Surface Seawater p CO 2 Variation after a Typhoon Passage in the Kuroshio off Eastern Taiwan

: In this study, two cruises were conducted across the mainstream of the Kuroshio off eastern Taiwan before and after the passage of Typhoon Saola in summer 2012. The continuous underway p CO 2 (i.e., partial pressure of CO 2 ) measurements revealed that surface seawater p CO 2 (SS p CO 2 ) displayed spatial variations in response to typhoon passage. The simulated results showed that the mixed-layer deepening after typhoon passage had a minor effect on SS p CO 2 variation because p CO 2 decrease driven by temperature dropdown and enhanced biological production fueled by nutrients input was largely compensated by p CO 2 increase caused by salinity increase and dissolved inorganic carbon input from the subsurface layer. By contrast, the advection pattern showed signiﬁcant change before and after the typhoon, which could play a major role in controlling the variation of SS p CO 2 . In the exit area of the cyclonic eddy, SS p CO 2 decreased, while in the area of its arrival, SS p CO 2 increased. Besides, the discharge of freshwater and the intrusion of the South China Sea subsurface could result in SS p CO 2 increase in the nearshore area. The present study highlights that more advection changes need to be considered to better understand the impact of the typhoon on SS p CO 2 , especially in the strong current area, such as the Kuroshio. Author Contributions: Conceptualization, L.-F.F. and W.-C.C.; Data curation, L.-F.F., C.H.C., G.-C.G. and W.-C.C.; Formal analysis, L.-F.F. and C.H.C.; Funding acquisition, W.-C.C.; Investigation, W.-C.C.; Methodology, C.H.C. and G.-C.G.; Resources, W.-C.C.; Software, C.H.C.; Validation, C.H.C. and W.-C.C.; Visualization, L.-F.F. and C.H.C.; Writing—original draft, L.-F.F., C.H.C. and W.-C.C.; Writing—review & editing, G.-C.G.


Introduction
Previous studies have demonstrated that typhoons (severe tropical storms also referred to as tropical cyclones or hurricanes) can cause dramatic hydrographic and biogeochemical impacts on marine ecosystems, such as cooling of sea surface temperature, upward mixing, and entrainment of deep salty water rich in nutrients and CO 2 , and the enhanced biological activities [1][2][3]. Typhoon-triggered entrainment/upwelling prompts strong vertical mixing and breaks stratification in the upper ocean, which can cause prominent sea surface cooling, thereby reducing partially pressure of CO 2 (pCO 2 ; [2][3][4][5]). The effect of salinity on pCO 2 is another thermodynamic manifestation of the changes in the solubility of CO 2 and the dissociation constants of carbonic acid in seawater [6]. For example, a wing pump by a tropical cyclone that supplies much salinity causing the increased sea surface pCO 2 was reported in the North Atlantic Ocean [7].
A number of studies have reported that hurricanes (or typhoons) increase surface pCO 2 and summertime CO 2 efflux, mainly due to upward mixing and entrainment of deep water enriched in CO 2 during the typhoon passage [2,[7][8][9][10]. In contrast, some recent observations have shown net uptake of sea surface CO 2 associated with the passage of typhoons over the continental shelf and continental margin, as a result of the enhancements in primary production fueled by the nutrients brought up from the deep water [3].
These previous studies have shown that the impact of typhoons on surface seawater pCO 2 (SS pCO 2 ) varies significantly among different regions. Therefore, observations in

Study Sites and Sampling
The Kuroshio is the western boundary current of the North Pacific's subtropical gyre, originating from the northward bifurcation of the North Equatorial Current off the east coast of Luzon and flowing more than 3000 km past Taiwan to the southeast coast of Japan [11]. Off east of Taiwan, the Kuroshio is a relatively stable current with a climatological width of 100 km and a maximum velocity of~1 ms −1 at 15 m depth [12]. Additionally, Kuroshio is also known to considerably change its water property in its onshore flank due to the South China Sea (SCS) subsurface outflow as it bypasses the Luzon Strait [13].
