Author's personal copy

values. Similarly, the N/P ratio of marine diazotrophs varies only by afactor 3, from 18 to 56 g/g (Okin et al., 2011 and the references therein).Assuming ameanN/P ratio of 37 g/g (applicable tomarine diazotrophs;Okin et al., 2011 and references therein), we have estimated the nitro-gen fixation rate based on the assumption that surface waters are de-pleted in P.

It is noteworthy that N-fixation in the ARS ranges from 0.001 to0.03 Tg-N yr−1 with an N/Fe ratio of about 50. Likewise, nitrogen fix-ation rate varies from 0.03 to 0.67 Tg-N yr−1 with an N/Fe of 1050. AsTrichodesmium species (a marine diazotrophs that require highabundance of iron) are identified both in the ARS and BoB, we havecompared the higher estimate (0.03–0.67 Tg-N yr−1) with otherestimates reported in the literature. This is of comparable mag-nitude with earlier estimates of 3.3 Tg-N yr−1 Bange et al. (2000),1.5 Tg-N yr−1 (Somasundar et al., 1990), 1.0 Tg-N yr−1 (Capone etal., 1997) and 3.5 Tg-N yr−1 (Brandes et al., 1998). Similarly, assum-ing P-limitation, the estimated N-fixation rate for Arabian Sea corre-sponds to ~0.5 Tg-N yr−1. In both these approaches, the N-fixationrate is quite similar and is comparable with the literature studies.However, the estimated N-fixation rate may vary by a factor of 3 to5 (Table 5). Recently, the N-fixation rate of ~15.6 Tg yr−1 has beenreported for theTrichodesmiumspecies in the Arabian Sea during springbloom (Gandhi et al., 2011). This estimate is somewhat higher com-pared to other studies for this region ((Bange et al., 2000 and referencestherein; Capone et al., 1997). Themodeling study by Krishnamurthy et al.(2010), suggests that an increase in the atmospheric input of N to the tro-pics could result in decrease in N-fixation by diazotrophs due to compe-tition with other diatoms and picoautotrophs. Although, our estimate of

N-fixation (Fig. 4) is relatively low compared to that reported in thebloom conditions of Trichodesmium species (Gandhi et al., 2011), it isreasonable to argue that atmospheric supply of iron from desert regionsis likely to be an important factor responsible for observed nitrogenfixation.

The N-fixation rate in the BoB, supported by atmospheric depositionof soluble Fe, is about 0.4 Tg-N yr−1 (for N: Fe = 50) and 8.2 Tg-N yr−1

(for N: Fe = 1050). Assuming P limitation in the surfacewaters of Bay ofBengal (area of 2.2 *1012 m2); average N-fixation rate is 0.9 Tg-N yr−1.Due to the lack of measurements of N-fixation rate for this region(BoB), an upper estimate of 1.0 Tg-N yr−1 has been suggested basedon a mass balance approach by (Naqvi et al., 2011). This is somewhatof the same order with our estimate in this study (0.4 and 0.9Tg-N yr−1 by assuming Fe and P limitation respectively). Their studyalso suggests that the N-fixation rate in the BoB is of comparable mag-nitude with that in ARS. Therefore, a higher estimate for N-fixationrate (~8.2 Tg-N yr−1) is very unlikely. We suggest N-fixation rate of0.5 Tg-N yr−1 for the Northern Indian Ocean (assessed based on thesoluble Fe fluxes). These calculations could be biased due to underlyingassumption that surface waters in the study region (BoB and ARS) arelimited by the availability of Fe and P. However, our approach providesa reasonable estimate for theN-fixation rate that is somewhat compara-ble with the earlier estimates.

3.5. Ecological impact

The limiting concentration of nutrients in surfacewaters can be tracedbased on the elemental ratios (N:P or Si:P andN:Si) in a particular aquatic

Fig. 4. Estimated potential carbon (Pg-C yr−1) and nitrogen fixation rates (Tg-N yr−1), supported by atmospheric deposition of Fe.

