Kongsfjorden (78.5–79°N, 11.6–12.6°E) is a fjord located in the north western part of the of the Norwegian Svalbard archipelago (Figure 1). It is approximately 20 km long, elongated northwest-southeast with a width varying from 4 to 10 km at its outer mouth. The bathymetry within the fjord varies from <100 m in the inner fjord east of the Lovénøyane and >400 m at its outer part. Glacier ablation, ice calving, snowmelt, and summer rainfall are the main sources for freshwater run-off (Svendsen et al., 2002). Within the Fjord there are strong influences of the subglacial meltwater from the four tidewater glaciers: Kronebreen, Kongsvegen at its head, Conwaybreen and Blomstrandbreen on its northern coast. These tidewater glaciers are also reported to be the main source of sedimentation (Svendsen et al., 2002). The circulation of the freshwater runoff, once it has reached the Kongsfjorden basin, is influenced by a number of factors such as tidal variation, meteorological conditions, subsurface currents and surface winds. Offshore circulation interacts with the hydrology inside the fjord and is constituted by the warm and salty Atlantic Water (AW, Ɵ > 3°C, S_A > 34.65 g/kg and σ_θ < 27.92 kg/m³) that is carried along the coast of Spitsbergen by the West Spitsbergen Current (WSC).

In summer it is mixed with Arctic water masses on the West Spitsbergen shelf and then advected inside the fjord as Transformed Atlantic Water (TAW), with characteristics and volume that significantly vary from year-to-year (TAW, Θ = 1.0–3.0°C, S_A > 34.65 g/kg and σ_θ < 27.92 kg/m³; (Stefano Aliani et al., 2016; Svendsen et al., 2002); (Hodal et al., 2012; Bazzano et al., 2017). In a simple two-layer system inside the fjord, TAW is close to the bottom with relatively cold, less salty water of polar origin above (Cottier et al., 2005). In the inner fjord and in the deep and bottom layer, two other waters masses are described by (Svendsen et al., 2002): the very cold and low saline Local Water (LW, Θ < 1.0°C) originated inside the fjord in autumn/winter due to conventional processes that sink water passing in front of glaciers; the Winter Cooled Water mass (WCW; T < −0.5°C, S > 34.56 g/kg) then sinks, filling the depressions on the bottom in front of glaciers. WCW is the result of a deep convection on very cold and dense water due to winter cooling and sea ice formation (Svendsen et al., 2002). Cottier et al. (2005) identified two additional water masses named Surface Water (SW, Θ > 1.0°C, S_A < 34.00 g/kg) and Intermediate Water (IW, Θ > 1.0°C, S_A = 34.00–34.65 g/kg). The SW spreads over the LW and is produced by glacial melting. This several metres-thick layer decreases as it moves from the inner part of the fjord to its mouth. The drivers of circulation in the inner Kongsfjorden are wind, basin oscillations due to changes in pycnocline and glacier discharge that creates a very stable stratification, but also fronts and small-scale turbulent eddies a few meters in diameter and vortex filaments in front of glaciers (Svendsen et al., 2002). Winds mainly act on the sea surface. In case of cross-shore winds there is an induced tilt in the free surface and the strong pycnocline, that separates the circulation of upper layers from deeper ones, but the total volume of water in the fjord remains constant. On the contrary, with alongshore winds the free surface at the mouth of the fjord drives either a baroclinic outflow (i.e., toward the fjord mouth) of waters or an inflow (i.e. toward the glacier front) (Stefano Aliani et al., 2016; S. Aliani et al., 2004). In summer and in a relatively ice-free environment, the general circulation of the water masses inside the fjord is anticlockwise with warm and salty mainly TAW and less AW entering from the outer fjord, reaching the tidewater glaciers Kronebreen-Kongsvegen. Then they move cooler and less salty north-westward along the coast east of the Lovénøyane shoals and exit in the outer fjord between the Blømstrand island and the northern coast, constrained by topography, with velocities less than 8–10 cm/s (Aliani et al., 2016; Svendsen et al., 2002; Aliani et al., 2004). Nutrient availability is one of the most direct mechanisms by which glacial discharge affects primary productivity of phytoplankton in polar regions (Hodal et al., 2012). Within Arctic regions most glacial discharge enters the ocean through fjord systems and as such these marine settings are heavily influenced by glacial runoff, and that is set to increase in the near future due to climate change. Consequently, there is a pressing need to generate robust in-situ analytical measurements on a continual basis at multiple locations to provide data on the ever-changing chemistries not only in Arctic waters but also on a global scale. Sample locations within the fjord are shown in Figure 2, water samples were acquired using 10 L Niskin bottles on board the MS Teisten research vessel, hydrological casts were also performed using a Seabird 19plus V2 SeaCAT Profiler CTD probe with turbidity and dissolved oxygen sensor (Figure 3).