In this study, nine hydrological stations along a transect across the Kuroshio off the east coast of Taiwan were visited twice aboard the R/V Ocean Research II ( Figure 1). The two cruises were conducted from 26 to 30 July and from 4 to 6 August 2012, respectively. During this period, typhoon Saola, a strong tropical cyclone with the highest winds up to 130 km h −1 and the lowest pressure approaching 960 kPa, passed through the study area from 31 July to 3 August. Therefore, the first and the second cruise were used to represent the pre-typhoon (4 days before) and post-typhoon (3 days after) conditions. During these cruises, the current velocity along the ship track was measured by a 150 kHz shipboard acoustic Doppler current profiler (ADCP; Teledyne RD Instruments Inc., San Diego, CA, USA). The underway pCO 2 measurements were made with a continuous flow equilibration system (AS-P2, Apollo SciTech Inc., Newark, DE, USA), which pumped seawater into the system from an inlet at 3-4 m on R/V Ocean Researcher II. The equilibrated headspace gas was dried by a Peltier cooler and a dry tube filled with magnesium perchlorate (Mg(ClO 4 ) 2 ) before it flowed into a nondispersive infrared spectrometer (LI-COR 7000), which was calibrated every 4 h against a CO 2 -free reference gas (N 2 ) and three gas standards (the xCO 2 of the three standards was 146.6, 363.2, and 523.2 ppm, respectively, traceable to NIST, National Institute of Standards and Technology). The measured xCO 2 data were converted into surface water pCO 2 by correcting to 100% humidity and to in situ water temperature, according to [14]. The accuracy and precision of the system were better than 2 ppm. Surface seawater temperature and salinity were measured just before the equilibration by the thermosalinograph (SBE 45, Sea-Bird Electronics, Bellevue, WA, USA).
At each hydrographic station, depth profiles of temperature and salinity were recorded using a Seabird SBE9/11-plus conductivity-temperature-depth (CTD) system. Discrete water samples were collected at twelve depths of 3,10,25,50,75,100,125,150,175,200,250, and 300 m, using 20 L Go-Flo bottles mounted onto a rosette sampling assembly. Subsamples for DIC and TA analyses were transferred into 350 mL pre-cleaned borosilicate bottles, and 200 µL of HgCl 2 -saturated solution was immediately added. Subsamples for the determination of nitrate concentrations were placed in 100 mL polypropylene bottles and frozen with liquid nitrogen immediately. Water samples for Chl a analysis were quickly filtered through a GF/F filter paper (Whatman, 47 mm, MERCK, Darmstadt, Germany), and the filter paper was stored at −20 • C. At each hydrographic station, depth profiles of temperature and salinity were recorded using a Seabird SBE9/11-plus conductivity-temperature-depth (CTD) system. Discrete water samples were collected at twelve depths of 3, 10, 25, 50, 75, 100, 125, 150, 175, 200, 250, and 300 m, using 20 L Go-Flo bottles mounted onto a rosette sampling assembly. Subsamples for DIC and TA analyses were transferred into 350 mL pre-cleaned borosilicate bottles, and 200 μL of HgCl2-saturated solution was immediately added. Subsamples for the determination of nitrate concentrations were placed in 100 mL polypropylene bottles and frozen with liquid nitrogen immediately. Water samples for Chl a analysis were quickly filtered through a GF/F filter paper (Whatman, 47 mm, MERCK, Darmstadt, Germany), and the filter paper was stored at −20 °C. The satellite-measured sea level anomalies (SLAs) and estimated absolute geostrophic currents were obtained from the CMEMS to study synoptic ocean currents and eddies. These satellite data are provided at 1/4° resolution, available daily since the end of 1992. To compare the observed temperature-salinity (T-S) properties of water with those climatological fields, we used the dataset from synoptic monthly gridded three-dimensional World Ocean Atlas version 2018, at 1° resolution, provided by the National Oceanic and Atmospheric Administration.

Analytical Methods
The measurements of TA, DIC, and pH followed the standard methods reported in The satellite-measured sea level anomalies (SLAs) and estimated absolute geostrophic currents were obtained from the CMEMS to study synoptic ocean currents and eddies. These satellite data are provided at 1/4 • resolution, available daily since the end of 1992. To compare the observed temperature-salinity (T-S) properties of water with those climatological fields, we used the dataset from synoptic monthly gridded three-dimensional World Ocean Atlas version 2018, at 1 • resolution, provided by the National Oceanic and Atmospheric Administration.

