Contents lists available at ScienceDirect

Ocean Engineering

journal homepage: www.elsevier.com/locate/oceaneng

Analysis of drift characteristic in conductivity and temperature sensors usedin Moored buoy systemRamasamy Venkatesan∗, Krishnamoorthy Ramesh, Manickavasagam Arul Muthiah,Karuppiah Thirumurugan, Malayath Aravindakshan AtmanandNational Institute of Ocean Technology, Pallikaranai, Chennai, India

A R T I C L E I N F O

Keywords:CTD sensorDrift in sensorBio-foulingMoored buoy system

A B S T R A C T

Conductivity and Temperature (CT) sensors are attached in the Ocean Moored buoy network in Northern IndianOcean (OMNI) at different depths upto 500m below the sea surface. Drift in these sensors were analyzed basedon the result of calibration. For detailed investigation, the sensor's drift is categorized based on the position ofsensors in the mooring as surface layer (above 50m), thermocline layer (50–200m) and deep layer (200–500m)and analysis was carried out at both Arabian Sea (AS) and Bay of Bengal BoB. The drift analysis revealed thatover the time of deployment the drift in temperature sensor was very minimal and well within the accuracy limit.However, the drift in the conductivity sensors was more significant. Drift in the conductivity sensor decreases asthe deployment depth of the sensor increases and also drift in the sensors which were deployed in AS is high ascompared to the sensors deployed in BoB. Biofouling of the sensor is measured spatially in AS and BoB and itindicates that biofouling and Abrasive scouring of cells are due to the flow of high concentrations of plankton inthe euphotic zone(top layer), which may cause the drift in the conductivity sensor.

1. Introduction

Temperature and salinity are the most important physical para-meters in oceanography to determine the accurate dynamic heightvariability (Maes, 1998) and to understand the dynamic nature of theglobal climate and marine ecosystem (Feistel et al., 2015), whereindifferences in these physical parameters drive the ocean circulation,called the thermohaline circulation (Toggweiler and Key, 2003:Tsuchiya, 1968). These physical parameters differ in BoB and AS. BoB islow saline and forms strong vertical density stratification due to thelarge inputs of freshwater through precipitation and river discharge(Shenoi et al., 2002; Venkatesan et al., 2013 b). However, AS is highlysalty as a result of the influences of the three major water masses in thenorthern side, namely, the Arabian Sea High Salinity Water (ASHSW),which is generally present in the upper layer of 100m, Persian GulfWater mass (PGW) between 200 and 250m and Red Sea Water (RSW)between 600 and 900m (Prasanna Kumar and Prasad, 1999). For betterunderstanding of the physical characteristic of these Oceans, tempera-ture is measured in ITS-90 °C (°C). ITS-90 is an instrument calibrationstandard agreed to by a number of Scientists in 1990, which providesfor comparison and compatibility of temperature measurements inter-nationally. For the ocean, temperatures typically range from −2 to

35 °C (28.4–95.0 °F).Salinity can be derived from the measurement of conductivity.

Conductivity is a measure of a material's ability to conduct electricalcurrent. Water with a large amount of dissolved salts has high con-ductivity, while fresh water has low conductivity. Temperature alsoaffects conductivity: warm water has high conductivity, while coldwater has low conductivity. The units for conductivity are Siemens/meter (S/m). For water ranging from freshwater to seawater, con-ductivity typically ranges from 0 to 7.5 S/m.

Under the Ocean Observation Network of India (I-OON), the OceanMoored Buoy Network for the Northern Indian Ocean (OMNI) withseven buoys in the Bay of Bengal and five in the Arabian Sea (Fig. 1) areattached with sensors to collect meteorological, surface and subsurfaceoceanographic parameters on a real-time basis through INMARSATsatellite (Venkatesan et al., 2013 a) for the past 7 years. This network ofOMNI buoy systems has been providing data which are of great re-levance to the climate research community to constantly monitor theseasonal, intra-seasonal, annual and inter-annual variations in thenorthern Indian Ocean.

