle of ITF transport ESTIMATES OF ITF from SODA accurately shows peak T where V ITF(t) is the total ITF volume TRANSPORT BASED ON transport during the southeast monsoon transport as a function of time (mean OBSERVATIONS AND MODELS (roughly July–October) and minimum removed), ηi(t) are sea-level variations at Long-term, net ITF transport has been during the northwest monsoon, as pre- location i (mean removed and normal- estimated from in situ observations in dicted by Wyrtki (1961). The annual ized to unit variance), and αi are weights some of the key Indonesian passages. cycle is most evident in the upper 200 determined from a least squares fi t. Understanding that these measurements m in the SODA results (similar to other Since that study, advancements in nu- have been acquired in different years, the models), and represents about 60 per- merical models and new in situ measure- mean ITF is estimated to be between 6.0 cent of the variance in this layer (9 Sv2 ments have lead to a better understand- and 10.0 Sv (Hautala et al., 2001); see compared to 15 Sv2). A deeper layer, 200 ing of the temporal and spatial variability Gordon (this issue) for a full review. to 500 m, has less overall variance (6.9 of ITF transport, particularly the vertical Model estimates of ITF transport Sv2), but only 6 percent is due to sea- variability of this fl ow (Potemra et al., are typically higher than the observed sonal variability. 2003). In addition, there are now more estimates. Perhaps due to insuffi cient This model result is consistent with than ten years of T/P measurements that resolution of the complex bathymetry, the hypothesis that lower-frequency forc- can be used to make a more robust ITF or uncertainties in wind forcing, most ing of the ITF, for example, that associ- index. For this study, version 1.0 NASA/ large-scale models, particularly coarse ated with El Niño-Southern Oscillation GSFC Ocean Pathfi nder gridded sea-level resolution models, give net ITF trans- (ENSO) variations, is evident in a deeper variations (obtained from ftp://iliad.gsfc. ports in excess of 15 Sv. layer of the ITF (Potemra et al., 2003). nasa.gov) were used. The model used in this study, the Interannual variability in ITF transport The P97 study is, therefore, revis- Simple Ocean Data Assimilation-Paral- is less well known than the seasonal cycle, ited, using both a longer satellite record lel Ocean Program (SODA POP 1.2) mainly due to a lack of suffi ciently long of sea level and a more sophisticated model (henceforth referred to as SODA) observations. One notable exception is ocean model to determine the best fi t produces a realistic ITF transport with a the nearly twenty-year repeat expend-
(the αi’s in equation 1). As in P97, the net mean of 12.7 Sv. This model has 40 able bathythermograph (XBT) section relationship between ITF and sea level vertical levels that vary from 10 m near from Shark Bay, Australia to the Sunda was determined using an ocean general the surface to 250-m thick in the deep Strait known as IX-1 (red stars in Figure circulation model (OGCM), and then ocean. The original SODA grid uses a 1). Meyers et al. (1995) computed upper- this relationship was applied to the T/P “displaced pole,” and is 0.4° in longitude ocean transport (0 to approximately 400 by 0.28° latitude, but the model results m) based on these XBT data for the early James T. Potemra ([email protected]) is were regridded to a regular 0.5° by 0.5° part of the record and found a response Assistant Researcher, School of Ocean and grid. The model was forced by ERA-40 to ENSO in the sense that warm El Niño Earth Science and Technology, International daily averaged winds from the European events were associated with reduced ITF Pacifi c Research Center, University of Ha- Center for Medium Range Weather Fore- transport (Meyers, 1996). A possible waii, Honolulu, HI, USA. casts (ECMWF). The model also uses a explanation is that during these warm