Analytical Methods
The measurements of TA, DIC, and pH followed the standard methods reported in [15] and were consistent with those used in our previous studies [16,17]. TA was measured using Gran titration on an automatic TA titrator (AS-ALK2, Apollo SciTech Inc.). DIC was measured using the nondispersive infrared method on a DIC analyzer (AS-C3, Apollo SciTech Inc.). The alkalinity titrator and DIC analyzer were calibrated using certified reference materials (CRMs) obtained from Dr. A. Dickson's laboratory at the Scripps Institution of Oceanography. The accuracy and precision for both analytical TA and DIC measurements were ±0.15% or better. Determination of pH was based on the method described [18] and spectrophotometrically determined at 25 • C, with a precision of 0.005 on the total hydrogen ion concentration scale. Nitrate was analyzed with a custom-made flow injection analyzer. The precision for the nitrate analysis was 0.3 µmol kg −1 , based on six duplicated measurements of a reference solution with a concentration of 10 µmol kg −1 . Chl a retained on the GF/F filters was determined fluorometrically on a Turner Design 10-AU-005 field fluorometer. Detailed descriptions of nitrate and Chl a analyses have been described earlier [19].

Statistical Analyses
Differences in the means of temperature, salinity, in situ pCO 2 , and npCO 2 at 29.1 • C before and after the typhoon passage were assessed using a Two-Sample t-test [20]. All statistical analyses were performed with a TTEST procedure in the SAS 9.4 software package [21].

Results
3.1. Spatial Variations of Temperature, Salinity, and pCO 2 in the Sea Surface Water before and after the Passage of Typhoon Saola The spatial variations of temperature, salinity, and pCO 2 in the sea surface water (3-4 m) before and after the passage of typhoon Saola are shown in Figure 2. Before and after the typhoon passage, sea surface temperature (SST) generally increased eastwards from the nearshore area to~122.5 • E and to 122.0 • E, respectively, and then decreased tõ 123 • E, and gradually increased again to the far coastal zone. The variation range was 28.1 to 30.9 • C and 28.6 to 29.5 • C before and after the typhoon passage, respectively ( Figure 2A). Generally, SST was not significantly different in the area west of~122.0 • E between the typhoon passage, while it decreased significantly by an average of ∼0.6 • C (29.0 ± 0.2 • C vs. 29.6 ± 0.6 • C; p < 0.01) in the area east of~122.0 • E after typhoon passage. SSS increased eastwards from the nearshore area to~122.0 • E and then remained at a relatively constant level of~33.7 and~34.2 before and after typhoon passage ( Figure 2B), respectively, in the east area of 122.0 • E, where the average SSS was significantly higher by~0.9 after typhoon compared with that before typhoon (34.0 ± 0.5 vs. 33.1 ± 0.5; p < 0.01). Mixed-layer depth (MLD), defined as the surface layer with a potential density variation less than 0.1 kg m −3 [19], generally deepened eastwards from the nearshore area to~122.5 • E and then shallowed to the far coastal zone. The variation range was 11 to 33 m and 19 to 59 m before and after the passage of the typhoon, respectively ( Figure 2C). In terms of average values, the MLD was significantly deeper after than those before the typhoon passage (41.8 ± 4.5 m vs. 21.8 ± 2.3 m; p < 0.05).