These buoy systems have a time-series of vertical profiles of tem-perature and conductivity (SBE37 IM CT sensor Make Seabird USA) upto 500m water column from the surface at ten discrete depths in the

https://doi.org/10.1016/j.oceaneng.2018.10.033Received 4 March 2018; Received in revised form 20 October 2018; Accepted 23 October 2018

∗ Corresponding author.E-mail address: venkat@niot.res.in (R. Venkatesan).

Ocean Engineering 171 (2019) 151–156

0029-8018/ © 2018 Published by Elsevier Ltd.

T

Bay (Fig. 2). Accuracy and range of the sensor used in OMNI Buoys aregiven in the Table 1.

The OMNI buoy programme addresses a long-standing need to un-derstand the observed variability of upper ocean thermohaline andcurrent structures on several timescales that has an important bearingon the evolution of seasonal monsoons and cyclones (Venkatesan et al.,2013 b). However, there is no study results regarding the time drift of

the temperature and conductivity sensors. Ando et al., 2004 studied thedrift in CT sensor using Triangle Trans Ocean Buoy Network (TRITON)and revealed that drift in temperature sensor were 0.1mK and con-ductivity drifted higher in the sensor moored in the upper ocean thanthe sensor moored in the deep ocean. Freitag et al., 2005 studied thedrift in temperature sensor in TAO/TRITON moored buoy Array in thetropical Pacific and the PIRATA Array in the tropical Atlantic, whichrevealed that drift in temperature sensor was−0.0095 °C.

This article describes the use of a different kind of technology tomeasure conductivity and temperature, conductivity and temperaturecalibration system, time drift of sensor using post deployment calibra-tion, bio fouling in moored instrument, and data correction.

2. Technology to measure conductivity

In the last century, two methods were employed to measure salinity,one is by drying a sample and weighing the residue (Forch et al., 1902)and the other method is by carrying out a complete chemical analysis ofthe sample's composition and adding up the constituent masses. Thesetwo methods show results with different values for the same seawater(millero et al., 2008). Efforts are being taken for obtaining consistencyin the measurement of salinity. However, the link has not been estab-lished between any of the salinity definitions and the internationalsystem of units (SI) (Maes, 1998). At present, conductivity definessalinity using suitable correction equations. As a result of recent ad-vancements, many techniques were used to measure conductivity(Nelson et al., 2003; Diaz-Herrera et al., 2006; Tengesdal, 2014).However, due to practical implications in the field, inductive andconductivity cells are mostly used by many of the ocean observations.

Conductive type sensors used in the OMNI buoys are SBE37, man-ufactured by the company viz., Sea Bird Electronics, USA (Sea Bird). Itis a conductivity cell type sensor. The conductivity measured by thesensor using a scale factor or cell constant reflects the ratio of lengthand cross-sectional area of the sampled water volume in which theelectrical current actually flows.

The conductivity derives from the relationship:

R= ρ L/A

where:R= resistance=1/conductance ρ= resistivity=1/conductivityL= length of sampled water volume A=cross-sectional area of sampledwater volume.

The conductivity cell is made up of borosilicate glass with threeinternal platinum electrodes; it has zero external fields by connecting itsouter electrodes together and no voltage difference exists to create anexternal electrical current (Fig. 3). Thus, it is immune to proximityerrors and readily protected from fouling by anti-biology (toxic) gate-keepers placed at the ends of the cell.

Stable cell geometry is important for an accurate conductivitymeasurement. Cell contaminated with oil, bio fouling or other foreignmaterials will reduce cell geometry. Through formula 1, contaminationof the cell increases the apparent resistance because of the reduction ofcell diameter, hence, the conductivity measurement becomes lowerthan the actual value. Abrasive scouring of cell due to flow of highconcentrations of plankton into the cell may increase the cell diameterand conductivity measurement becomes higher than the actual value.Generally, it is called positive drift of sensor (Freitag et al., 1999). The

Fig. 1. Location of OMNI buoy systems.

Fig. 2. Mooring configuration of the OMNI buoy system.

