Seasonal and Interannual Variability of Tidal Mixing Signatures in Indonesian Seas from High-Resolution Sea Surface Temperature

: With their complex narrow passages and vigorous mixing, the Indonesian seas provide the only low-latitude pathway between the Paciﬁc and Indian Oceans and thus play an essential role in regulating Paciﬁc-Indian Ocean exchange, regional air-sea interaction, and ultimately, global climate phenomena. While previous investigations using remote sensing and numerical simulations strongly suggest that this mixing is tidally driven, the impacts of monsoon and El Niño Southern Oscillation (ENSO) on tidal mixing in the Indonesian seas must play an important role. Here we use high-resolution sea surface temperature from June 2002 to June 2021 to reveal monsoon and ENSO modulations of mixing. The largest spring-neap (fortnightly) signals are found to be localized in the narrow passages/straits and sills, with more vigorous tidal mixing during the southeast (boreal summer) monsoon and El Niño than that during the northwest (boreal winter monsoon) and La Niña. Therefore, tidal mixing, which necessarily responds to seasonal and interannual changes in stratiﬁcation, must also play a feedback role in regulating seasonal and interannual variability of water mass transformations and Indonesian throughﬂow. The ﬁndings have implications for longer-term variations and changes of Paciﬁc–Indian ocean


Introduction
With their complex geography and narrow passages, the Indonesian seas provide the only pathway for low-latitude Pacific Ocean water to flow into the Indian Ocean ( Figure 1). Transport and water-mass transformation associated with the Indonesian throughflow (ITF) directly impacts the heat and freshwater budgets of the Pacific and Indian Oceans, and influences the El Niño Southern Oscillation (ENSO) and Asian-Australian monsoons, e.g., [1][2][3]. Furthermore, the Indonesian seas are a prime location of atmospheric convection that drives the Walker Circulation. Hence any changes in sea surface temperature (SST) within these regions will have global impacts on weather and climate events, e.g., [4][5][6]. Ocean mixing in the Indonesian seas affects the mean state of the Indo-Pacific region and its interannual variability [7].
Along the ITF pathways, waters experience strong tidal mixing and air-sea interactions. The detailed geography of nonlinear interactions between tides and tidally-induced mixing, and the influence of the ITF and strait geometry in the Indonesian seas are complicated and not fully understood. The temperature and salinity stratification are significantly altered within the Indonesian Seas' complex topography by turbulent mixing processes [2,8]. Although the core of the ITF is at about 100 m depth, the mixing drives Sea Surface Temperature (SST) changes by several • C [2]. While many aspects of the water-mass modification remain poorly understood, it is clear that tidally-driven mixing is a crucial Along the ITF pathways, waters experience strong tidal mixing and air-sea interactions. The detailed geography of nonlinear interactions between tides and tidally-induced mixing, and the influence of the ITF and strait geometry in the Indonesian seas are complicated and not fully understood. The temperature and salinity stratification are significantly altered within the Indonesian Seas' complex topography by turbulent mixing processes [2,8]. Although the core of the ITF is at about 100 m depth, the mixing drives Sea Surface Temperature (SST) changes by several °C [2]. While many aspects of the watermass modification remain poorly understood, it is clear that tidally-driven mixing is a crucial component [2,[8][9][10][11][12][13][14]. Ffield and Gordon [15] argued for tidal mixing by identifying a pronounced spring-neap tidal cycle in satellite SST data centered on the Banda Sea. A more recent satellite study by Ray and Susanto [16], based on much more superior SST measurements, produced significantly different results with the strongest fortnightly SST signals localized to relatively small straits, channels, and sills-Sulu Sill and exit passages of the Indonesian seas from Bali to the Timor Islands. The analysis of Ray and Susanto [16] identified little significant fortnightly SST signals in the central Banda Sea, which is consistent with in situ microstructure measurements of weak mixing there [17]. It also Using a numerical model that includes tides, Nugroho et al. [22] show a similar map of fortnightly variability in SST. All the mixing hotspots occur in regions that display strong M 2 current velocities [23] and where the model of Nagai and Hibiya [24] predicts significant dissipation of baroclinic tidal energy.