100 Oceanography Vol. 18, No. 4, Dec. 2005 events, sea level in the western Pacifi c is ern end of the Java/Australia section et al. (1995). The model velocity along anomalously low, thus reducing the Pa- (England and Huang, 2005; McClean et this section is highest in the near-surface cifi c-to-Indian Ocean pressure gradient al., submitted) is longer than those gener- layer, particularly near the coast of Java. that is thought to drive the ITF. These ated in the Indian Ocean, and thus there Zonal velocity through this section is results have been confi rmed to a certain is not a coherent ENSO signal in the net, maximum between 11°S and the coast degree by recent modeling studies (e.g., along-track integrated transport. and reaches almost 50 cm s-1 in the up- England and Huang [2005] and McClean per 30 m. It is half that value at 100 m, et al. [submitted]) in that time-fi ltered SODA ITF Transport and by 200 m the velocity is 12 cm s-1. model transport values show a correla- ITF transport was computed as the South of 11°S the surface fl ow is between tion to ENSO indices. The correlation, depth and cross-strait integral of zonal 0 and 10 cm s-1 (westward). while signifi cant, is not very high (–0.37 velocity from SODA along a line from These velocities integrate to a net ITF between upper ocean ITF transport and Australia to South Java (Figure 2). This transport of 12.7 Sv. The variability of ENSO) (England and Huang, 2005), and section is close to the IX-1 line of Meyers this transport is different in distinct re- depends on the vertical and horizontal section limits of the transport calcula- tion. For example, transport near the coast of Australia shows a higher corre- lation with ENSO than transport varia- tions near South Java (McClean et al., submitted). The ITF-to-ENSO relationship is somewhat obscured when looking at to- tal depth integrated transport because local winds act on the near-surface layer, while Pacifi c (ENSO) forcing controls ITF variations at a deeper depth. Further, ITF transport computed as the integral of velocity along the IX-1 section from Aus- tralia to Indonesia not only includes ITF infl uences, but also infl uences from In- dian Ocean circulation (from the south- ern subtropical gyre) and from the South Java Current, a seasonal current along the southern coasts of Sumatra/Java. Indian Ocean forcing, such as from the Indian Ocean Dipole (IOD) (Saji et al., 1999), also complicates this signal, and directly infl uences transport on the northern edge of the Java/Australia section, while Figure 2. Data from transect shown in Figure 1 (green line). ITF transport occurs in dis- Pacifi c forcing controls low-frequency crete sections, both in the vertical and horizontal. Th e upper panel shows the bathymetry variations on the southern edge (Wijffels along a section from Australia (left side) to Java (right side). In the lower panel, the mean (
Oceanography Vol. 18, No. 4, Dec. 2005 101 gions along the track. The upper 200 m independent and in fact are uncorrelated of the section near Australia. This result (upper 14 model levels) carry 83 percent in the SODA results. is consistent with Pacifi c forcing driving of the total transport (10.6 Sv), and 80 The variance in transport is greatest the interannual variations, because these percent of this occurs in the northern in the lower layer of each of the Java/ signals would follow a wave guide along part of the section near the coast of Java. Australia subsections (Figure 2), but the the coasts of New Guinea and Australia It is hypothesized that Indian Ocean ef- variance in total transport along the sec- (Wijffels and Meyers, 2004). fects, both direct wind forcing and coast- tion is largest in the surface layer, 15 Sv2 In summary, fl ow between Australia al waves (e.g., Sprintall et al., 2000; Wi- compared to 7 Sv2 for the lower layer, and south Java is infl uenced by local, sea- jffels and Meyers, 2004) are seen near the suggesting that there is recirculation in sonal winds that mostly affect the upper coast of Java, while other forcing, includ- the lower layer (i.e., the variance in each ocean fl ow, and low-frequency forcing ing remote Pacifi c winds and local (Aus- horizontal subsection cancels each oth- from the Indian and Pacifi c Oceans that tralian) alongshore winds are responsible er). Most of the variability in transport is seen at thermocline depth. A closer for the variability in the southern part of is at the annual and semi-annual peri- look at ITF transport from the SODA the section. Thus, transport in these two ods (Figure 3), but interannual varia- results shows that most of the ITF trans- ends of the Java/Australia section appear tions are strongest in the southern end port, computed along a line from Austra-