Table 1Detail of temperature and conductivity sensor.

Temperature (°C) Conductivity (mS/cm)

Measurement Range −5 to +35 0 to 70Initial Accuracy 0.002 0.003Typical Stability per month 0.0002 0.003

Fig. 3. Conductive type cell –cross sectional view.

R. Venkatesan et al. Ocean Engineering 171 (2019) 151–156

152

characteristics of sensor, therefore, have a natural tendency to driftduring the contamination. Sea-Bird moored CTDs use EPA approvedanti-fouling devices to keep the insides of the sensors clean, so thatfouling will not affect the measurements. The Anti-Fouling Device is anexpendable device that is installed on each end of the conductivity cell,so that any water that enters the cell is treated (Seabird, 2016b). Anti-Fouling Devices have no effect on the calibration because they do notaffect the geometry of the conductivity cell in any way (Marc le men,2012; Seabird, 2009). In OMNI buoy system, the SBE 37 CT sensor withpump (Fig. 4) is fixed at particular depth, the pump flushes the con-ductivity cell at a faster rate than the changes in the environment, sothe T and C measurements stay closely synchronized with the en-vironment (i.e., even slow or varying response times are not significantfactors in the salinity calculation).

3. Technology to measure temperature

The Platinum Resistance Thermometer (PRT) provides accuratetemperature measurement and is used as the laboratory standard.Generally, platinum-resistance thermometer is used as interpolationdevice, for the temperatures commonly found in the ocean. It has aloosely wound, strain-free, pure platinum wire whose resistance is afunction of temperature.

Thermistors are mostly used because of the lower cost involvedcompared to other temperature sensors. The sensing element of the SBE37 sensors is glass-coated thermistor bead, pressure-protected in a thin-walled stainless steel tube. Thermistors differ from ResistanceTemperature Detectors (RTDs) based on the material used. A thermistoris generally made up of ceramic or polymer, whereas RTDs use puremetals.So, a thermistor can be interchanged without applying any ca-libration on it thereby saving the cost of calibration.

4. Calibration system

Calibration of sensors is performed in computer-controlled tem-perature baths for temperature and conductivity by M/s SeabirdElectronics USA. Calibrations in pre and post deployment are importantfor the data correction. SBE3 temperature sensor is used as secondary

reference to calibrate the temperature sensor used in the mooring (SBE37). SBE3 sensor is calibrated by using Standard Platinum ResistanceThermometer (SPRT), which is calibrated using physical standards astriple-point-of-water cells and gallium melt cells that are certified byNIST USA. Temperature calibrations are done at 7 points over theoceanographic range of 1–32.5 deg C.

IAPSO Standard Seawater is used as a primary standard for con-ductivity calibration (Seitz et al., 2011). The practical salinity of IAPSOis obtained by a series of conductance measurements at OSIL relative toKCL standard solutions, prepared using precise weights of KCL crystals(Bacon et al., 2007). Salinometer with IAPSO water is used to calibratethe SBE4 conductivity sensor, which is the secondary standard to cali-brate the conductivity sensor used in mooring (Ando et al., 2004;Venkatesan et al., 2012; Bihana et al., 2014). Conductivity calibrationsare done at 8 points over the range of 0–6 S/m. Post-deployment cali-bration is done without cleaning conductivity cell. Based on the resultof post-deployment, the pre deployment calibration is performed eitherby re-platinizing the cell or after cleaning the internal part of the cell.

5. Drift of the sensor estimated using laboratory calibration

Sensors retrieved from the OMNI buoy systems are sent back to M/sSeabird Electronics USA, for pre and post deployment calibration. Fromthese calibrations the drift of the sensor is calculated as described in thedocument appnote31 released by M/s Seabird Electronics USA,(Seabird Electronics, 2016a). Totally 80 sensors were calibrated and itsdeployed mean time at sea was 580 days and the distribution of thesensor at sea reflects that around 50% of the sensors were calibratedafter 450–600 days of operation at sea (Fig. 5). The regular serviceperiod of the OMNI buoys at sea is scheduled for one year. However, inafew cases the early replacement happened due to vandalism or driftingof buoy system. The longer service periods were due to the non-avail-ability of ship time for the regular maintenance or due to redeploymentof same sensors at sea.