Throughout this paper, we will use MSf as shorthand for the 14.77-days beat period between the M 2 and S 2 semidiurnal tides (i.e., spring-neap cycle). The long-period MSf tidal constituent has the same frequency, but it plays no role here as it is very weak [21,25]. Even when appearing as a nonlinear compound tide, MSf tidal currents are far too weak to affect stratification. Any mixing at the MSf period is the result of nonlinearity arising from the fortnightly modulations of M 2 by the solar S 2 .
Remote Sens. 2022, 14, 1934 3 of 14 At longer periods, ocean-atmosphere dynamics of the Indo-Pacific Maritime continent are strongly affected by oceanic/atmospheric climatic processes associated with Madden-Julian Oscillation (MJO), monsoon, ENSO, and the Indian Ocean Dipole (IOD). The most substantial SST variability is associated with the Asian-Australian monsoon [5,26,27]. Stratification in this region is highly dependent on these climatic conditions, such as the phase of ENSO or the phase of the monsoon season. Any stratification changes will, of course, potentially change tidal mixing. Although previous studies have shown intense mixing occurrences within the Indonesian seas and their impacts of water mass transformation into the Indian Ocean, little effort has been made to investigate the modulation of monsoon and ENSO on tidal mixing. Our understanding of tidal mixing within the Indonesian seas has been hampered by a lack of in situ and long-term observational time series.
Here we use long-term, high-spatiotemporal resolution SST to investigate monsoon and ENSO modulation of tidal mixing within the Indonesian seas. Our previous results [16,21] primarily show the mean mixing signal. Fortunately, the SST time series are now sufficiently long that we can begin to investigate some of these monsoon and ENSO modulations, at least in regions where the mixing signal is robust.

Expected Tidal Mixing Frequencies in Sea Surface Temperature
The primary approach of analysis of SST for tidal mixing signature is to decompose SST time series using harmonic analysis and search for spring-neap and monthly cycles. Under an assumption that the nonlinear product (e.g., quadratic) of tidal current velocities is proportional to changes in ocean temperature due to tidal mixing, the more intense the tidal current, the colder the SST [15,16,21,28]. Suppose the temperature of a region is influenced by the interaction between two tidal constituents with combined currents: Taking the square of this time series will generate a nonlinear term with new frequencies of 2ω M2 , 2ω S2 , (ω M2 + ω S2 ), and |ω M2 − ω S2 |: where A 1 and A 2 are the amplitudes of M 2 and S 2 , respectively, and ω M2 and ω S2 are the angular frequency of M 2 and S 2 , respectively. The lower frequency |ω M2 − ω S2 | has a period of 14.77 days (fortnightly, MSf). Similarly, the square of SST time series of the diurnal luni-solar tide (K 1 ) and diurnal lunar tide (O 1 ) will have a modulation envelope with a period of 13.66 days (near-fortnightly signal, Mf). The strength of semidiurnal and diurnal cycles in the region dictates the spring-neap cycle of either MSf or Mf signals. For example, Indonesian seas exhibit mostly an MSf signal while the South China Sea displays a substantial Mf signal. For some regions, such as New England's coast, the N 2 tidal constituent is larger than S 2 . Therefore, the square of N 2 and M 2 would generate a monthly tidal envelope with 27.55 days [16].
The following section discusses the harmonic analysis of SST to extract spring-neap MSf tidal signals and determine monsoon and ENSO modulations.

Sea Surface Temperature Data and Method
In remote sensing of SST, there are generally two types of sensors: infrared and microwave. The infrared measurements provide high spatial resolutions on the order of 1-2 km, i.e., the Multiscale Ultrahigh Resolution SST (MURSST) product [29]. These data were used by Ray and Susanto [16,21] and Susanto et al. [28] to extract tidal mixing signals in Indonesian seas and Hong Kong coastal waters, respectively. However, infrared measurements are often obstructed by clouds, which can be problematic in some regions such as Southeast Asia.