4 0.5yr 1yr 3yr 6yr 4 0.5yr 1yr 3yr 6yr 10 10 Figure 3. Th e variability of transport from the subsec- Australia mid−passage 3 3 tions in Figure 2 is estimat- 10 10 ed by power spectra. Th e surface fl ow in each section 2 2 10 10 is given with a green line, the mid-depth transport with a purple line, and the 1 1 10 10 spectra for total depth-in- spectral energy tegrated transport is given 0 − 200m 0 − 200m 0 0 with the black line. All 10 200 − 3000m 10 200 − 3000m total total three subsections, as well as the total, show peaks at the annual and semian- 0 1 2 3 4 0 1 2 3 4 10 10 10 10 10 10 10 10 10 10 nual periods. ENSO-type 4 4 10 0.5yr 1yr 3yr 6yr 10 0.5yr 1yr 3yr 6yr variability is seen most strongly in the section near S. Java TOTAL ITF Australia. 3 3 10 10
2 2 10 10
1 1 10 10 spectral energy
0 − 200m 0 − 200m 0 0 10 200 − 3000m 10 200 − 3000m total total
0 1 2 3 4 0 1 2 3 4 10 10 10 10 10 10 10 10 10 10 Period (months) Period (months)
102 Oceanography Vol. 18, No. 4, Dec. 2005 lia to South Java, does occur in distinct where, ηSJ is sea level (mean removed, equatorial Indian Ocean was included regions: near the coast of Java, near the normalized to unit variance) averaged for the new linear model of ITF trans- coast of Australia, and a section between over a region south of Java; ηWP is sea port. These fi ve areas correspond to the two. A simpler, geostrophic-type in- level (mean removed, normalized to areas of high correlation, in the SODA dex, based on the cross-strait pressure unit variance) averaged over a region results, between sea level and ITF trans- difference, would not capture all these in the western Pacifi c warm pool; ηDAR port (Figure 4). The Indian Ocean wave processes. To estimate ITF transport from is sea level (mean removed, normalized guide along the southern coasts of Su- sea level, therefore, locations that capture to unit variance) at Darwin, Northwest matra and Java, as well as the Pacifi c these dynamics should be included. Australia; and ηDAV is sea level (mean re- Ocean wave guide along the west coasts moved, normalized to unit variance) at of New Guinea and Australia, are evi- ESTIMATE OF ITF TRANSPORT Davao, Philippines. dent in the continuous regions of high BASED ON SEA LEVEL The OGCM used to determine the correlation in the upper layer and total P97 pursued the original work of Wyrtki optimal fi t of sea level to ITF transport transport, respectively. Total ITF trans- (1987) and incorporated sea level from was the Parallel Ocean Climate Model port, as correlated to sea level (Figure four locations into a linear model of (POCM) of Semtner and Chervin 4), is consistent with an “island-rule” geostrophic ITF transport. The four lo- (1992); nine years of monthly mean (Godfrey, 1989) type of balance, with cations were chosen to represent physi- POCM sea level and velocities were used high correlations evident along the west- cal processes in the Pacifi c and Indian to derive the fi t. ern edge of New Guinea and Australia Oceans that were thought to control In this new study, results from as well as the equatorial Pacifi c. Some of variability in the ITF. The best fi t was SODA POP 1.2 are used. To incorpo- these lower-frequency effects are found given as: rate low-frequency variability from the on higher vertical modes in the ITF, so Indian Ocean, separate from local, In- the new ITF index will use lagged sea VT = 2.40 η – 0.91 η [2] ITF SJ WP donesian effects, a fi fth station in the level as well as contemporaneous. – 0.23 ηDAR – 1.41 ηDAV
Figure 4. Sea level from the SODA model was correlated to upper layer (left) and total depth-integrated transport (right). Regions that were used to index ITF transport variability are given with the red boxes. Th ese regions have relatively high correlations (sea level to ITF) and also are in regions of dynamical importance: the equatorial Indian Ocean; along South Java; Davao (Philippines); Darwin (Australia); and the western Pacifi c warm pool. An index of ITF transport was therefore made based on sea level in these discrete locations.