The drift in temperature sensors is very minimum and it is wellwithin the accuracy limit of the sensor even if it was redeployed con-tinuously for three years of the moored period. It also indicates that thedrift does not depend on the time in function and depth at which thesensor is moored (Fig. 6). The post deployment calibration indicatesthat about 93% of the sensors drifted in negative side (Fig. 7) and theaverage drift was −0.00012 °C and −0.00027 °C per year and standarddeviation were 0.00012 °C and 0.00015 °C in AS and BoB, respectively.The post deployment calibration clearly indicates that the temperaturesensor is very stable over time and drift is not dependent on depth(Fig. 8) and time.

However, the conductivity sensor used in mooring drifted moresignificantly (Fig. 9). The post deployment calibration indicates thatabout 87% of the sensors are drifted in the positive side and 13% driftedin the negative side because of the abrasive scouring and contamination

Fig. 4. Left panel: Pumped flow through the sensor. Right panel: Anti-Foulantdevice.

Fig. 5. Distribution – Day in the measurement at sea.

R. Venkatesan et al. Ocean Engineering 171 (2019) 151–156

153

of the cell, respectively. Calibration result of conductivity sensor in-dicates that the sensors moored in 200m and 500m depth are wellwithin the sensor accuracy limit, whereas few sensors which weremoored in the euphotic zone (above 200m depth) drifted more than theaccuracy limit.

For detailed investigations, the drift in the sensor is categorizedaccording to the places where the sensors are moored as surface layerabove 50m which includes 10, 15, 20 and 30m moored sensors,thermocline layer from 50 to 200m depth, which includes 50, 75 and100m moored sensors and Deep layer from 200 to 500m depth, whichinclude 200 and 500m moored sensors. Further, it is categorized spa-tially as BoB and AS.

The drift of the conductivity is higher in surface layer and decreasesin thermocline layer and it further decreases in deep layer (Fig. 10).

Drift in the conductivity sensors that were deployed in AS is high whencompared to the sensors deployed in BoB. The Average drift of con-ductivity in the surface layer was very high ie. 0.00335 and 0.00275PSU/month for AS and BoB, respectively (Table 2). In the thermoclinelayer, the drift is 0.00318 and 0.00196 for AS and BoB, which iscomparatively lesser than the surface layer.

Thus, the conductivity measurement over the period becomeshigher than it should be measured. The standard deviation also in-creased in the surface and thermocline layers, indicating a large var-iance of drift for each sensor. On the other hand, the drift of the con-ductivity in the deep layer (below the euphotic zone) was very small0.00014 and 0.00013 for AS and BoB, respectively which are not sig-nificant when compared to the accuracy limit of the sensor. This ana-lysis revealed that the conductivity sensors are very stable. The drift inconductivity sensor in the surface and thermocline layer was not causedby the sensor itself but by its environment. Bio fouling on the con-ductivity sensor is a limiting factor to provide high quality data in theeuphotic zone(surface and thermocline layer). To overcome these

Fig. 6. Drift in the temperature sensors used on the OMNI buoy system. Boldgray line indicates the accuracy of the sensor; horizontal axis indicates thesensor's moored period in days.

Fig. 7. Distribution – Drift in Temperature sensor.

Fig. 8. Drift in Temperature sensor with respect to the moored depth. SouthernAS (SAS) – Average of AD09 and AD10. Northern AS – Average of AD06, AD07and AD08. Northern BoB – Average of BD 09, BD08, BD10 and BD13. SouthernBoB – Average of BD14, BD11 and BD12.

Fig. 9. Drift in the Conductivity sensors used in the OMNI buoy system. Boltedgray line indicates the accuracy of the sensor and horizontal axis indicates thesensor's moored period in days.

Fig. 10. Drift in conductivity sensor with respect to the moored depth.