Microwave sensors provide data in nearly all weather conditions but provide considerably coarser spatial resolution on the order of 25-50 km [30]. They require a mask of 50-75 km around the land, depending on the sensor specifics, owing to the large microwave footprint and the possible contamination when land falls within the antenna sidelobes, e.g., [31].
This study uses the global daily gridded compilation from the Group for High-Resolution Sea Surface Temperature (GHRSST) [32,33]. The GHRSST products aim to achieve the best combination of these different systems, a combination of infrared and microwave sensors with a spatial resolution of 0.09 • × 0.09 • . The data were available from the NASA-PODAAC from June 2002 to the present [34] and were accessed on 28 July 2021.
The SST time series from June 2002 to June 2021 were partitioned into segments according to the monsoon or ENSO phases, and then harmonic analysis was applied separately to each. Within the Indonesia Maritime Continent, the monsoon phase is usually defined based on the combination of annual rainfall patterns and winds, i.e., [35]. Southeast monsoon (SEM) is defined as April to September, characterized by dry air and less rainfall. Meanwhile, the northwest monsoon (NWM) is characterized by high humidity and wet conditions from October to March. For the ENSO phase, we use the Oceanic Niño Index (ONI; https://www.cpc.ncep.noaa.gov/data/indices/and accessed on 1 September 2021). El Niño defines as ONI ≥ 0.5 • C and La Niña defines as ONI ≤ −0.5 • C. Figure Figure 3 shows the main results of extracting 14.77-day MSf amplitude from the SST time series from June 2002 to June 2021 for the southeast monsoon (boreal summer) and northwest monsoon (boreal winter). During southeast monsoon, easterly wind-induced upwelling occurs along the southern coasts of the Lesser Sunda Island chain from Java to Timor and Banda Sea with clearly distinctive SST pattern ( Figure 2A). Hence, stronger  Figure 3 shows the main results of extracting 14.77-day MSf amplitude from the SST time series from June 2002 to June 2021 for the southeast monsoon (boreal summer) and northwest monsoon (boreal winter). During southeast monsoon, easterly wind-induced upwelling occurs along the southern coasts of the Lesser Sunda Island chain from Java to Timor and Banda Sea with clearly distinctive SST pattern ( Figure 2A). Hence, stronger stratifications in the exit passages of ITF in the Lombok, Sape, and Ombai Straits translate to stronger mixing signatures, as seen in Figure 3A. Meanwhile, during the northwest monsoon, the westerly winds from Asia/South China Sea do not generate a distinct pattern within the Indonesia seas (except in the South China Sea). The relationship between SST, monsoonal winds, and tidal mixing has been investigated using a numerical model [27,36,37]. Their results (especially Nagai and Hibiya) [37] suggested that strong tidal mixing in the narrow tidal straits between the Lesser Sunda Islands creates a cold SST signal that propagates as baroclinic Kelvin waves along the Lesser Sunda Islands. The Ekman transport associated with monsoonal winds regulates the northward and/or southward propagation of these baroclinic Kelvin waves. Their results are consistent with our results. Figure 3 shows stronger tidal mixing signatures during the boreal summer than in the boreal winter. The vigorous mixing locations are consistent with previous results [16]. They are primarily concentrated in the narrow exit passages of ITF into the Indian Ocean along the Lesser Sunda Island chain and Sulu Sill. Moderate mixings are observed around Halmahera/Ceram Sea, Lifamatola passage, and Manipa Strait. Meanwhile, the interior Banda Sea shows a weak mixing signature.

Results
These results are consistent with in situ data of stratification based on historical Indonesian throughflow measurements of Conductivity Temperature and Depth (CTD) taken in June 2005 to represent the southeast monsoon ( Figure 4A), and in December 2019 to represent the northwest monsoon ( Figure 4B). The southward ITF flow of waters entering the Lombok Strait is strongest and stratified during the southeast monsoon. Along pathways in the Lombok Strait from the northern entrance to the southern side of the sill, TS profiles are nearly the same. However, when the water reaches south of Lombok Strait's sill, the temperature-salinity (TS) properties become nearly linear, with the salinity maximum and the deeper salinity minimum mixed away, especially from σ θ = 23.5 kg/m 3 to σ θ = 27.25 kg/m 3 . On the other hand, ITF southward flow to the Indian Ocean is the lowest during the northwest monsoon, even sometimes undergoing northward flow from the Indian Ocean into the Indonesian seas [38][39][40][41]. The TS properties are nearly linear from the north to the south of the Lombok Strait.