110 B. Srinivas, M.M. Sarin / Science of the Total Environment 456-457 (2013) 104–114

　 　 　

　 AS BoB

N 0.2-18.6 2-167

P 0.3-0.9 0.5-4.8

Fe 0.001-0.015 0.02-1.2

aerosol 　Natural Anthropogenic

(solubility) small large

(umol m-2 day-1)

A-OS27 インド洋域の物理・生物地球化学・生態系と相互連関

（総説）ベンガル湾の生物地球化学におけるエアロゾルの役割本多牧生、松本和彦、Eko Siswanto

金谷有剛、竹谷文一、宮川拓真(海洋研究開発機構)

AOS27-P01 JPGU2017 20-25 May 2017, 幕張メッセ

SIBER-7 Scientific Steering Committee Meeting, January 31, 2017, Perth, Australia ------------------------------------------------------------------------------------------------------------

7th Scientific Steering Committee Meeting, January 31st, 2017, Perth, Australia

Final Agenda

Planned Attendees: SIBER SSC Members: Raleigh Hood (Chair), Mike Roberts, (co-Chair), Jerry Wiggert, Dwi Susanto, Mike Landry, Greg Cowie, Makio Honda, Lynnath Beckley, M. Ravichandran, Somkiat Khokiattiwong SIBER IPO Staff: Satya Prakash IOC Perth Office: Nick D’Adamo, Louise Wicks SIBER-7 will be held in conjunction with the13th Indian Ocean Regional Panel (IORP) meeting, the 13th Indian Ocean GOOS (IOGOOS) meeting, the 7th IndOOS Resources Forum (IRF) meeting and the 1st IIOE-2 Steering Committee meeting. SIBER will participate in a joint IndOOS Review Workshop with IOP and IOGOOS on January 30 through February 1 (3-days, see IOP Workshop agenda). Also note that there will be a group dinner (for IOP, IOGOOS and SIBER delegates and guests, date and location TBA). The SIBER SSC business meeting will be convened from 8:30 am – 11:30 am on January 31st. An additional 1.5 hour time slot has also been set aside on February 1st, from 4:00 pm – 5:30 pm, to continue the SIBER business meeting focusing on IIOE-2 updates. There will be a SIBER BBQ on Thursday evening, February 2nd, hosted by Lynnath Beckley.

SIBER Sustained Indian Ocean Biogeochemistry and Ecosystem Research

Sustained Indian Ocean Biogeochemistry and Ecosystem Research

2000年の反応性N（Nr）供給量に対する2030年のNr供給量予測値の比(Duce et al. 2008 Science 320. 893-897)

　冬季〜春季にかけては、アラビア海（AS）より、 ベンガル湾（BoB）への陸起源エアロゾル供給量が多い

アラビア海とベンガル湾における冬季ー春季のN, P, Feの乾性沈着量の比較　(Srinivas and Sarin. 2013 Science of the Total

Environment 456-457. 104-114)

　ベンガル湾とアラビア海におけるN、P収支

　Fe供給時の窒素固定藻(diazotroph）による 潜在的基礎生産力（炭素同化量）およびN2固定量

(after Srinivas and Sarin. 2013 Science of the Total Environment 456-457. 104-114)

左： C/Fe比=125000とした時の炭素同化量(Pg-C yr-1)、中央：N/Fe=50とした時のN2固定量、右：N/Fe=1050とした時のN2固定量 (Tg-N yr-1)

(Srinivas and Sarin. 2013 Science of the Total Environment 456-457. 104-114)

　白鳳丸航海（2019年1月頃）の予定航跡とエアロゾル関係観測項目案

　 　 　 　