Water 2022, 14, x FOR PEER REVIEW 5 of 15 0.01), while the average npCO2 was significantly higher after the typhoon than that before typhoon (386 ± 9 μatm vs. 376 ± 12 μatm; p < 0.01). Depth distributions of temperature, salinity, DIC, TA, pH, nitrate, and Chl a before and after the typhoon passage at nine hydrological stations are shown in Figure 3. As a typical feature in the ocean, temperature generally decreases with increasing depth. On close examination, a striking feature, however, is readily seen in the subsurface layer (~50-200 m), where the temperature was higher after the typhoon than that before the typhoon at the same depth at stations one, six, seven, and eight, whereas temperature was lower after the typhoon than that before the typhoon at stations three and four ( Figure 3A). Salinity generally increased with increasing depth to a maximum of ~34.9 at the depth range between 100 to 150 m and then decreased slightly with depth. A sharper vertical salinity increasing trend was observed at the nearshore station (stations one and two; Figure 3B), suggesting the influence of freshwater input in the nearshore area. Similar to the vertical distribution of temperature, salinity in the subsurface layer also showed some remarkable variations before and after the typhoon passage. For example, salinity in the subsurface layer seemed to be higher after the typhoon than that before the typhoon at stations three and four, while it appeared to be lower at stations one, two, six, and seven.  Depth distributions of temperature, salinity, DIC, TA, pH, nitrate, and Chl a before and after the typhoon passage at nine hydrological stations are shown in Figure 3. As a typical feature in the ocean, temperature generally decreases with increasing depth. On close examination, a striking feature, however, is readily seen in the subsurface layer (~50-200 m), where the temperature was higher after the typhoon than that before the typhoon at the same depth at stations one, six, seven, and eight, whereas temperature was lower after the typhoon than that before the typhoon at stations three and four ( Figure 3A). Salinity generally increased with increasing depth to a maximum of~34.9 at the depth range between 100 to 150 m and then decreased slightly with depth. A sharper vertical salinity increasing trend was observed at the nearshore station (stations one and two; Figure 3B), suggesting the influence of freshwater input in the nearshore area. Similar to the vertical distribution of temperature, salinity in the subsurface layer also showed some remarkable variations before and after the typhoon passage. For example, salinity in the subsurface layer seemed to be higher after the typhoon than that before the typhoon at stations three and four, while it appeared to be lower at stations one, two, six, and seven.   For carbonate chemistry parameters, pH generally decreased with increasing depth ( Figure 3C), whereas DIC increased with depth ( Figure 3D). The vertical distribution of TA mimicked the pattern of salinity, increased gradually to a maximum at~100-150 m, and then decreased slightly with depth ( Figure 3E). The close correlation between TA and salinity suggests that salinity was the dominant factor controlling the vertical distribution of TA. Likewise, carbonate chemistry parameters in the subsurface layer also revealed some notable variations before and after typhoon passage. Generally, pH in the subsurface layer was higher after the typhoon than that before the typhoon at stations one, two, seven, and eight, but it was lower at stations three and four. In contrast, DIC in the subsurface layer was higher after typhoon than that before typhoon at stations three and four, but it was lower at stations one, two, seven, and eight.
Nitrate concentration was generally under the detection limit (0.2 µM) within the top 50 m and then increased with depth, while the surface low-nitrate layer expanded to the depth of approximately 100 m before typhoon ( Figure 3F), suggesting that the nitracline should be shallower after typhoon passage. Chl a concentration generally increased with depth before reaching the subsurface Chl a maximum (SCM), which was between 50 and 100 m, and then decreased sharply ( Figure 3G). However, the SCM was not found at station one before the typhoon passage.
In summary, the most striking variations in the vertical distributions of hydrological and carbonate parameters before and after typhoon passage were observed in the subsurface layer, which showed two contrasting patterns between stations 3 and 4 and stations 7 and 8. In stations 3 and 4, temperature and pH were lower, but salinity and DIC were higher after typhoon than those before the typhoon. In contrast, in stations 7 and 8, temperature and pH were higher, but salinity and DIC were lower after typhoon than those before the typhoon. This contrasting pattern in the variations of hydrological and carbonate parameters in the subsurface layer implies that there might be spatially various alterations in the horizontal advection pattern before and after typhoon passage, which has been further examined in the following Discussion section.