The main ITF pathway from the western Pacific enters the Sulawesi Sea into the Makassar Strait. It then splits into two pathways: the direct exit into the Indian Ocean via the Lombok Strait and the rest toward the Banda Sea ( Figure 5). In situ observations show significant water-mass modification occurring in the narrow southern passages where the flow exits the Indonesian archipelago, i.e., [40]. TS curves from CTD measurements show a continual erosion of a salinity maximum centered around σ θ = 24 kg/m 3 from north to south through the Makassar and Lombok Straits ( Figure 5B,C). Between the CTD taken in the middle of Lombok Strait and one south of Lombok Strait, the TS properties become nearly linear, with the salinity maximum and the deeper salinity minimum (centered on σ θ = 26.5 kg/m 3 ) having been mixed away. This abrupt change in water-mass properties indicates elevated mixing over the sills at the southern end of Lombok Strait. Remote sensing and numerical simulations strongly suggest that this mixing is tidally-driven [16,21,22,24].    to σθ = 27.25 kg/m 3 . On the other hand, ITF southward flow to the Indian Ocean is the lowest during the northwest monsoon, even sometimes undergoing northward flow from the Indian Ocean into the Indonesian seas [38][39][40][41]. The TS properties are nearly linear from the north to the south of the Lombok Strait. The main ITF pathway from the western Pacific enters the Sulawesi Sea into the Makassar Strait. It then splits into two pathways: the direct exit into the Indian Ocean via the Lombok Strait and the rest toward the Banda Sea ( Figure 5). In situ observations show significant water-mass modification occurring in the narrow southern passages where the flow exits the Indonesian archipelago, i.e., [40]. TS curves from CTD measurements show a continual erosion of a salinity maximum centered around σθ = 24 kg/m 3 from north to south through the Makassar and Lombok Straits (Figures 5B,C). Between the CTD taken in the middle of Lombok Strait and one south of Lombok Strait, the TS properties become nearly linear, with the salinity maximum and the deeper salinity minimum (centered on σθ = 26.5 kg/m 3 ) having been mixed away. This abrupt change in water-mass properties indicates elevated mixing over the sills at the southern end of Lombok Strait. Remote sensing and numerical simulations strongly suggest that this mixing is tidally-driven [16,21,22,24]. Note that a relatively strong signature is seen in Figure 3 at the southern sill of the Sulu Sea for both seasons, but unlike the Lesser Sunda Island chain, the signature is more pronounced in the northwest season.
At Lombok Strait, the largest spring-neap SST amplitudes during the southeast monsoon ( Figure 3A) and the annual average occur almost 75 km south of the sill. SST spectra south of Lombok and Alas Straits show narrow MSf peaks well above the background ( Figure 6A). However, an SST spectrum at the northern end of Lombok Strait shows no fortnightly enhancement (green line in Figure 6A), despite there being a clear MSf signal in velocity at the INSTANT (International Nusantara Stratification and Transport) [38] mooring ( Figure 6B). Aiki et al. [42] argued that the southward ITF causes the asymmetry by displacing internal tides generated at the sill southward. If southward throughflow Note that a relatively strong signature is seen in Figure 3 at the southern sill of the Sulu Sea for both seasons, but unlike the Lesser Sunda Island chain, the signature is more pronounced in the northwest season.
