Dataset: BATS Validation bottle data
Deployment: BATS_cruises

Cruises
Data were collected on BATS Validation (BVAL) cruises, from cruise #1 (April 1991) through BATS cruise # 60 (June 2023). Please note that BVAL cruises 4, 25, 43, and 54 were cancelled and hence no reporting. However, all bottle files are included for this dataset, even those that do not have any BATS core parameters, to provide data from ancillary measurements for re-use.  

Research was conducted  on the R/V Weatherbird II through 2005 and thereafter on the R/V Atlantic Explorer. There were numerous Chief Scientists for these cruises including Rachel Dow, Anthony Michaels, Kjell Gundersen, Rodney Johnson, Paul Lethaby, Mike Lomas, Steven Bell, Gwyn Evans, and Claire Medley. Cruise track information showing CTD station locations and cruise dates for each BVAL cruise conducted between 1991 and 2022 are available in the Supplemental Files section below. 

Water sampling
Full depth water sampling and data collection at the BATS site are achieved with a total of three hydrocasts using a General Oceanics Intelligent Rosette® with an array of 24 12L water bottles and a Sea-Bird Scientific CTD system. Water samples are collected during the upcast with a 1-minute resting period between reaching the sampling depth and triggering the bottle to close.  Bottom measurements/ sampling are achieved within 20 meters from the bottom, as determined using an altimeter.

Water samples are taken right after Rosette® recovery. On any cast, if only a single water bottle is collected to sample all biogeochemical parameters, then gas samples are collected first due to their exposure to air when opened. However, if enough bottles are available, two bottles can be taken for a single depth. Water is usually split between large-volume particulate samples (POCN, HPLC, POP and PSi) and all other small volume samples, including gas samples. When two bottles are taken for a single depth, particulate samples are collected first to prevent settling within the Niskin bottle. Samples are fixed or frozen once all same-sample bottles from one cast have been collected. Particulate samples are filtered as soon as collected.

Nutrients
The BATS nutrient methodology is based on the Protocols for the Joint Global Ocean Flux Study (JGOFS) Core Measurements (Intergovernmental Oceanographic Commission, 1994) which describes the method for the determination of dissolved inorganic macronutrients in seawater: nitrite (NO2 – ), nitrate + nitrite (NO3 – + NO2 – ), orthophosphate (PO4 3 – ) and reactive silicate (Si(OH)4) using Continuous Flow Analysis (CFA). 

While the definition of the dissolved fraction has changed throughout the years that the BATS time series has operated, the pore size used has remained constant in order to create a comparable temporal dataset. While similar studies in oligotrophic ocean regions have opted to forego the use of nutrient filters under the assumption that the particulate nutrient pool is negligible, we continue the use of filters for the sake of continuity. Sample filtration also removes the potential for turbidity-derived uncertainties during analysis, and may aid preservation of frozen samples.

Discrete samples are collected at the Bermuda Atlantic Time-series Study (BATS) site from surface to bottom depths (∼4,200 meters). Sea water is filtered directly from the Niskin spigot using a 0.8 µm membrane to remove particulates. Collected sea water is preserved by freezing until analysis. Replicate samples are taken during each cast to ensure quality control standards are met during analytical and data processes. Dissolved inorganic nutrients are measured using a SEAL AA500 Autoanalyzer by Continuous Flow Analysis (CFA). During this process, a subset of sample is drawn and further split into four different channels driven by a peristaltic pump. The sample stream is segmented with air or nitrogen bubbles throughout the flow path to enhance the mixing of reagents with the sample. The nutrients, NO₂-, nitrate + nitrite (NO₃– + NO2-) , PO₄^3– and Si, are chemically reacted in the separate channels to produce a color change and are measured colorimetrically at different wavelengthsusing a flow-through colorimeter located at the end of the flow path. The light absorption by the sample-reagent mixture is proportional to the concentration of nutrient in the sample according to the principles of the Beer-Lambert Law. Raw absorbance units are converted into nutrient concentrations according to a linear calibration curve formulated from known standards.