Typhoon Effects on Surface pCO 2
Typhoon generally leads to surface cooling, precipitation, and mixed-layer deepening, which may change temperature, salinity, and chemical characteristics such as TA, DIC, and nutrient concentrations in surface water, thereby influencing local variations of SS pCO 2 [2,3,[7][8][9]22,23]. To clarify the processes controlling the SS pCO 2 variations as typhoons pass through the study area, we quantified the respective contributions from temperature, salinity, TA and DIC input, and potential biological effect (i.e., nutrients input) to the observed SS pCO 2 variations at nine hydrographic stations by the following equations: where ∆pCO 2.Obs represents the observed pCO 2 difference before and after the typhoon passage (i.e., pCO 2.aft − pCO 2.bef ), which can be further expressed as the sum of For the calculation of ∆pCO 2.T , ∆pCO 2.S , ∆pCO 2.C , and ∆pCO 2.B , surface temperature and salinity, and the average DIC, TA, and NO 3 in the MLD (DIC .M , TA .M , NO 3.M ) before the typhoon was set as the initial condition [24]. ∆pCO 2.T , ∆pCO 2.S , and ∆pCO 2.C are computed based on the differences in temperature, salinity, and TA and DIC before (T .bef , S .bef , TA .M.bef , DIC .M.bef ) and after (T .aft , S .aft , TA .M.aft , DIC .M.aft ) typhoon passage on initial pCO 2 (Equations (2)-(4)). Biological effect (∆pCO 2.B ) was estimated based on the differences in NO 3.M.aft and NO 3.M.bef , and the subsequent uptake C/N ratio (R) by the stimulated biological production (Redfield ratio = 6.6, R in Equation (5)). Finally, ∆pCO 2.O represents the part of ∆pCO 2.Obs that cannot be explained by the changes in temperature, salinity, and TA, DIC, and nitrate inputs induced by the enhanced vertical mixing after typhoon passage, which is computed as the difference between ∆pCO 2.Obs and the sum of ∆pCO 2.T , ∆pCO 2.S , ∆pCO 2.C , and ∆pCO 2.B (Equation (6)).
The calculated results are shown in Figure 4 and Table 1. Because thermodynamically temperature decrease leads to pCO 2 reduction, the observed temperature decrease induced by the cooling effect and the enhancement of vertical mixing after typhoon passage would have produced negative ∆pCO 2.T values [2,3], ranging from −2 to −29 µatm with an average value of −13 µatm ( Figure 4B). The most pronounced temperature-induced pCO 2 decrease (∆pCO 2.T ) was observed at stations 3, 4, 5, and 6, located in Zones II and III, which also showed the largest temperature decrease ( Figure 2B). In contrast, the temperature decrease was generally smaller, and thus, ∆pCO 2.T was lower at the other stations in Zones I (stations 1 and 2) and IV (stations 7, 8, and 9). The larger temperature decrease in Zones II and III could be associated with the combined effect of MLD deepening ( Figure 4H) and the arrival of cold eddy ( Figure 5) after typhoon passage, while the smaller temperature decrease in Zones I and IV could be related to the movement of Kuroshio towards the shore and the disappearance of cold eddy after the typhoon passage ( Figure 5), respectively, both of which could partially compensate the temperature decrease induced by cooling effect and MLD deepening.
Since salinity generally increased with depth before reaching the salinity maximum depth ( Figure 3B), the enhanced vertical mixing would have resulted in an increase in SSS in the study area, except at station 1, where SSS decreased after typhoon passage, probably due to freshwater discharge ( Figure 2B). An increase in salinity would cause an increase in pCO 2 , and vice versa, by affecting the thermodynamic constants. Consequently, the calculated results showed positive ∆pCO 2.S values at all stations (except station 1) within a narrow range between 3 and 7 µatm, while ∆pCO 2.S was −2 µatm at station 1 ( Figure 4C). Compared with temperature, salinity had a minor effect on SS pCO 2 variations in response to typhoon passage.
The enhanced vertical mixing may bring subsurface water that is replete in DIC and TA into the surface layer and therefore may cause SS pCO 2 variations. The calculated results showed a generally positive ∆pCO 2.C at all stations (except station 8; Figure 4D), suggesting that the input of DIC may play a more significant role than TA input in regulating SS pCO 2 variations, providing that DIC increment would increase pCO 2, but TA increment would decrease pCO 2 . At station 8, MLD merely deepened 3 m after typhoon passage so that ∆pCO 2.C was very close to null.