At Lombok Strait, the largest spring-neap SST amplitudes during the southeast monsoon ( Figure 3A) and the annual average occur almost 75 km south of the sill. SST spectra  Figure 6A). However, an SST spectrum at the northern end of Lombok Strait shows no fortnightly enhancement (green line in Figure 6A), despite there being a clear MSf signal in velocity at the INSTANT (International Nusantara Stratification and Transport) [38] mooring ( Figure 6B). Aiki et al. [42] argued that the southward ITF causes the asymmetry by displacing internal tides generated at the sill southward. If southward throughflow currents are stronger than the tidal group speed at the sill, they will block propagation to the north to produce the north-south asymmetry observed in the fortnightly SST signatures. When coastal-trapped Kelvin waves propagate from the west along the south coast of Java, ITF diminishes and even reverses [43]. Figure 3B shows that the internal tide SST signature is substantially more symmetric during the northwest monsoons.  Figure 3A). Locations of observations for the spectrum analysis (red, green, and blue curves).
To determine the ENSO/IOD modulations of tidal mixing in the Indonesian seas, similar to our monsoon approach, we divided the SST time series into three classes, El Niño, La Niña, and Neutral, based on the ONI value of ± 0.5 °C threshold. Figure 7 shows the MSf amplitude of SST during the (a) El Niño year, (b) La Niña year, and (c) Neutral year. Figure 7 shows that the tidal mixings in the Sulu Sill and exit passages of ITF into the Indian Ocean are more vigorous during the El Niño year than during La Niña and Neutral years.  Figure 3A). Locations of observations for the spectrum analysis (red, green, and blue curves).
To determine the ENSO/IOD modulations of tidal mixing in the Indonesian seas, similar to our monsoon approach, we divided the SST time series into three classes, El Niño, La Niña, and Neutral, based on the ONI value of ± 0.5 • C threshold. Figure 7 shows the MSf amplitude of SST during the (a) El Niño year, (b) La Niña year, and (c) Neutral year. Figure 7 shows that the tidal mixings in the Sulu Sill and exit passages of ITF into the Indian Ocean are more vigorous during the El Niño year than during La Niña and Neutral years.  Based on Hybrid Coordinate Ocean Model (HYCOM), South China Sea throughflow is increased through the Sulu Sill during El Niño, inhibiting surface injection of tropical Pacific water into the Makassar Strait [44], and enhancing the stratification and the tidal mixing over the Sulu Sill (Figure 7). For the Lesser Sunda Island chain/exit passages of ITF into the Indian Ocean, more vigorous mixing occurs during El Niño and Neutral years, and is weaker during La Niña. Even though ITF in the Makassar Strait is weaker during El Niño and stronger during La Niña [44][45][46], ITF transport in the exit passages to the Indian Ocean remains strong during El Niño, concurrent with IOD positive. There is phase delayed of ITF transport in the outflow passage relative to the inflow passage due to ENSO variability, i.e., [47,48].
During the concurrent El Niño and IOD positive events, an anomalously strong easterly wind occurs along the southern coast of the Lesser Sunda Island chain, inducing intense upwelling, colder temperature, shallower thermocline depth, lower sea surface height, and high chlorophyll-a concentration, i.e., [39,[49][50][51][52][53][54][55]. For example, prolonged anomalously early winds during concurrent El Niño 2006 and IOD events lowered sea level on the Indian Ocean side, and enhanced and extended the ITF southward flow until December 2006, i.e., [38,39,49]. This suggests that Indian Ocean dynamics may be more important than the Pacific Ocean dynamics in controlling the outflow transport and stratification during the concurrent El Niño and IOD events [13]. Therefore, these interannual events may suppress the intraseasonal and seasonal signals, and induce stronger stratification and tidal mixing signatures, as seen during the boreal summer in Figure 3A.
