CHEMISTRY
Marine chemistry data can represent measured or estimated values of pH, oxygen concentration, nutrient concentration, etc.
Data Collection
The chemical composition of seawater is influenced by many processes, such as ocean currents, atmospheric conditions, seafloor processes (erosion, underwater volcanoes, hydrothermal vents), and not least human activity and climate change. Temperature and salinity are the core parameters that are frequently measured, as these parameters determine the density of seawater and thus can be used to model the large-scale density-driven currents in the world ocean. pH changes due to ocean acidification have become increasingly important as scientists gain a better understanding of how the ocean takes up excess carbon dioxide from human activities such as the burning of fossil fuels. When carbon dioxide dissolves in seawater, the chemical process leads to the production of carbonic acid, which in turn increases the amount of hydrogen ions in the water, thus lowering the pH. This may have great implications for marine biology, as some species are highly sensitive to pH and may, therefore, not be able to survive in their habitats.
Water chemistry at the benthic boundary layer of the world’s oceans represents exchange, and sometimes equilibrium, between the underlying sediments and the water column above. Concentrations of solutes that chemical oceanographers, geological oceanographers, biogeochemists, and marine geologists, among others, are interested in measuring can differ between the sediment porewaters and water column due to interaction with the sediment. Therefore, the sediment-water interface typically contains a high rate of change across the interface relating to the fluxes between sediment and water. Thus, measuring chemical gradients at the benthic boundaries is important for many types of interdisciplinary research efforts. As such, data collection and management procedures must be defined to achieve intercomparability of different research efforts.
In Florida marine and coastal waters, the sediment-water interface includes carbonate sediments formed from precipitation in seawater, lithogenous sediments from physical and chemical weathering of terrestrial rocks, organic-rich peat deposits from marshes and mangroves, and coral reef carbonates. These diverse bottom types involve different types of data generation and treatment. The difficulty of sampling the open ocean by ship limits chemical measurements to a few specialized tools. Ocean chemists and chemical oceanographers will measure data from the entire water column, generally stopping profiling data loggers (CTDs, for instance, a term commonly used for a set of instruments measuring Conductivity, Temperature, and Depth in conjunction with other sensors) approximately one meter above the surface to avoid contaminating the water column sensors with higher porewater concentrations and sediment. Those interested in the interface between sediments and water more so than the water column above, such as marine geologists and geological oceanographers, generally attempt to capture gradients at this interface with box cores and multicores. However, they disregard the sediment-water interface when obtaining deeper cores obtained by gravity core or piston core. Between these two sampling regimes – water-column and surficial-sediment – biogeochemists have depended on coring techniques that preserve the interface (multicoring and some box coring) for measurements over finite temporal and spatial domains, as well as landers that can make microscale measurements of chemical exchange at the seawater-sediment interface continuously for an amount of time determined by battery power, storage, and measurement payload.
Environments such as marshes, mangroves, and coral reefs in and around Florida can often be sampled by divers from shore or small boats, or even by researchers on foot. Although tools like bottom landers are not frequently used in these environments due to the complex bathymetry of bottom type and tidal influence, sensors can be used to measure water levels, pressure, temperature, conductivity, and some chemical variables. Data collection by individuals in coral reefs and coastal ecosystems ranges from collecting visual semi-quantitative data (quadrats, species identification, tree biomass measurements) to sampling of water and sediment (including coral fragments) for more detailed laboratory analysis. Because such sampling efforts typically arise from individually defined scientific questions and do not depend on shared research facilities like research vessels, the types of data collected in any one experiment or monitoring project can differ entirely from other efforts.
No matter the technique, some benthic environments (coastal systems, estuaries) are more dynamic than others (abyssal environments). In more dynamic environments, when discrete samples are taken, it may not be possible to cover all states of the system with adequate sampling. Therefore, it is necessary to describe variables like time, meteorologic conditions, tide phase, and other pertinent observations with any benthic chemistry data to contextualize discrete measurements. Such observations can also be helpful when time series data are collected, for instance from a lander, so deviations from the background conditions may be explained during data analysis.