試料 測定項目 測定成分 測定手法、解析手法

海水酸素、栄養塩、炭酸系

DO, NO3, NO2, NH4, DIC, Alkalinity, pCO2, DOC, pH

溶存微量金属 Fe, Al, Pb, Ca, Znなど

炭素・窒素安定同位体

13C, 15N 質量分析計

懸濁粒子 微量金属 Fe, Al, Pb, Ca, Znなど ICP-MS/AES

炭素・窒素安定同位体

13C, 15N

沈降粒子 微量金属 Fe, Al, Pb, Ca, Znなど ICP-MS/AES

炭素・窒素安定同位体

13C, 15N

黒色炭素

植物プランクトン 種組成 珪藻、円石藻、窒素固定藻など CHEMTAX, HPLC

現存量 クロロフィル他色素 ターナー蛍光光度計

基礎生産力 現場・擬似現場方法

新生産 15NO3取込速度測定

添加実験 栄養塩、微量金属（Fe）添加実験

窒素固定 15Nトレーサー法

動物プランクトン 種組成 顕微鏡観察

エアロゾル 微量金属 Al, Ba,Ca, Cd, Co, Cr, Cu, Fe, K, Mg, Mn, Na, Pb, Znなど ICP-MS/AES

（PM2.5, PM10) イオン 陽イオン：Na+, NH4+, K+, Mg2+, Ca2+など イオンクロマトグラフィ

陰イオン：NO2-, NO3

-, Br-, PO32-, Cl-, SO4

2-, HCO3-など

黒色炭素

起源・輸送経路 後方流跡線解析

光学的観測 エアロゾルX線回折、MAX-DOAS

サンフォトメーター

微量金属・栄養塩溶出速度

雨水 イオン 陽イオン：Na+, NH4+, K+, Mg2+, Ca2+など イオンクロマトグラフィ

　 　 陰イオン：NO2-, NO3

-, Br-, PO32-, Cl-, SO4

2-, HCO3-など 　

　インド洋ダイポールモード現象（IOD）発生時の インドネシア森林火災由来エアロゾルの供給による植物プランクトン増加の可能性

Normal 1997 IOD

海面水温

基礎生産力

エアロゾル光学深度

通常時（左）と1997年インド洋ダイポールモード現象発生時（右）の東部インド洋赤道域（インドネシア西方沖）における海面水温（上）、基礎生産力（中）、エアロゾル光学深度（下）。基礎生産力の増加は亜表層からの栄養塩供給

に加え、同時期に発生した山火事起源エアロゾル中栄養塩の供給による可能性がある (Siswanto et al. 2015 GRL 42. 5378-5386)

　表層混合層海水中溶存FeおよびAl濃度測定結果から推定された エアロゾルによるFe供給量

(a)  2007年2-4月、南部インド洋ーベンガル湾観測の航跡、(b) 航跡に沿った溶存Al、Fe濃度、(c) 河川・亜表層からの供給量、滞留時間、Fe, Al溶解率、混合層深度を考慮してプログラムMADCOWを用いて計算されたダスト

沈着量、(d) 緯度別エアロゾルFe供給量の比較 (Grand et al. 2015 GBC 29. 357-374)

Principal Enquries: JPO Perth

Office (/IIOE-2/PerthOffice.jsp) (https://www.facebook.com/International-Indian-Ocean-Expedition-2-888081817972325/)

(https://twitter.com/iioe_2)

Home (/IIOE-2/index.jsp)

Sitemap (/IIOE-2/Sitemap.jsp)

Search (/IIOE-2/Search.jsp) Contact Us

(/IIOE-2/ContactUs.jsp)

Second International Indian Ocean Expedition (IIOE-2)"to advance our understanding of the Indian Ocean and its role in the Earth System in order to enable informed decisions in

support of sustainable development and the well-being of humankind"

This international program of discovery will be guided by a rich set of contemporary science questions. It will provide newobservational data that will be processed, curated, managed and used to underpin characterisation of key oceanic and relatedatmospheric processes at a wide spectrum of spatial and temporal scales. This will support improved modelling andpredictability, and as an integrated body of new information, the transfer of knowledge for societal benefit with an emphasis on

capacity development as a key vehicle to achieve this objective.   Read more (/IIOE-2/iioe-2.jsp)

Flash News

(links still under development)

!! Working Groups, Research Themes and activities !!

Education, training, mentoring

relationship building,

empowerment through science

and applications

(/IIOE-2/cb.jsp)

CapacityDevelopment

Translating

New knowledge on Indian Ocean bio-physical oceanographic and coupled

atmospheric science

(/IIOE-2/sr.jsp)

Science & Research

!! Six Research Themes !!

" # $

% &

'

(   *** Endorsement Form for IIOE-2 Science Activities *** (/IIOE-2/EndorsementForm.jsp)) * +

Theme 4: Circulation, climate variability and change

Endorsed Projects & Activities

(/IIOE-2/Endorsed_Projects.jsp)Read more

International Indian Ocean Expedition (IIOE-2) http://www.iioe-2.incois.gov.in/IIOE-2/index.jsp#

1 / 3 2017/05/12 9:31

Atmosphere

Dry deposition

0.01 – 0.38 Tg-N yr-1

6 – 50 Gg-P yr-1

Surface water

Deep water

Bay of Bengal

Rivers

0.9 – 2.0 Tg-N yr-1

35 – 61 Gg-P yr-1

Primary Production

99 - 566 mg-C m-2 d-1

f-ratio = 0.5

(a) Atmosphere

Dry deposition

0.01 – 0.19 Tg-N yr-1

8 – 21 Gg-P yr-1

Surface water

Deep water

Arabian Sea

Rivers

0.9 – 0.43 Tg-N yr-1

7 – 13 Gg-P yr-1

Primary Production

0.8 - 2.4 g-C m-2 d-1

f-ratio = 0.5

(b)

www.nature.com/scientificreports/

2Scientific RepoRts | 6:30416 | DOI: 10.1038/srep30416

shipping corridors3,4. Since there are no oceanic sources of NO2 over this region, the observed increase in tropo-spheric NO2 values due to seagoing ships over BoB and associated warming is significantly high.