The aforementioned suggests that ENSO classification may not always be the most appropriate in all regions, e.g., [35]. Indonesian seas are affected by both ENSO and IOD events, and Indian Ocean dynamics may, at times, win over the Pacific Ocean dynamics [13]. Generally, El Niño is concurrent with IOD positive, and La Niña is concurrent with IOD negative. However, both events do not always remain in phase or perfectly aligned ( Figure 8). Analysis based on IOD alone may be useful in some regions. As the SST time series continues to lengthen, we may classify the interannual variability based on IOD only and both ENSO and IOD conditions. Based on Hybrid Coordinate Ocean Model (HYCOM), South China Sea throughflow is increased through the Sulu Sill during El Niño, inhibiting surface injection of tropical Pacific water into the Makassar Strait [44], and enhancing the stratification and the tidal mixing over the Sulu Sill (Figure 7). For the Lesser Sunda Island chain/exit passages of ITF into the Indian Ocean, more vigorous mixing occurs during El Niño and Neutral years, and is weaker during La Niña. Even though ITF in the Makassar Strait is weaker during El Niño and stronger during La Niña [44][45][46], ITF transport in the exit passages to the Indian Ocean remains strong during El Niño, concurrent with IOD positive. There is phase delayed of ITF transport in the outflow passage relative to the inflow passage due to ENSO variability, i.e., [47,48].
During the concurrent El Niño and IOD positive events, an anomalously strong easterly wind occurs along the southern coast of the Lesser Sunda Island chain, inducing intense upwelling, colder temperature, shallower thermocline depth, lower sea surface height, and high chlorophyll-a concentration, i.e., [39,[49][50][51][52][53][54][55]. For example, prolonged anomalously early winds during concurrent El Niño 2006 and IOD events lowered sea level on the Indian Ocean side, and enhanced and extended the ITF southward flow until December 2006, i.e., [38,39,49]. This suggests that Indian Ocean dynamics may be more important than the Pacific Ocean dynamics in controlling the outflow transport and stratification during the concurrent El Niño and IOD events [13]. Therefore, these interannual events may suppress the intraseasonal and seasonal signals, and induce stronger stratification and tidal mixing signatures, as seen during the boreal summer in Figure 3A.
The aforementioned suggests that ENSO classification may not always be the most appropriate in all regions, e.g., [35]. Indonesian seas are affected by both ENSO and IOD events, and Indian Ocean dynamics may, at times, win over the Pacific Ocean dynamics [13]. Generally, El Niño is concurrent with IOD positive, and La Niña is concurrent with IOD negative. However, both events do not always remain in phase or perfectly aligned ( Figure 8). Analysis based on IOD alone may be useful in some regions. As the SST time series continues to lengthen, we may classify the interannual variability based on IOD only and both ENSO and IOD conditions.

Conclusions and Discussions
With their complex geography and narrow passages, the Indonesian seas provide the only pathway for low-latitude Pacific Ocean water to flow into the Indian Ocean. Transport and water-mass transformation associated with the ITF directly impacts the heat and freshwater budgets of the Pacific and the Indian Ocean. In situ observations show extensive water-mass modification occurring in the narrow southern passages where the flow exits the Indonesian archipelago.
The influence of significant climatic phenomena associated with Southeast Asia monsoon and longer time scales associated with ENSO or Indian Ocean Dipole can be expected to affect tidal mixing in the Indonesian seas. In fact, Koch-Larrouy et al. [7] show that the

Conclusions and Discussions
With their complex geography and narrow passages, the Indonesian seas provide the only pathway for low-latitude Pacific Ocean water to flow into the Indian Ocean. Transport and water-mass transformation associated with the ITF directly impacts the heat and freshwater budgets of the Pacific and the Indian Ocean. In situ observations show extensive water-mass modification occurring in the narrow southern passages where the flow exits the Indonesian archipelago.
The influence of significant climatic phenomena associated with Southeast Asia monsoon and longer time scales associated with ENSO or Indian Ocean Dipole can be expected to affect tidal mixing in the Indonesian seas. In fact, Koch-Larrouy et al. [7] show that the interaction goes both ways: they show that mixing in the Indonesian seas affects eastern equatorial Pacific temperature and thereby affects ENSO itself. This paper has attempted to identify monsoon and ENSO influence on tidal mixing in the Indonesian seas using high-resolution SST. The GHRSST data from 2002 to 2021 provides a map of seasonal and interannual variability of tidal mixing signatures in the Indonesian seas, which reconfirm average tidal mixing signatures derived from Multi-Ultra High-Resolution SST (MURSST) [16,21].
