Salinity (± 0.01) and water temperature (± 0.002 °C) were measured using a thermosalinograph (SBE 45 MicroTSG). A ion-sensitive field-effect transistor (Honeywell Durafet II) was used to determine pH (± 0.005) as described by Martz et al. (2010). Dissolved oxygen (DO, ±0.06%) was measured with an oxygen optode (Aanderaa 3835) calibrated against Winkler titration. Fluorescent dissolved organic matter (fDOM) was analysed with a fluorometer on a multiparameter Sonde (YSI EXO2), which has a detection limit of 0.07 ppb QSE. Dissolved inorganic carbon samples were acidified in a DIC analyser as described by Friederich et al. (2002), to transfer all carbonate species into CO~2~, and subsequently analysed with a non-dispersive infrared absorption (NDIR) detector (LI-COR LI-7000), with a precision better than 0.1%. Continuous CO~2~, N~2~O, CH~4~ (1 min intervals) and radon-222 (222^Rn, min intervals) concentrations were measured via a showerhead equilibrator coupled to gas analysers (Santos et al., 2012). Concentrations of CO~2~ and CH~4~ were analysed during SV1 and time series at GI with a mobile gas concentration analyser (Picarro GasScouter^TM G4301) with a raw precession of 0.4 ppm and 3 ppb, respectively. Nitrous oxide concentrations (±2 ppb), as well as CO~2~ (±10 ppm) and CH~4~ (±10 ppb) concentrations for SV2 and SR time series, were analysed with by cavity ring-down spectroscopy (Picarro G2308) (Maher et al., 2013b). A radon detector (Durridge RAD7) was used to measure ^222^Rn (Burnett et al., 2001). Equilibration times of all gases were corrected according to Webb et al. (2016). For TAlk and OAlk analysis, samples were filtered with 0.7 µm GFF filters and measured within one day, using a titrator (Metrohm 888 Titrando with Tiamo light), which has a precision better than 5 µM. Deviations and drifts in the acid concentration (0.05 M hydrochloric acid) were accounted for, using certified reference materials (CRM batch 175 and CRM batch 178) according to Dickson (2010). The method developed by Cai et al. (1998) was adapted for OAlk determination. Following the first Gran titration, which determines TAlk, samples were purged with high purity nitrogen for five minutes to remove the CO~2~ and then back-titrated to the original pH with sodium hydroxide (0.05 M). Subsequently, a second Gran titration was undertaken on the sample, which determines the non-carbonate alkalinity. To calculate OAlk, results were corrected for borate alkalinity, which was calculated from TAlk, pH, temperature and salinity using CO2SYS (Lewis & Wallace, 1998). For DOC analysis, samples were filtered with 0.7 µm GFF filters, preserved with phosphoric acid and analysed with a total organic carbon analyser (Shimadzu TOC-L CSH/CSN), which has a precision better than 2%.