In addition to gaseous emissions (carbon dioxide, nitrogen oxides, carbon monoxide, sulphur dioxide, etc), ships exhaust also consist of unburned hydrocarbons and particulate matter such as black carbon (BC), sulphate and organic carbon. Marine vessels are contributing 7–9% of global diesel BC emissions in 2000 and is expected to rise due to increased shipping demand5,6. BC emissions further enhance the warming, as the fossil-fuel domi-nated black-carbon plumes are known to be more efficient warming agents7,8. The BoB, despite being a potentially very energetic region for the development of monsoon depressions, have experienced a noticeable decrease in their frequency during the recent years9,10. Also, the summer monsoon precipitation averaged over the Indian region10 declined by 7% during 1951–2005. Recent studies have attributed the observed declining trend of sum-mer monsoon to anthropogenic aerosol forcing11–14. There are only few measurement studies over the BoB on the regional impact of ship emissions. The Continental Tropical Convergence Zone (CTCZ) experiment conducted during summer 2012 provided us with observations over southern BoB for determining the impact of ship emis-sions on black carbon warming and cloud condensation nuclei (CCN) concentration. This campaign was in part motivated by the Bay of Bengal Monsoon Experiment15 in 1999 and the Arabian Sea Monsoon Experiment16 in 2003 to broaden our understanding of monsoon processes.

ResultsOn-board the research vessel Sagar Nidhi, the CTCZ campaign had deployed ground-based measurements dur-ing July 21 to August 20, 2012, over the southern BoB directly down-wind of shipping lane as shown in Fig. 1b (see Methods section). CTCZ intercepted aerosol plumes from coastal, shipping, and relatively-pristine marine sources, which were classified into three categories based on NO2 concentrations (Fig. 1) and the regional flow: (1) Coastal plumes, (2) Shipping plumes and (3) relatively-pristine Marine plumes, which consisted of marine sources and plumes from other regions.

BC concentrations and associated solar heating rates. The temporal variation of BC (at 880 nm), aer-osol concentration (CN) and cloud condensation nuclei (CCN) concentrations along the cruise track are shown in Fig. 2. The BC mass concentrations were typically in the range of 30 to 2500 ng.m−3, with higher BCs (> 200 ng.m−3) for anthropogenic plumes. Aerosol concentrations decreased rapidly (BC: 200 ng.m−3, CN: 10,420 cm−3) as the vessel moved 100 km away from the coast (BC: 2500 ng.m−3, CN: 135,490 cm) which suggested that the heavy anthropogenic emission-laden air masses encountered at the coast were transported over to the BoB. The BC (and CN) concentrations further decreased to very low values at 8oN (BC: 63 ng.m−3, CN: 775 cm−3), comparable to those reported for background marine concentrations of 140 ng.m−3 measured in the Southern Ocean17. We intercepted the shipping corridor at 89oE and 85oE longitudes (on Aug 12–13 and Aug 16 respectively). Owing to this, BC mass concentrations went up to 523 ng.m−3 with a mean value of 282 ± 184 ng.m−3; showing a four-fold

Figure 1. NO2 signature of shipping emissions in the Bay of Bengal. Mean tropospheric NO2 retrieved by OMI (ozone monitoring instrument) satellite for (a) Jan-Feb, 2012 and (b) July-Aug, 2012. Prevailing mean atmospheric circulation at 925 mb level during the respective time periods based on NCEP/NCAR reanalysis streamlines data provided by the NOAA-CIRES Climate Diagnostic Center are shown in white color. In (b) the CTCZ cruise track over southern Bay of Bengal is shown in red color. S1 and S2 shown in (b) are the locations where ship was stationary for continuous measurements at 8oN 85oE and 8oN 89oE respectively. Figure is generated using MATLAB R2015a software available at http://in.mathworks.com/products/matlab/. (License no: 927142).

