An In Situ Analyzer for Accurate Dissolved Carbon Dioxide and Total Inorganic Carbon Measurements

Inland waters have generally been viewed as passive conduits for carbon (C) movement to the sea. However, recent estimates suggest that approximately half of the C entering aquatic systems from the land actually reaches the ocean. Understanding the processes and mechanisms responsible for net C losses en route to the sea is a leading challenge in biogeochemistry. Formulation of an integrated C budget for inland waters will require advances in chemical sensor networks for both organic and inorganic forms of C. We describe here a new in situ chemical analyzer for the inorganic carbon system and illustrate findings for lake and stream deployments. In conjunction with thermodynamic acid/base equilibrium calculations, a complete characterization of the dissolved inorganic carbon system can be accomplished by accurately measuring both pCO2 (partial pressure of dissolved CO2) and TCO2 (total inorganic carbon). Unfortunately, due to the expense and complexity of current instruments, continuous (e.g., hourly) pCO2 in situ monitoring has been performed for long periods only in ocean systems. Affordability and complexity still hinder routine deployments of similar pCO2 instrumentation in freshwater systems. Moreover, in situ devices for continuous TCO2 monitoring are still in development and are not yet practical for routine deployments in freshwater or marine ecosystems. Further, though manual monitoring of CO2 (via headspace sampling and laboratory analysis) and TCO2 (via grab sampling and laboratory analysis) are both feasible, high frequency sampling programs are difficult to sustain long-term; thus, the diurnal and seasonal cycles of events are not resolvable practically with such an approach. Our new device provides for continuous (high resolution) or manual in situ measurements of both pCO2 and TCO2 in groundwater and surface water, allowing complete characterization of the inorganic carbon system. Initial lake and stream deployments for continuous and manual sampling were performed in 2006 and 2007. Accuracy and precision (<0.15% coefficient of variation) met or approximated the data quality requirements for large-scale biogeochemical research initiatives (e.g., Joint Global Ocean Flux Study). A continuous buoy deployment for pCO2 provided a high resolution (30 minute intervals) record of diurnal lake metabolism during the summer of 2007. In a 24-hr test deployment we obtained a 3 minute temporal resolution. Synoptic surveys of 51 northern Wisconsin lakes revealed that pCO2 was sub-atmospheric (<370 ppm) in a majority of lakes (n=40), suggesting the lakes generally served as atmospheric C sinks, rather than atmospheric C sources, during early and mid summer. Further, TCO2 data suggested that carbonate anions helped suppress pCO2 below atmospheric levels during periods when R > P. Surface maps (0.5 m depth) of selected lakes revealed dynamic and heterogeneous pCO2 in near shore areas and relatively homogenous patterns in deeper zones. High resolution spatial patterns along the perimeter of one lake revealed likely entry points of groundwater discharge. Similarly, longitudinal surveys along a small stream corridor revealed patterns of pCO2 and TCO2 consistent with known groundwater discharge and recharge features and showed the primary importance of the groundwater discharge segments as sources of CO2 efflux to the atmosphere.