The potentially subsequent elevated biological production stimulated by the nutrient input via enhanced vertical mixing had a nearly negligible impact on SS pCO 2 variation at stations 3 to 9, as shown by ∆pCO 2.B , which was very close to zero ( Figure 4E). This may reflect the fact that the nitracline is relatively deep in the study area, so the enhanced deepening of MLD could transport limited nitrate into the surface layer ( Figure 3F), which was further supported by the little change in Chl a concentration before and after the typhoon passage ( Figure 3G).    Table 1. Summary of the difference in surface seawater temperature (∆SST), surface seawater salinity (∆SSS), mixed-layer depth (∆MLD), pCO 2 (∆pCO 2 ), temperature-normalized npCO 2 (∆npCO 2 ), as well as average dissolved inorganic carbon (∆DIC), TA (∆TA), and nitrate (∆NO 3 ) in the mixed layer before and after the typhoon passage at the nine hydrological stations, which were grouped into four subzones based on the pCO 2 variations (Zone I, stations one and two; Zone II, station three; Zone III, stations four and five; and Zone IV, stations six, seven, eight, and nine; see text for the details). I  II  III  IV   Station  1  2  3  4  5  6  7  8  9 Parameter  Since salinity generally increased with depth before reaching the salinity maximum depth (Figure 3B), the enhanced vertical mixing would have resulted in an increase in SSS in the study area, except at station 1, where SSS decreased after typhoon passage, probably due to freshwater discharge ( Figure 2B). An increase in salinity would cause an increase in pCO2, and vice versa, by affecting the thermodynamic constants. Consequently, the calculated results showed positive ΔpCO2.S values at all stations (except station 1) within a narrow range between 3 and 7 μatm, while ΔpCO2.S was −2 μatm at station 1 ( Figure 4C). Compared with temperature, salinity had a minor effect on SS pCO2 variations in response to typhoon passage.

Zone
The enhanced vertical mixing may bring subsurface water that is replete in DIC and TA into the surface layer and therefore may cause SS pCO2 variations. The calculated results showed a generally positive ΔpCO2.C at all stations (except station 8; Figure 4D), suggesting that the input of DIC may play a more significant role than TA input in regulating SS pCO2 variations, providing that DIC increment would increase pCO2, but TA increment would decrease pCO2. At station 8, MLD merely deepened 3 m after typhoon passage so that ΔpCO2.C was very close to null.
The potentially subsequent elevated biological production stimulated by the nutrient input via enhanced vertical mixing had a nearly negligible impact on SS pCO2 variation at stations 3 to 9, as shown by ΔpCO2.B, which was very close to zero ( Figure 4E). This may reflect the fact that the nitracline is relatively deep in the study area, so the enhanced deepening of MLD could transport limited nitrate into the surface layer ( Figure 3F), which was further supported by the little change in Chl a concentration before and after the ty-

Non-Typhoon on Surface pCO 2
As shown in Figure 4G, the calculated ∆pCO 2O (i.e., another effect on SS pCO 2 variation) varied substantially within a range between −16 to 37 µatm, which was generally larger than the ∆pCO 2.typhoon . Furthermore, ∆pCO 2.O revealed a very similar spatial variation pattern as ∆pCO 2.Obs . From these lines of evidence, we suggest that ∆pCO 2.O could be the dominant process controlling the variations of SS pCO 2 after typhoon passage in the study area.
According to previous studies, the core of cyclonic eddies is characterized by lower sea level and shallower pycnocline [25] due to eddy-driven upwelling, consequently increasing SS pCO 2 [9,10,26]. Thus, in this study, we examined the potential impact of ocean eddies/currents on SS pCO 2 variations after the typhoon passage by comparing the spatial distribution of SLAs and geostrophic flow before and after the typhoon from 25 July to 8 August 2012 ( Figure 5). On July 25, before the typhoon occurred, a cyclonic eddy was detected far offshore in Zone IV and east of Zone III ( Figure 5A). Then, the cyclonic eddy propagated westward and reached Zone III on 8 August ( Figure 5B), after the typhoon. Such propagating cyclonic eddy could induce upwelling in Zone IV before the typhoon and in Zone III after the typhoon. Thus, the observed SS pCO 2 decreased in Zone IV but increased in Zone III ( Figure 4G) after the typhoon passage in this area, probably due to the exit and arrival of the cyclonic eddy in Zone IV and Zone III, respectively.