925mbにおける風向。背景はNO2濃度推定値。（Ramana and Devi 2016. Sci Rep 6)

　ベンガル湾の人為起源Nの増加量は世界最大級であると予想される

　陸から海洋へ供給される大気塵（エアロゾル）が海洋の様々な生物地球化学に影響を与えていることが報告されてきた。現代のみならず氷期においても、陸域起源のエアロゾルが海洋の栄養塩濃度を増加させ、基礎生産力、沈降粒子量の増加を誘引する可能性が指摘されている。一方、陸域エアロゾルの海洋供給は有害物質を海洋へ供給したり、海洋の酸性化を促進させたり等負の効果も考えられる。インド洋北東部に位置するベンガル湾は北側および東西を陸に囲まれた半閉鎖的な熱帯海域である。強い日射に加え、ガンジス川、ブラマプトラ川という世界最大級河川による淡水流入、さらに北半球夏季モンスーン（南西風）による多量の降雨により成層構造が発達するため亜表層からの栄養塩供給が極めて少ない貧栄養海域である。このベンガル湾において海洋表層の基礎生産力を高めるための栄養塩供給メカニズムとしては、海洋内部の中規模渦、サイクロンのような気象擾乱に加え、陸域エアロゾル供給が挙げられる。インド洋特有のモンスーンシステムにより、冬季は北東風が卓越し陸から海に向かって風が吹くため、陸域から天然起源あるいはバイオマスバーニングや化石燃料使用による人為起源のエアロゾルが大量にベンガル湾に供給される。ベンガル湾の特徴として、湾周辺国には世界人口の約1/4が集中している。特にインドやバングラディッシュは大気汚染度（PM2.5濃度）が高いことで知られており、大量の人為起源のエアロゾルがベンガル湾周辺で発生し大気中へ放出されている。アラビア海に比べると、冬季—春季はより多くの天然起源・あるいは人為起源のエアロゾルが供給されやすい海域であるとの報告がある。人為起源エアロゾルに含まれるmicronutrientである鉄は、天然起源エアロゾルに比較すると、濃度が高く、かつ溶出しやすい（換言すれば、生物に利用されやすい）との報告がある。したがって将来的にはより多くの鉄が海洋に供給された結果、鉄不足の海域では基礎生産力が増加する可能性がある。またベンガル湾南東部に位置するインドネシアからも山火事、火山噴火、バイオマスバーニングによる大量のエアロゾルが発生し同海域の海洋の生物地球化学に影響を与えている可能性も指摘されている。大量のエアロゾルがベンガル湾海上大気に輸送された場合は日射を遮断し、海洋の一次生産力を低下させる効果もあるはずである。従って生物地球化学におけるエアロゾルの役割を研究する上でベンガル湾は“陸—海洋相互作用、大気—海洋相互作用、人間活動—海洋相互作用の研究におけるホットスポット”と言える。 　本発表ではベンガル湾の生物地球化学におけるエアロゾルの役割に関する研究例をまとめ、2018年度に計画されている白鳳丸航海における観測例について紹介する。

increased deposition further into open ocean re-gions (21, 22). The ratio of 2030-to-2000 dep-osition rates (Fig. 1C) shows up to a factor of 2increase in Southeast Asia, the Bay of Bengal, andthe Arabian Sea; up to a 50% increase off westernAfrica; and up to 30% across essentially all themid-latitude North Atlantic and North Pacific. AsGalloway et al. (9) conclude, controlling NOx

emissions using maximum feasible reductionscould substantially decrease future emissions, so

the increases we predict on deposition rates (Fig.1C) may represent upper limits.

Impact on New Primary Productionand the Biological PumpPresent global open ocean primary production isestimated at ~50 Pg C year−1 (23), equivalent to~8800 Tg N year−1, assuming Redfield stoichi-ometry (Table 2). Because ~78% of this produc-tion is driven by regeneration of Nr within surface

waters (24) (a in Fig. 2), it is more relevant toevaluate the impact of AAN deposition onoceanic productivity and biogeochemistry bycomparing AAN with global new production,estimated at ~11 Pg C year−1 (24–26). Newproduction (b in Fig. 2 and Table 2) is dominatedby nitrate regenerated at depth from sinkingorganic matter and subsequently returned to theeuphotic zone via physical transport (b′ in Fig. 2)(27). Over sufficiently large space and time scales

A D

E

F

B

C

Nr 1860

0 - 1415 - 4243 - 7071 - 140141 - 210211 - 280281 - 420421 - 560561 - 700701 - 840841 - 1,1201,121 - 1,4001,401 - 2,1002,101 - 2,8002,801 - 3,500