Water column data from oceanographic expeditions are almost universally collected from Seabird CTDs. In addition to conductivity, temperature, and depth (generally measured through pressure), multiple other sensors, including oxygen concentration, transmissivity, fluorescence, and others that may be important to benthic chemistry, can make up the payload of these versatile instruments. CTDs are most often connected to rosette bottle samplers with multiple Niskin bottles that can be closed remotely based on CTD data communicated to the ship through the wireline during real-time deployment. Typically, CTDs record data at a high frequency during the downcast and upcast, with water sampling bottles being closed (fired) near the bottom and at targeted depths during the upcast based on real-time visualization and analysis of water column data from the downcast through an undisturbed water column. Other data collections methods include landers and submersible and peristaltic pumps.
Data Processing
Generally, Seabird data files need to be processed with proprietary Seabird software to produce files that are readable in most data analysis and visualization programs like Excel, Matlab, or Python infrastructure. Once processed, any such platform can be used to obtain metadata, such as ship latitude and longitude (measured one time for each cast), timestamps, and depths of bottle samples. Finite water column standards, like those used to measure nutrients, are often tied to the CTD metadata. Seabird, for instance, generates bottle files (.bl) with their sample equipment that relate the time it took for a bottle firing sequence to be sent and received to the indices of specific water column data recorded by mounted sensors during the sequence. These data can be averaged from the upcast or linked to the downcast by depth to characterize conditions during bottle sampling or conditions targeted for bottle sampling, respectively. Bottle samples, however, allow chemists to measure variables not recorded by CTD sensors. Almost without exception, concentrations of these targeted variables will be higher if the bottles were fired within the rapidly changing boundary layer depth nearest the sediment-water interface. Often, it is necessary to dilute these samples differently than other water column samples, and turbidity or transmissivity data from the CTD sensors can help determine when this may be necessary. Chemical analysis of water samples from the benthic boundary layer must then clearly include information about dilution because this can affect the propagation of uncertainty through the dilution factor. It is recommended that the measured variable and the dilution be treated as two observations rather than reporting the interpretation of the undiluted concentrations without explicitly reporting the dilution factor and actual measurement.
Data Management
Often, many of the required metadata are produced directly during sampling. For the others, Jiang et al. (2022) include information about data header conventions, conventions for quality control flagging of data and missing value, and other specific metadata information. These standards also use the Exchange data file format of the World Ocean Circulation Experiment (WOCE), the Climate and Ocean Variability, Predictability, and Change Program (CLIVAR), and the Carbon Hydrographic Data Office (CCHDO). Observational data are frequently stored in a tabular format to accommodate the substantial number of potential measured variables at a single observation. This format does lend itself to relational database designs frequently used for spatially fixed observation data. However, little direction exists on how water quality data should be formatted in terms of tabular content (e.g., pertinent, optional, or required fields) or how it should be distributed for outside models. As referenced earlier in the sampling design considerations, proper planning of a sampling strategy is crucial to maximize the effectiveness of data collection and relaying a project’s findings. Data distribution and documentation should be no different by providing a set schema of common units, field order, attribute domains, and field nomenclature conventions.
Even when an opportunistic or passive sampling program sampling design is chosen, a potential environmental impact guidance report can aid with post-collection data handling, modeling, and distribution by providing additional metadata considerations that outside managers/researchers may not be familiar with in the study area. Environmental checklists are available from state, national, and international agencies that provide a checklist, guidance, scoring systems, and instructions to identify potential impacts across various habitats. One comprehensive example is the RESTORE Act Environmental Checklist from the United States Department of the Treasury. This form contains questions and contact information to identify applicable laws that may apply to planned research activities, as well as necessary forms and documents from listed agencies that apply to the planned research activities. The importance of environmental impacts will ultimately depend on the objective and the target sampling variables. Hence, any agency checklist must be adapted for the specific monitoring program.