Oceanography Vol. 18, No. 4, Dec. 2005 103 1’ RESULTS England and Huang, 2005). In fact, the V ITF(t) = 2.42 η'SJ(t) – 1.15 η'SJ(t–1) [5]
Sea Level-ITF Relationship wave guide from the Pacifi c and Indian – 1.16 η'DAR(t–1) from SODA Oceans are along the coasts, so no sig- Given the results from the SODA model nifi cant skill is lost by just using the The linear model given by [4] reproduces 1 (specifi cally the identifi cation of lower- lagged sea level at Darwin for Pacifi c the 0 to 200 m SODA transport, V ITF(t), frequency, fi rst baroclinic mode signals interannual variability and at south Java with a correlation of 0.85 (Figure 5a). from the Pacifi c), similar locations were for interannual variability in the Indian Note in this case that only sea level from used as in P97, but these were combined Ocean. Using 44 years of monthly mean either end of the ITF section is required, with lagged signals at these same sites. sea level from SODA, the best skill is but that a component from each at a later The new model for ITF transport is obtained when fi tting sea level just at month is necessary to get the baroclinic therefore: South Java and Australia (Darwin) to portion of the signal. Much of the skill upper layer ITF transport; including sea- could be due to the dominance of the
VITF(t) = Σ αi ηi(t) + Σ γi ηi(t–δt), [3] level variations in the equatorial Indian annual cycle and coincident variability Ocean, western Pacifi c warm pool, and in sea level. To examine this relationship, In this case, the lag (δt) was taken to Davao do not improve the skill. The best a fi t was made to the upper layer ITF be from one to eight months, roughly fi t is thus: transport with the mean seasonal cycle the time for Pacifi c oceanic signals to removed (V1’ (t ), equation 5). In this V1 (t) = 4.96 η (t) – 2.16 η (t–1) [4] ITF reach the west coast of Australia (see ITF SJ SJ case, the correlation between sea-level fi t – 1.62 ηDAR(t–1)
10 Figure 5. Th e estimate of ITF transport from a 5 sea level at discrete locations (green line) is compared to the actual transport from the 0 SODA model (purple line). Four cases are Sv presented: upper layer transport with and −5 0−200m transport (mean removed) without the mean seasonal cycle (upper two sea level fit −10 panels) and total transport with and without b the mean seasonal cycle (lower two panels). 5 Th e long-term mean has been removed, and negative numbers indicate an increase in 0 Sv fl ow from the Pacifi c to the Indian Ocean. Th e high correlations in each case indicate −5 0−200m transport (seasonal cycle removed) sea level fit the ability of estimating ITF transport from −10 sea level at specifi c locations. Diff erent fi ts c were determined for each of these cases (see 5 text) because various forcing mechanisms 0 aff ect transport in the upper layer and at dif- Sv ferent time scales. −5 total transport (mean removed) sea level fit −10 d 5
0 Sv
−5 total transport (seasonal cycle removed) sea level fit −10 1960 1965 1970 1975 1980 1985 1990 1995 2000
104 Oceanography Vol. 18, No. 4, Dec. 2005 T and model transport is 0.69 (Figure 5b), V ITF(t) = 3.89 ηSJ(t) – 2.04 ηSJ(t–1) [6] ITF Transport Estimate Using but the same two locations provide the – 1.24 ηDAR(t–1) – 1.24 ηDAV(t+1) TOPEX/Poseidon highest skill in capturing the ITF signal. Now that a simple model for ITF trans- T’ The relative sizes, as well as the signs, V ITF(t) = 1.06 η'EIO (t) – 1.38 η'DAR(t) [7] port based on sea level has been deter- of the coeffi cients αi and γI give an indi- – 0.89 η'WP(t–1) + 1.06 ηDAR(t–6) mined, the fi nal step is to apply this to cation of the importance of each forcing the T/P measurements of sea level. One region. The concurrent sea level at South In this case, the sea-level fi t to trans- pass of the T/P orbit passes very close Java is about twice as important