Nr 2000

0 - 1415 - 4243 - 7071 - 140141 - 210211 - 280281 - 420421 - 560561 - 700701 - 840841 - 1,1201,121 - 1,4001,401 - 2,1002,101 - 2,8002,801 - 3,500

Nr 2000

0 - 1415 - 4243 - 7071 - 140141 - 210211 - 280281 - 420421 - 560561 - 700701 - 840841 - 1,1201,121 - 1,4001,401 - 2,1002,101 - 2,8002,801 - 3,500

NO3- (!M)

0 - 12 - 45 - 78 - 1011 - 1314 - 1617 - 1920 - 2122 - 2324 - 2526 - 2728 - 33

Nr 2030:2000

11.01 - 1.051.06 - 1.11.11 - 1.21.21 - 1.31.31 - 1.51.51 - 22.01 - 4

Nr 2000:upwell N

0 - 0.050.06 - 0.10.11 - 0.30.31 - 0.50.51 - 11.01 - 5

(mg N m-2 year-1)

(mg N m-2 year-1) (mg N m-2 year-1)(<4 !M NO3

-)

Fig. 1. (A) Total atmospheric reactive nitrogen (Nr) deposition in 1860 inmg m−2 year−1 [NHx and NOy are derived from (3), with the addition of 30%of the total nitrogen as organic nitrogen]. Total atmospheric Nr deposition in1860 was ~20 Tg N year−1, AAN was ~5.7 Tg N year−1. (B) Total atmo-spheric reactive nitrogen (Nr) deposition in 2000 in mg m−2 year−1 [derivedfrom (21) with the addition of 30% of the total nitrogen as organic nitrogen].Total atmospheric Nr deposition in 2000 was ~67 Tg N year−1, AAN was~54 Tg N year−1. (C) Ratio of the projected flux of Nr to the ocean in 2030 to

that in 2000. (D) Nitrate concentrations (mM) in the surface (0 to 1 m) watersof the ocean (43). (E) Similar to (B), but with regions where surface nitrate>4 mM has been masked out. Total atmospheric Nr deposition in 2000 to thenonmasked areas was ~51 Tg N year−1, AAN was ~41 Tg N year−1. (F) Ratio oftotal Nr deposition to dissolved inorganic nitrogen (DIN) supply into the upper130 m as diagnosed from a model fitted to oceanic tracer observations (44).To reduce noise, computation of the ratio has been limited to areas with DINsupply exceeding 0.05 mol m−2 year−1.

16 MAY 2008 VOL 320 SCIENCE www.sciencemag.org894

REVIEWS

2. Materials and Methods2.1. Sample Collection and Analysis

Samples were collected aboard the R/V Roger Revelle during the U.S. CLIVAR CO2 Repeat HydrographyI08S and I09N cruises. The I08S transect was occupied from 15 February to 13 March 2007, starting at theAntarctic continental shelf break and continuing northward until 28°S. Sampling along the I09N linebegan on 27 March 2007 at 28°S and was completed on 27 April in the Bay of Bengal (BoB) (Figure 1).A total of 85, 12-depth vertical profiles spaced at approximately 1° intervals were sampled along thiscruise track, which extended from 65.8°S to 18°N, along 82–95°E. There are gaps in coverage near 55°S and50°S in the Southern Ocean due to rough seas that prevented the deployment of the trace metal rosettefrom the stern of the ship. Since this paper is concerned with total dust deposition and its impact onupper ocean biogeochemistry, only data within the mean climatological mixed layer of each station areconsidered here (up to 185m). The full-depth data set is described in detail in a companion manuscript[Grand et al., 2015].

Figure 1. Near-surface circulation along the CLIVAR I08S and I09N cruise tracks. The thin blue arrows represent the uppermean 50m LADCP velocities grouped into 1° bins. SZ: Southern Zone, SACCF: South ACC Front, AZ: Antarctic Zone, PF:Polar Front, PFZ: Polar Frontal Zone, SAF: Subantarctic Front, SAZ: Subantarctic Zone, STF: Subtropical Front, ACC: AntarcticCircumpolar Current, SIC: South Indian Current, SICC: South Indian Countercurrent, SEC: South Equatorial Current, ITF:Indonesian Throughflow, SECC: South Equatorial Counter Current, SMC: Southwest Monsoon Current, and BoB: Bay of Bengal.The red diamonds show the position of the stations sampled for dFe and dAl during the GEOTRACES-JAPAN expedition[Vu and Sohrin, 2013; Nishioka et al., 2013].

