Autonomous in Situ Measurements of Seawater Alkalinity Reggie S. Spaulding,* ,† Michael D. DeGrandpre, ‡ James C. Beck, † Robert D. Hart, ‡ Brittany Peterson, ‡ Eric H. De Carlo, § Patrick S. Drupp, § and Terry R. Hammar ⊥ † Sunburst Sensors, 1226 W Broadway, Missoula, Montana 59802, United States ‡ Department of Chemistry and Biochemistry, University of Montana, 32 Campus Drive, Missoula, Montana 59812, United States § Department of Oceanography, University of Hawaii, Manoa, Hawaii 96822, United States ⊥ Woods Hole Oceanographic Institution, 266 Woods Hole Road, Woods Hole, Massachusetts 02543, United States * S Supporting Information ABSTRACT: Total alkalinity (A T ) is an important parameter for describing the marine inorganic carbon system and understanding the effects of atmospheric CO 2 on the oceans. Measurements of A T are limited, however, because of the laborious process of collecting and analyzing samples. In this work we evaluate the performance of an autonomous instrument for high temporal resolution measurements of seawater A T . The Submersible Autonomous Moored Instru- ment for alkalinity (SAMI-alk) uses a novel tracer monitored titration method where a colorimetric pH indicator quantifies both pH and relative volumes of sample and titrant, circumventing the need for gravimetric or volumetric measure- ments. The SAMI-alk performance was validated in the laboratory and in situ during two field studies. Overall in situ accuracy was -2.2 ± 13.1 μmol kg -1 (n = 86), on the basis of comparison to discrete samples. Precision on duplicate analyses of a carbonate standard was ±4.7 μmol kg -1 (n = 22). This prototype instrument can measure in situ A T hourly for one month, limited by consumption of reagent and standard solutions. ■ INTRODUCTION Atmospheric CO 2 has increased from 280 to 400 ppm over the past ∼150 years as a consequence of industrialization. 1,2 The oceans have prevented an even greater increase by absorbing a significant amount of anthropogenic CO 2 . 3,4 Seawater pH has decreased by more than 0.1 pH units from conversion of the absorbed CO 2 to carbonic acid, 5,6 an undesirable side-effect of ocean CO 2 uptake. The decreased pH will potentially change biological processes in the oceans. For example, as the oceans become more acidic, aragonite and calcite saturation states become lower, 5-7 and it becomes more difficult for calcifying organisms to produce shells. 7,8 The predicted detrimental effects on marine calcifiers could propagate through the food chain, altering entire ecosystems. 9,10 The inorganic carbon parameters, including saturation states, can be calculated if two of the primary parameters, partial pressure of CO 2 (pCO 2 ), total hydrogen ion concentration (pH T ), total alkalinity (A T ), and total dissolved inorganic carbon (C T ), are known. The carbonate system is most accurately defined by measuring either pH or pCO 2 in combination with either A T or C T . 11,12 In situ instruments are currently available to measure only pH 13-15 and pCO 2 , 16-18 although some progress has been made toward in situ C T measurements. 19-21 Benchtop automated, flow-through instru- ments have been developed for A T analysis, 22-24 and an in situ potentiometric A T titrator was developed, 25 but to our knowledge no in situ A T data have been reported. Thus, A T data are limited to shipboard determination during cruises or intense efforts with manual or robotic sample collection followed by manual analysis. 26 In many cases it is possible to estimate A T using a relationship with salinity 27 or salinity and temperature, 28 because evaporation, precipitation, and mixing strongly control A T . However, these conservative relationships are not always followed. For example, in the Saragasso Sea, A T is generally conservative with salinity, but a several month-long drawdown of 25-30 μmol kg -1 , likely due to coccolithophore calcification, was detected. 29 After spring blooms in the northern Bay of Biscay, A T is consistently depleted by coccolithophore calcification, and thus nonconservative with salinity. 30 In many coastal areas consistent salinity-A T relationships do not exist because of riverine inputs, CaCO 3 dissolution, and anaerobic processes. 31 On the Bering Sea shelf the relationship between A T and salinity can be highly variable, due to coccolithophore calcification and CaCO 3 mineralization. 32 In coral reef ecosystems, A T variability is typically dominated by CaCO 3 formation and dissolution. 26 Thus, A T -salinity relation- Received: April 2, 2014 Revised: July 13, 2014 Accepted: July 22, 2014 Published: July 22, 2014 Article pubs.acs.org/est © 2014 American Chemical Society 9573 dx.doi.org/10.1021/es501615x | Environ. Sci. Technol. 2014, 48, 9573-9581