as the port has lower skill, correlations of 0.70 to the IX-1 line (see Figure 1). Sea level T other two in both [4] and [5], so at zero for the total transport, V ITF(t), and 0.51 at a nearly ten-day interval along this T’ lag, the SODA ITF transport is consistent for the seasonal anomalies (V ITF(t); pass shows similar variability to the with geostrophic balance; an increase in Figure 5c and d, respectively). Here, SODA ITF transport. Spectral analysis of the along-track pressure gradient (South variability from the North Pacifi c sea-level variability (Figure 6) along this Java minus Darwin) gives eastward fl ow, (Davao) is required to improve the line as measured by T/P shows a peak in or a reduced ITF transport. skill for total transport, while variabil- annual energy along the entire section More processes from both the Pacifi c ity from the equatorial Indian Ocean from Australia to south Java. Energy off and Indian Oceans infl uence the total and the western Pacifi c warm pool are the coast of Java is dominated by semi- depth-integrated ITF transport, and the needed to estimate seasonal anomalies annual and interannual variations. In- best fi t was found using the following: of total transport. terannual variability in the southern end
Figure 6. Th e power spectra from a single along-track pass of the T/P satellite (close to the XBT line in Figure 1) is shown in three dimensions: the period is given along the front axis, location along the track (from Australia to Java) is given along the right axis, and energy is shown verti- cally. Similar to the SODA results, the T/P has strong semiannual variability all along the section (red ridge at 180 days), strong annual variability near Indonesia, and strong interannual vari- ability near Australia.
Oceanography Vol. 18, No. 4, Dec. 2005 105 of the T/P pass, near Australia, is smaller the coasts where many of the indices variations of T/P agree with the trans- and occurs at lower frequency than at the are based (although sea level was aver- port estimates from Hautala et al. (2001), northern end of the pass. This is consis- aged over a larger region), but the T/P but it is not clear what is controlling tent with the SODA results and hypoth- sea level in the regions used in this study these interannual variations, and why the eses developed previously. have extremely high correlations to the model might be overestimating them. Therefore, the linear model given by model sea level; correlations are between equations [4] to [7] was applied to the 0.87 and 0.95 for sea level both with and CONCLUSIONS T/P sea level averaged over the same without the seasonal cycle. Sea-level measurements from the TO- regions given in Figure 4. The results, The estimates of transport with sea- PEX/Poseidon altimeter have been used given in Figure 7, are consistent with sonal cycle included match the model to derive an estimate of ITF transport the SODA model in terms of variabil- and observed transport (Figures 7a and variability. Locations were chosen that ity. The estimate based on T/P, however, c) better than the anomalies (without highlight specifi c processes that deter- shows a smaller seasonal cycle, about a a seasonal cycle), despite the high cor- mine the variations in ITF transport, and 6 Sv range. The seasonal anomalies are relation in the input time series. The a least-square fi t was used to determine even smaller, between +/– 2 Sv, and the relatively small differences become more the relative contributions. The utility of relationship to ENSO is not obvious. apparent in the transport estimate. For this approach is two-fold: (1) the rela- However, it should be cautioned that the example, in late 1999 through 2002, the tive size and sign of the weights give an fi t to total transport anomalies has the model transport anomalies range from indication of the important regions, and weakest skill. There are certainly limi- –5 to +8 Sv, while the estimate from T/P by extension the forcing, for ITF varia- tations to the T/P measurements near are much smaller (+/– 3 Sv). The smaller tions, and (2) such an index may be used