Global Biogeochemical Cycles 10.1002/2014GB004898

GRAND ET AL. ©2015. American Geophysical Union. All Rights Reserved. 359

3.1. Southern Ocean (65.8°S!38°S)

The circulation of the Southern Oceandomain is dominated by the eastwardflowing Antarctic Circumpolar Current(ACC). The ACC transport is concentratedalong three fronts that separate zoneswith relatively uniform water massproperties. Using the property criteria ofOrsi et al. [1995], we identified the SouthACC Front (SACCF) at 63°S, the PolarFront (PF) from 57 to 54°S, and theSubantarctic Front (SAF) at 49°S. The ACCcirculation is bounded to the north bythe Subtropical Front (STF), near 40–38°S,and to the south by the SouthernBoundary, which appears to merge withthe SACCF along this section. The ACCfronts separate four zones [Talley et al.,2011]: the Southern Zone (SZ) lyingpoleward of the SACCF, the AntarcticZone (AZ) located between the SACCFand the PF, the Polar Frontal Zone (PFZ)between the PF and the SAF, and the

Subantarctic Zone (SAZ), which extends from the SAF to the STF (Figure 1). The SAZ near the cruise trackexperiences the deepest wintermixed layers of the South Indian Ocean, with depths generally exceeding 500min winter [Rintoul and Bullister, 1999; Wong, 2005].

Surface dAl levels in the eastern Indian sector of the Southern Ocean were extremely low (0.54 ± 0.25 nM,n= 25), and seven samples in this region had dAl levels less than or equal to the analytical detection limit(0.3 nM). The concentration of dAl increased northward (Figure 2), with significantly lower values south of thePF (66–54°S) relative to waters farther north (54–38°S) in the SAZ (two tailed t test, p< 0.05). These surface dAlvalues, which reflect minimal dust inputs in the circumpolar domain, are among the lowest reported in theworld ocean and are consistent with surface observations in other sectors of the Southern Ocean thattypically lie below 1 nM [van Beusekom et al., 1997; Measures and Vink, 2000; Obata et al., 2004; Middag et al.,2011, 2012, 2013].

Surface dFe exhibited an oppositemeridional trend with values decreasingnorthward from 0.46 ± 0.22 nM (n= 5)in surface waters of the SZ to0.16±0.03nM (n=7) in the SAZ (Figure 2).We note that a similar meridional trendin dFe was also observed in a recentcompilation of historical dFe data fromthe upper 100m of the Indian sector ofthe Southern Ocean [Tagliabue et al.,2012]. The elevated surface dFe valuesfrom the Antarctic margin to the PF(54°S) may reflect localized inputs frommelting sea ice and icebergs [Croot et al.,2004a; Lannuzel et al., 2007, 2010; Linet al., 2011], as illustrated by the drop insurface salinity near 65°S (Figure 3). Theseelevated dFe levels could also reflect alateral supply from shelf sediments

32

32.5

33

33.5

34

34.5

35

35.5

36

Surfa

ce S

alinit

y

−70−65−60−55−50−45−40−35−30−25−20−15−10 −5 0 5 10 15 200

10

20

30

40

50

60

70

80

90

100

110

Mixe

d La

yer D

epth

[m]

Latitude

Surface SalinityMixed Layer Depth

Figure 3. Surface salinity and observed mixed layer depth (MLDOBS)along the cruise track.

20

25

30

35

dAldFe

0

1.5

3

4.5

6

7.5

9

10.5

12

13.5

15

Surfa

ce d

Al [n

M]

ER−12 ER−11

ER−10

ER−9

ER−2

SICC

SEC

SECC

−70−65−60−55−50−45−40−35−30−25−20−15−10 −5 0 5 10 15 200

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Surfa

ce d

Fe [n

M]

Latitude

Figure 2. Surface dAl and dFe (<25m) measured along the cruise track.Note the change in the dAl scale (top). The broken lines delineate thelatitudinal bands of the major zonal currents in the gyre and SECC region.SICC: South Indian Counter Current (northeastward flow), SEC: SouthEquatorial Current (westward flow), and SECC: South Equatorial CounterCurrent (eastward flow). The circled crosses show the mean surface dAlvalues (<25m) from Vu and Sohrin [2013]: ER-2: 8.5°N, 86°E; ER-9: 5°S 68°E;ER-10: 20°S 72°E, ER-11: 30°S 64°E; and ER-12: 38°S 57°E.