Table 2Drift of conductivity sensor in PSU/month.

Location Arabian Sea BOB

Surface Layer(above 30m) Average 0.00335 0.00275Std Dev 0.01183 0.00139Data Points 12 21

Subsurface Layer(50–100m) Average 0.00318 0.00196Std Dev 0.00154 0.00121Data Points 14 13

Deep Layer(200–500m) Average 0.00014 0.00013Std Dev 0.00121 0.00007Data Points 5 4

R. Venkatesan et al. Ocean Engineering 171 (2019) 151–156

154

issues, NIOT implements various methods of anti-fouling approach,which includes copper guard protection, polyester tape on sensor casingand anti-fouling paints on frames. These resulted in reduced bio foulingon the surface layer.

6. Biofouling in the moored instrument

Euphotic is the zone where enough light for photosynthesis isavailable; in this zone many plants and other organism live and food isalso abundant. Biofouling on an underwater instrument is caused by thedevelopment of microscopic life forms (algae or bacteria) in the eu-photic zone. After growth in stages, they become large organisms, suchas shells and barnacles and get attached to the submerged objects in theocean. Biofouling is one of the limiting factors in ocean monitoring andit disrupts the quality of the measurements (Marc le menn, 2012;Delauney et al., 2010; Venkatesan et al., 2017). Thus, the underwatersensors moored in the OMNI Buoys system are prone to biofouling.

Venkat et al., studied biofouling in OMNI Buoy mooring which re-vealed that biofouling is predominant only up to a depth of 50m andLepas anatifera (goose neck barnacle) is the common biofoulant irre-spective of the location and water conductivity. For detailed in-vestigation, the biofouling deposited over the sensor was measuredupon retrieval of the OMNI buoy, which revealed that there were morebio fouling deposits in the surface layer and got reduced gradually in

Fig. 11. Bio Fouling in the CT sensors corresponds to the moored depth withthe mooring diagram.

Fig. 12. Fouling trend in the AS and BoB.

Fig. 13. Comparison of uncorrected and corrected data with reference shipbased CTD sensor.

Fig. 14. Picture of the field calibration setup (Left panel) a) Temperaturereading b) conductivity reading from all sensors at 1000m depth.

R. Venkatesan et al. Ocean Engineering 171 (2019) 151–156

155

the deep layer (Fig. 11). This analysis shows that the sensor moored inAS is highly affected by biofouling when compared to the sensors in thebuoy deployed in BoB (Fig. 12). This could be due to the high pro-ductivity in the Euphotic zone of AB when compared to the productivityin the BoB (Madhupratap et al., 1996).

7. Data correction

The abrasive scouring and contamination of cell is the limitingfactor for accurate conductivity measurement. We could make thecorrection in the measured data using the post cruise calibration co-efficient. Behavior of drift over the period of deployed duration is un-known. However, assuming the linear drift in the conductivity, the datawas corrected. The corrected data from the moored buoy was comparedusing the in-situ ship based CTD data which was taken just before theretrieval of buoy system and it revealed that the error in the data isreduced as compared to the uncorrected data (Fig. 13). Large drift wasobserved in the surface and thermocline layer in the uncorrected dataand the drift were reduced after the correction.

8. Field calibration

It is important to have post-cruise calibration in the same state ofcell geometry as identical to that during the retrieval of the sensor tohave a better understanding about the drift and data correction accu-rately. Allowing the cell to dry or keeping in salt water between theretrieval and post calibration could change the cell geometry. Fieldcalibration provides an opportunity to have similar cell geometry toperform the calibration.

Ocean is the good source for getting variable temperature andconductivity, as we go deep the stability is also much better. By usingthis ocean characteristic, the authors made comparative calibrationusing a reference sensor (Ship CTD) and this system was lowered alongwith the retrieved CT sensor moored in the buoy as shown in Fig. 14.CTD was lowered along with the retrieved sensors to pre-defined fixeddepths (200,400,600,800 and 1000m) to get a range of temperatureand salinity measurements in the ocean environment. During thesefixed depths the CTD was held until the sensor gets stable readings.Field calibration also revealed that the same result as laboratory cali-bration, i.e. temperature sensors are very stable and drift does not de-pend on the depth at which sensors are moored (Fig. 14a). However,conductivity sensors moored in euphotic zone (above 200m) driftedsignificantly (Fig. 14b).