Figure 7. Pacifi c to Indian Ocean transport A was estimated with sea level from the T/P satellite (purple lines). Th e estimate was constructed based on results from the 3V ´ SODA model (green line). Results from in )8´ ESTIMATE BASED ON 8"4 MEAN REMOVED FROM 3PRINTALL ET AL situ measurements are shown with brown ´ ´M TRANSPORT FROM 3/$! MEAN REMOVED ´M TRANSPORT ESTIMATED FROM 4/0%8 MEAN REMOVED shading. Similar to Figure 5, the upper pan- ´ els are for transport in the upper 200 m B (with and without the seasonal cycle), and the lower panels are for the total transport. Th e best estimates are seen for the transport 3V ´ with the seasonal cycle included; seasonal )8´ ESTIMATE BASED ON 8"4 MEAN REMOVED FROM 3PRINTALL ET AL ´ ´M TRANSPORT FROM 3/$! SEASONAL CYCLE REMOVED anomalies of transport are not as accurately estimated from sea-level variability. ´ ´M TRANSPORT ESTIMATED FROM 4/0%8 WITHOUT SEASONAL CYCLE C 3V ´ )8´ ESTIMATE BASED ON 8"4 MEAN REMOVED FROM 3PRINTALL ET AL ´ 4OTAL TRANSPORT FROM 3/$! MEAN REMOVED ´ 4OTAL TRANSPORT ESTIMATED FROM 4/0%8 MEAN REMOVED D 3V ´ )8´ ESTIMATE BASED ON 8"4 MEAN REMOVED FROM 3PRINTALL ET AL ´ 4OTAL TRANSPORT FROM 3/$! SEASONAL CYCLE REMOVED 4OTAL TRANSPORT ESTIMATED FROM 4/0%8 WITHOUT SEASONAL CYCLE ´
106 Oceanography Vol. 18, No. 4, Dec. 2005 to monitor ITF changes with either the (equations 4 and 5), but more complex, the northern and eastern Indian Ocean. Journal of Physical Oceanography 24:1,224–1,235. global sea-level fi eld from T/P or other wider-area forcing is required for total England, M.H., and F. Huang. 2005. On the interan- satellites or from in situ tide gauges. depth ITF transport. nual variability of the Indonesian throughfl ow and its linkage with ENSO. Journal of Climate By using SODA POP 1.2, the weights The large-scale nature of ITF forcing, 18:1,435–1,444. of sea level change somewhat from those while still being understood, demon- Godfrey, J.S. 1989. A Sverdrup model of the depth- integrated fl ow for the world ocean allowing for derived in P97. In that study, time- strates the need for suitably large-scale island circulations. Geophysical and Astrophysical lagged sea level was not considered. This observations for an accurate index. The Fluid Dynamics 45:89–112. Gordon, A.L., R.D. Susanto, and A. Ffi eld. 1999. result leads to an index based on the satellite estimates of sea level provide Throughfl ow within Makassar Strait transport. difference between sea level at South such a measurement and thus allow for Geophysical Research Letters 26:3,325–3,328. Hautala, S.L., J. Sprintall, J.T. Potemra, A.G. Ilahude, Java and a combination of sea level at a potential near-real-time monitoring J.C. Chong, W. Pandoe, and N. Bray. 2001. Veloc- Darwin, Davao, and the warm pool. The of ITF transport. The net ITF transport, ity structure and transport of the Indonesian throughfl ow in the major straits restricting fl ow three latter locations all refl ect Pacifi c which is forced by several factors in both into the Indian Ocean. Journal of Geophysical forcing. In the present study, the intro- the Pacifi c and Indian Oceans at several Research 106:19,527–19,546. McClean, J.L., D.P. Ivanova, and J. Sprintall. Submit- duction of a time lag allows the elimi- frequencies, is quite complex, and as ted. Remote Origins of biennial and interannual nation of the warm-pool location, but new in situ measurements become avail- variability in the Indonesian throughfl ow region from WOCE IX1 XBT data and a global eddy- again shows an index that is in the geo- able, an index such as this can certainly permitting ocean model. Journal of Geophysical strophic sense: Indian Ocean sea level be improved. Research. Meyers, G. 1996. Variation of Indonesian through- (South Java) minus Pacifi c Ocean sea fl ow and ENSO. Journal of Geophysical Research level (Davao and Darwin). ACKNOWLEDGEMENTS 101:12,255–12,264. Meyers, G., R.J. Bailey, and A.P. Worby. 