The most robust SST signatures are all localized to narrow straits and sills where the semidiurnal currents are very strong [23]. These regions are primarily associated with the exit passages of ITF into the Indian Ocean along the Lesser Sunda Island chain from Bali to Timor Islands, Sulu Sill, Lifamatola passage, and Manipa Strait. Most of these regions show vigorous tidal mixing during both seasons (boreal summer and winter), but boreal summer shows a stronger signal. For interannual variability, stronger tidal mixing signatures are observed during the El Niño and Neutral years, while the La Niña year shows moderate signals.
While the boreal summer finds a more significant tidal mixing signature in the Lombok Strait than boreal winter, tidal mixing signatures in the Sulu Sill are the opposite. More vigorous mixing was observed in the Sulu during boreal winter than in summer. This is consistent with remote sensing observations of tidally-induced internal waves in the two regions, i.e., [56,57]. Higher internal wave activities were observed in Sulu Sill during boreal winter than in the boreal summer, i.e., [58]. The Sebutu passage over the Sulu Sill exhibits stronger southward flow toward the Sulawesi Sea from August to December. Reversal northward flow was observed from December to July [59]. Tidally-generated solitons near exit passage sills of the Indonesian seas [14,42,56,60,61] may be responsible for the tidal mixing inferred from satellite SST. Isopycnal heaving can exceed 100 m in 6 min ( [56] their Figure 3). Based on synthetic aperture radar (SAR) data, arc-like internal waves are frequently detected during boreal winter to spring when the southward throughflow is low or absent [61]. The waves may be responsible for the nearly linear stratification seen in Figure 4B. On the other hand, during boreal summer, when the southward throughflow is strongest, irregular internal waves frequently occur, which may enhance more vigorous mixing [61].
The analysis of Ray and Susanto [16], as well as the figures shown here, identified little significant fortnightly SST signal in the central Banda Sea, which is consistent with in situ microstructure measurements of weak mixing in the Banda Sea [17,19,20]. Moderate fortnightly SST signals in the Lifamatola Passage and Manipa Strait during both monsoons are consistent with in situ measurements showing intense tidal mixing with dissipation rates ranging from 10 −7 to 10 −6 W kg −1 [18,19]. Along the southern coasts of the Nusa Tenggara Island chain, especially south of the Lombok Strait, Sape Strait/Sumba Sea, and Alor/Savu Sea, even stronger fortnightly SST signals can be seen. Lombok Strait is collocated with the most significant water-mass changes ( Figures 4A and 5C). This result is consistent with in situ measurements in the Lombok Strait and Sape Strait [19].
Although there is compelling remote sensing evidence for vigorous tidal mixing along the exit passages of ITF or the Nusa Tenggara Island chain, this critical process has not been sufficiently explored with in situ observations. Nagai et al. [19] measured mixing in the Lombok and Sape Straits using a microstructure profiler, although their measurements were limited to two or three profiles at each station (ten near the Lombok Sill), so did not cover a complete spring-neap cycle as observed by Alford et al. [17] in the Banda Sea. Recently, Susanto led extensive tidal mixing measurements in collaboration with scientists from Indonesia and China in the south of the Lombok Strait for fifteen consecutive days. The data are being analyzed and will be reported in future publications. It would be ideal to use shipboard and moored observations that cover boreal summer and winter to similarly explore in more detail the underlying processes and variability. Validating these results in the Lombok Strait will open up the possibility of using remotely sensed SST data to estimate tidal mixing in other regions around the global ocean. Data Availability Statement: All remotely sensed data are publicly available from the NASA-PODAAC (https://podaac.jpl.nasa.gov/) accessed on 28 July 2021, while the in situ CTD data are available upon request. The Oceanic Niño Index is https://www.cpc.ncep.noaa.gov/data/indices/ and was accessed on 1 September 2021. The Dipole Mode Index (DMI) is publicly available from https://stateoftheocean.osmc.noaa.gov/sur/ind/dmi.php and was accessed on 1 September 2021.