9. Conclusion

Drift of conductivity and temperature sensors moored with OMNIbuoy system in AS and BoB was investigated using pre and post de-ployment calibration. It is identified that the temperature sensor is verystable and its drift is very minimum, within the accuracy limit and alsorevealed that drift does not depend on the sensor moored depth andperiod in function. However, the drift in the conductivity sensor is moresignificant and it depends on its moored position. Euphotic zone(Surface layer) of the ocean is highly prone to phytoplankton produc-tion and bio fouling due to availability of nutrient; hence, more drift inconductivity at surface layer and very less drift in the deep layer. Datacorrection based on post cruise calibration shows an improvement inconductivity measurement. Many techniques were evolved to preventthe biofouling on the instrument by researchers. However, only a fewhave been tested in-situ.

Acknowledgements

The authors thank the Ministry of Earth Sciences, Government ofIndia for funding this project. Authors are grateful to David Murphy,

Seabird Electronics USA and Robert A. Weller Woods HoleOceanographic Institution, USA for their suggestions and valuablecomments to improve this paper. The authors are indebted to theDirector of National Institute of Ocean Technology and the staff ofOcean Observation System (OOS) for their support. The Master, Crewand Vessel Management Cell are thanked for their support on board thevessel. The authors also thank Seabird Electronics USA for extendingtechnical support.

References

Ando, K., Matsumoto, T., Nagahama, T., 2004. Drift characteristics of a moored con-ductivity–temperature–depth sensor and correction of salinity data”. J. Atmos.Ocean. Technol. 22, 282–291.

Bihana, Caroline Le, Salvetat, Florence, Lamandé, Nolwenn, Compère, Chantal, 2014. Theeffect of the salinity level on conductivity sensor calibration. EPJ Web Conf. 7700015.

Diaz-Herrera, N., Esteban, O., Navarrete, M.C., Le Haitre, M., 2006. In situ salinitymeasurements in seawater with a fibre-optic probe. Meas. Sci. Technol. 17 2227.https://doi.org/10.1088/0957-0233/17/8/024.

Delauney, L., Compère, C., Lehaitre, M., 2010. Biofouling protection for marine under-water observatories sensors. Ocean Sci. 6, 503–511 – Europe.

Freitag, H.P., Sawatzky, T.A., Ronnholm, K.B., McPhaden, M.J., 2005. Calibration pro-cedures and instrumental accuracy estimates of next generation atlas water tem-perature and pressure measurement. NOAA Tech. Memo. ERL PMEL 115 89.

Freitag, H.P., McCarty, M.E., Nosse, C., Lukas, R., McPhaden, M.J., Cronin, M.F., 1999.COARE SEACAT DATA: calibrations and quality control procedures. NOAA Tech.Memo. ERL PMEL 115 89.

Wielgosz, Feistel, R., Bell, S.A., Camões, M.F., Cooper, J.R., Dexter, P., Dickson, A.G.,Fisicaro, P., Harvey, A.H., Heinonen, M., Hellmuth, O., Kretzschmar, H.-J., Lovell-Smith, J.W., McDougall, T.J., Pawlowicz, R., Ridout, P., Seitz, S., Spitzer, P., Stoica,D., Wolf, H., 2015. Metrological Challenges for Measurements of Key ClimatologicalObservables. Part 1: Overview, Part 2: Oceanic Salinity, Part 3: Seawater PH, Part 4:Atmospheric Relative Humidity. Metrologia.

Forch, C., Knudsen, M., Sorensen, S.P.L., 1902. Berichte uber dieKonstantenbestimmungen zur Aufstellung der hydrographischen Tabellen. DetKongelige Danske videnskabernes selskabs skrifter, 6 Raekke, Naturvidenskabelige ogMathematiske Afdeling. XII, 1. p 151.