1995. Volume It is interesting to note that the best fi t The author gratefully acknowledges Drs. transport of Indonesian throughfl ow. Deep Sea to total transport is found with sea level Roger Lukas and Gary Mitchum for in- Research Part I 42:1,163–1,174. Potemra, J.T., S.L. Hautala, and J. Sprintall. 2003. at South Java both contemporaneous spiring the initial idea in an earlier pa- Vertical structure of Indonesian throughfl ow in a (indicative of local or near-local forcing) per, and Drs. Susan Hautala and Niklas large-scale model. Deep Sea Research II 50:2,143– 2,162. and one month prior (indicative of re- Schneider for encouraging pursuit of Potemra, J.T., R. Lukas, and G. Mitchum. 1997. motely forced coastal Kelvin waves from this revised paper. Dr. Ben Giese gener- Large-scale estimation of transport from the Pa- cifi c to the Indian Ocean. Journal of Geophysical the Indian Ocean), as well as at Darwin ously provided the SODA output, and Research 102:27,795–27,812. leading by one month (again indica- the APDRC at the IPRC serves this to the Saji, N.H. and T. Yamagata. 2003. Possible impacts of Indian Ocean Dipole mode events on global tive of remotely forced coastal Kelvin research community. The diligent work climate. Climate Research 25:151–169. waves but from the Pacifi c Ocean), and of Dr. Arnold Gordon and three anony- Saji, N.H., B.N. Goswami, P.N. Vinayachandran, and T. Yamagata. 1999. A dipole mode in the tropical at Davao. But, the best fi t for total trans- mous reviewers is sincerely appreciated. Indian Ocean. Nature 401:360–363. port occurs when sea level at Davao is This research was supported by NOAA Semtner, A.J., and R.M. Chervin. 1992. Ocean gener- al circulation from a global eddy-resolving mod- included with a one-month lag. In other through grant No. 654477 and by the el. Journal of Geophysical Research 97:5,493–5,550. words, processes determining total ITF Japan Agency for Marine-Earth Science Sprintall, J., A.L. Gordon, R. Murtugudde, and R. D. Susanto. 2000. A semi-annual Indian Ocean transport are correlated with sea level at and Technology (JAMSTEC) through forced Kelvin wave observed in the Indonesian Davao a month later. This correlation its sponsorship of the International Pa- seas in May 1997. Journal of Geophysical Research 105:17,217–17,230. could be an Indian Ocean effect, perhaps cifi c Research Center. This manuscript is Wijffels, S., and G. Meyers. 2004. An intersection of a result of ITF transport, which then has SOEST publication 6672 and IPRC pub- oceanic waveguides: Variability in the Indonesian throughfl ow region. Journal of Physical Oceanog- an effect on Pacifi c winds, which in turn lication 353. raphy 34:1,232–1,253. adjusts sea level at Davao. Wyrtki, K. 1987. Indonesian Throughfl ow and the associated pressure gradient. Journal of Geophysi- Finally, it should be noted that the REFERENCES cal Research 92:12,941–12,946. Carton, J.A., G. Chepurin, X. Cao, and B.S. Giese. sea-level locations of Clarke and Liu Wyrtki, K. 1961. Physical Oceanography of the 2000. A Simple Ocean Data Assimilation analy- Southeast Asian Waters, Naga Report 2, Scripps sis of the global upper ocean 1950–1995, Part 1: (1994), and those proposed by Wyrtki Institution of Oceanography, 195 pp. (1987) do, in fact, produce a good es- Methodology. Journal of Physical Oceanography 30:294–309. timate of upper-ocean ITF transport Clark, A., and X. Liu. 1994. Interannual sea level in
Oceanography Vol. 18, No. 4, Dec. 2005 107