Moreover, from 25 July to 8 August 2012, the Kuroshio flowed through Zone I and varied with the cyclonic eddy approaching from the east ( Figure 5). The Kuroshio became weaker and narrower ( Figure 5C,D) when impinging with the cyclonic eddy after the typhoon passage, concurrent with that reported by [25]. In the same region after the typhoon, near-surface salinity decreased significantly above the 22-kg/m 3 isopycnic at Station 1 (Figure 6), suggesting the influence of the freshwater discharge near the surface in this area. Additionally, the water T-S properties at Station 1 before the typhoon were very close to that of the SCS (Figure 6), suggesting the influence of the outflowing SCS subsurface water. Previous studies have shown that the outflowing SCS water may induce high biological production and activities in the Kuroshio off eastern Taiwan [13,27,28], which could be favorable for the SS pCO 2 dropdown before the typhoon. In contrast, freshwater from terrestrial sources generally tends to have higher pCO 2 . Therefore, the SCS water intrusion into the area east of Taiwan before the typhoon, along with more freshwater discharge after the typhoon, may explain the increase in the observed SS pCO 2 in Zone I after typhoon ( Figure 4G).
Water 2022, 14, x FOR PEER REVIEW 12 of 15 Figure 6. Comparison of T-S diagrams before (blue symbols) and after (red symbols) the typhoon passage at the nine hydrological stations. Cyan and green lines refer to the typical T-S diagrams for the South China Sea water and the Kuroshio water, respectively, according to data from [13,29].

Conclusions
In this study, SS pCO2 showed spatially different variations from the near coast area to the far offshore area after the typhoon passage (  Figure 6. Comparison of T-S diagrams before (blue symbols) and after (red symbols) the typhoon passage at the nine hydrological stations. Cyan and green lines refer to the typical T-S diagrams for the South China Sea water and the Kuroshio water, respectively, according to data from [13,29].

Conclusions
In this study, SS pCO 2 showed spatially different variations from the near coast area to the far offshore area after the typhoon passage ( Figure 7 linity increase and DIC input from the subsurface layer. We suggest that the condition in Zone II could be representative of the impact of typhoon passage on the SS pCO2 variation without the influence of advection changes in the study area. Therefore, the exit of cyclonic eddy in Zone IV could well explain the observed SS pCO2 decrease after typhoon passage, as eddy-driven upwelling generally leads to higher SS pCO2 than that in the surrounding water. By contrast, the arrival of cyclonic eddy could be the dominant process resulting in the observed SS pCO2 increase in Zone III after typhoon passage. In addition, the intrusion of SCS water before the typhoon and terrestrial input after the typhoon could collectively explain the observed SS pCO2 increase after typhoon passage in Zone I. In summary, the present findings suggest that in addition to the enhanced vertical mixing effect, the alteration of advection pattern, e.g., eddy movement and the Kuroshio itself, could also play an important role in controlling the variation of SS pCO2 after typhoon passage, and thus demonstrate the need for more focus on advection changes to better understand the impact of the typhoon on SS pCO2, particularly in the strong current area (such as Kuroshio).   Among the four sub-areas, Zone II was the only area that did not reveal noticeable alteration in advection pattern before and after typhoon passage (i.e., arrival and exit of the cyclonic eddy, intrusion of SCS water, and freshwater discharge). Also, Zone II showed the smallest SS pCO 2 variation after typhoon passage, in which pCO 2 decrease driven by temperature dropdown and potential biological production enhancement fueled by nutrients input was nearly compensated by the increase in pCO 2 caused by salinity increase and DIC input from the subsurface layer. We suggest that the condition in Zone II could be representative of the impact of typhoon passage on the SS pCO 2 variation without the influence of advection changes in the study area. Therefore, the exit of cyclonic eddy in Zone IV could well explain the observed SS pCO 2 decrease after typhoon passage, as eddy-driven upwelling generally leads to higher SS pCO 2 than that in the surrounding water. By contrast, the arrival of cyclonic eddy could be the dominant process resulting in the observed SS pCO 2 increase in Zone III after typhoon passage. In addition, the intrusion of SCS water before the typhoon and terrestrial input after the typhoon could collectively explain the observed SS pCO 2 increase after typhoon passage in Zone I.
In summary, the present findings suggest that in addition to the enhanced vertical mixing effect, the alteration of advection pattern, e.g., eddy movement and the Kuroshio itself, could also play an important role in controlling the variation of SS pCO 2 after typhoon passage, and thus demonstrate the need for more focus on advection changes to better understand the impact of the typhoon on SS pCO 2 , particularly in the strong current area (such as Kuroshio).