Global Biogeochemical Cycles 10.1002/2014GB004898

GRAND ET AL. ©2015. American Geophysical Union. All Rights Reserved. 361

of the PF at 54°S (Mann-Whitney Utest, p< 0.01; Figure 5). Higherdeposition in the northernmostreaches of the Southern Ocean (>54°S:100 ± 60mgm!2 yr!1) probably reflectslong-distance transport of dustoriginating from South Africa [Gaudichetet al., 1989; Tanaka and Chiba, 2006;Mahowald, 2007], combined with apossible local Kerguelen Island source.This trend is also seen in globaldust models and extrapolations ofland-based measurements [Duce et al.,1991; Mahowald et al., 2005]. AlthoughPatagonian dust may also contribute tothe northward increase in depositionobserved in the Southern Ocean[Tanaka and Chiba, 2006; Mahowald,2007; Dammshäuser et al., 2011], recentobservations along the meridionalCLIVAR A16S section at ~30°W, wherethe bulk of Patagonian dust is shown tobe deposited in atmospheric models,do not show any evidence of enhanced

surface dAl signals south of 20°S that could be attributed to the deposition of Patagonian dust [MarikoHatta, personal communication].

Our mean Southern Ocean total dust deposition flux is ~3.6 times lower than that measured over a 2 yearsampling period using open collectors collecting both dry and wet depositions at Kerguelen Island (Table 2).It is possible that the MADCOW estimations for the Southern Ocean underestimate the true flux, because the3.6% solubility applied in MADCOW may be high for the Southern Ocean. In this regard, we note that threesamples from the South Pacific (67.5–38°S) suggest that the fractional solubility of Al in this area was on theorder of 1% [Buck et al., 2013]. It is also possible that our total deposition flux is underestimated because ofdilution of the mixed layer dAl signal during winter deep mixing leading to a residence time of dAl shorterthan 5 years along our cruise track in the Southern Ocean [Measures and Vink, 2000; Middag et al., 2013].However, it cannot be ruled out that orographic rainfall at Kerguelen yields enhanced precipitation ratesrelative to our open-ocean study area. Considering that wet deposition appears to dominate total deposition(wet + dry) at Kerguelen [Heimburger et al., 2012], the Kerguelen Island data may not be strictly comparable to

Table 2. Mean MADCOW Total Dust Deposition Fluxes and Comparisons to Atmospheric Model, Land- and Island-BasedExtrapolated Measurements and Shipboard Aerosol Collection Estimatesa

Southern Ocean(66°S–38°S)

Gyre/SECC(38°S–5°S)

Equator(5°S–5°N)

Bay of Bengal(5°N–10°N)

This study: MADCOW 66 ± 60 480 ± 280 490 ± 80 2500 ± 570f

Atmospheric modelsb 67 ± 27 280 ± 170 700 ± 230 3000 ± 1000f

Land extrapolationsc 10–100 10–1000 100–10,000 >10,000Land-based measurementsd 240Shipboard aerosol collection 14 ± 5e 300–6000g

aData aremean ± 1σ or reported ranges in mgm!2 yr!1. The 1σ represents the geographical variations in the flux anddoes not include uncertainties, which are on the order of a factor of 3.5.

bMahowald et al. [2005].cDuce et al. [1991].dHeimburger et al. [2012].eWagener et al. [2008].fFluxes exclude data north of 10°N (see text for details).gSrinivas and Sarin [2013].

111021103110411051106110

This Study: MADCOW

MADCOW N. BoB

Composite Aerosol Models

020406080

100

−70 −65 −60 −55 −50 −45 −40 −35 −30 −25 −20 −15 −10 −5 0 5 10 15 20

Latitude

110

210

310

410

510

610

710

810

910

1010

Dust

Dep

ositio

n [m

g m

−2 y

r−1]

Southern Oce

an

South Indian G

yre

SEC & SECC Equator. Domain

&

BoB

Figure 5. MADCOW total dust deposition overlaid with composite atmo-spheric model total deposition data along the cruise track. Model dataare the mean of three reanalysis-based atmospheric models [Jickells et al.,2005; Mahowald et al., 2005]. The grey shading highlights the MADCOWestimations in the northern Bay of Bengal that are not at steady state dueto riverine inputs from the Ganges-Brahmaputra river system and areexcluded from the discussion (see text for details).

Global Biogeochemical Cycles 10.1002/2014GB004898

GRAND ET AL. ©2015. American Geophysical Union. All Rights Reserved. 366

(a) (c)

(b) (d)

SIBER URL:　http://www.incois.gov.in/Incois/siber/siber.jsp

JAN-FEB 2012