Marc le men, 2012. Instrumentation and Metrology in Oceanography. ISTE - Wiley,Bookhttps://doi.org/10.1002/9781118561959.ch2.

Maes, C., 1998. Estimating the influence of salinity on sea level anomaly in the ocean.Geophys. Res. Lett. 25, 3551–3554.

Madhupratap, M., Prasanna Kumar, S., Bhattathiri, P.M.A., Dileep Kumar, M.,Raghukumar, S., Nair, K.K.C., Ramaiah, N., 1996. Mechanism of the biological re-sponse to winter cooling in the northeastern Arabian Sea. Nature 384, 549–552.

Millero, F.J., Feistel, R., Wright, D.G., McDougall, T.J., 2008. The composition of standardseawater and the definition of the reference-composition salinity scale deep-sea. Res.Ind. 55 (1), 50–72.

Nelson, R., Yamanaka, H., Maria Valnice, B., 2003. Electrochemical sensors: a powerfultool in analytical chemistry. J. Braz. Chem. Soc. 14 (No. 2), 159–173.

Prasanna Kumar, S., Prasad, T.G., 1999. Formation and spreading of Arabian sea high-salinity water mass. J. Geophys. Res. 104, 1455–1464.

Seitz, S., Feistel, R., Wright, D.G., Weinreben, S., Spitzer, P., de Bievre, P., 2011.Metrological traceability of oceanographic salinity measurement results. Ocean Sci.7, 45–62.

Shenoi, S.S.C., Shankar, D., Shetye, S.R., 2002. Differences in heat budgets of the near-surface Arabian Sea and Bay of Bengal: implications for the summer monsoon. J.Geophys. Res. 107 (c6). https://doi.org/10.1029/2000jc000679. 3052.

Seabird Electronic, 2009. Available at: http://www.seabird.com/document/conductivity-sensors-moored-and-autonomous-operation (last accessed 02.10.2017).

Seabird Electronic, 2016a. Available at http://www.seabird.com/sites/default/files/documents/appnote31Jun16.pdf (last accessed 28.09.2017).

Seabird Electronics, 2016b. Anti-foulant device. Available at. http://www.seabird.com/glossary, Accessed date: 14 October 2016.

Toggweiler, J.R., Key, R.M., 2003. Thermohaline Circulation. Encyclopedia of OceanSciences.

Tsuchiya, M., 1968. Upper Waters of the Intertropical Pacific Ocean. 4 Johns HopkinsOceanographic Series 50.

Tengesdal, Ø.A., Hauge, B.L., Helseth, L.E., 2014. Electromagnetic and optical methodsfor measurements of salt concentration of water. J. Electromagn. Anal. Appl. 6,130–139.

Venkatesan, R., Arul Muthiah, M., Ramesh, K., Ramasundaram, S., Sundar, R., Atmanand,M.A., 2013 a a. Satellite communication systems for ocean observational platforms:societal importance and challenges. J. Ocean Technol. 8 3.

Venkatesan, R., Shamji, V.R., Latha, G., Mathew, Simi, Arul Muthiah, R. R. Rao,Atmanand, M.A., 2013 b b. In situ ocean subsurface time-series measurements fromOMNI buoy network in the Bay of Bengal. Curr. Sci. 104 9.

Venkatesan, R., Yadava, Y.S., Atmanand, M.A., 2012. Training manual on best of prac-tices for instruments and methods of ocean observation. In: Regional Workshop onBest Practices for Instruments and Methods of Ocean Observation. WMO/DBCP, India200.

Venkatesan, Ramasamy, Kadiyam, Jagadeesh, SenthilKumar, Puniyamoorthy, Lavanya,Rajagopalan, Vedaprakash, Loganathan, 2017. Marine biofouling on moored buoysand sensors in the northern Indian Ocean. Number 2, March/April 2017. Mar.Technol. Soc. J. 51, 22–30 (9).

R. Venkatesan et al. Ocean Engineering 171 (2019) 151–156

156