Electronic Control was achieved via an Arduino Mega 2,560 microcontroller board (Radionics, Ireland). Raw data in BIT values were displayed on an LCD (Radionics, Ireland) while an 8 GB SD card was connected via a Data Logger. The LED intensity within the platform was controlled using a PWM (Pulse Width Modulation) driver, which enabled the intensity of the LED for the chosen colorimetric method to be optimised. The benchtop sensing platform shown in Figure 4 was developed as a low-cost sensing platform for the detection of NO²⁻, NO³⁻ and PO₄ ³⁻ in natural waters. The platform was fabricated using 4 mm thick black polymethylmethacrylate (PMMA) panels which were laser cut using an Epilog Zing Laser cutter and PMMA was solvent bonded together using 1,2-Dichloroethane (Sigma Aldrich). The sensing platform (20 cm × 6 cm x 3.5 cm) comprised of four detection chambers: chamber one and two for the detection of PO₄ ³⁻ and chamber three and four for detection of NO₂ ⁻ and NO₃ ⁻. LEDs were selected based on the lambda max (λ _max ) specific for each colorimetric chemistry (Donohoe et al., 2018; García-Robledo et al., 2014), optical detection within each chamber was achieved using a 375 nm LED (Roithner, Wien, Austria) for PO₄ ³⁻ and 540 nm LED (Roithner, Wien, Austria) for both NO₂ ⁻ and NO₃ ⁻. The photodiode (PD) (Roithner, Wien, Austria) was used for absorbance detection in all chambers.

The alignment of the LED and photodiode was achieved through a 3D printed (Stratasys Objet260 Connex1) custom designed holder, through which the LED photodiode was easily removed and replaced while maintaining alignment (Figures 4C,D). This improved detector stability, while also reducing unit cost and simplifying servicing of the detection unit which is critical during field deployments in remote locations.

Figure 5 presents the linear calibrations generated from the validation of the colorimetric chemistries in each chamber of the sensing platform. These results demonstrate that the benchtop sensing platform achieved excellent correlations for all linear calibrations within each chamber; chamber one and two; 0.0–20 μM PO₄ ³⁻ (R² = 0.999), chamber three; 0–10 μM NO₂ ⁻ (R² = 0.999), chamber four; 0–10 μM NO₃ ⁻ (R² = 0.999), the tightness of the error bars in all cases indicates the excellent precision of each method. The limit of detection (LOD) of the sensing platform was estimated using three times the standard deviation (n = 10) of the lowest concentration test solution, corrected with the blank measurement and was found to be 0.05 μM for NO₂ ⁻ and NO₃ ⁻, and 0.03 μM for PO₄ ³⁻.

During a cruise in Kongsfjorden on the June 22, 2016 a total of 55 water samples (10 ml aliquots) were acquired for nutrient analysis using 10 L Niskin bottles. Water samples were acquired from the surface <1 m, middle and bottom depths depending on the sample station. A total of 23 hydrological casts were performed using a Seabird 19plus V2 SeaCAT Profiler CTD probe with turbidity and dissolved oxygen sensors. Using the Seabird 19plus V2 SeaCAT vertical profiles were performed with a frequency of 1 Hz, from the surface to the bottom and the data was download at the end of the cruise. Nutrient concentrations found in surface samples (<0.5 m) are shown in Figure 6). Nutrient concentrations varied throughout the fjord depending on the depth and influences (e.g, Atlantic waters, glacial melt, suspended solids) associated with location of the sampling station. In surface samples (<0.5 m), elevated nutrient values were observed south and east of Blomstrand Island (019; 9.74 μM NO₃ ⁻, 7.14 μM PO₄ ³⁻, 018; 6.54 μM PO₄ ³⁻, 013; 7.39 μM PO₄ ³⁻).

At the southern-most part of the fjord, nearest the Kronebreen-Kongsvegen glacial fronts at station 009b (see Figure 7), NO₃ ⁻ concentrations increased with depth in samples analysed from the <0.5 m (4.23 μM NO₃ ⁻) to 50.0 m (9.54 μM NO₃ ⁻) and 101.1 m (9.94 μM NO₃ ⁻). At the time of sampling cast 009b was located in front of the Kongsbreen glacier inside a depression of a semi-closed bay with a maximum depth of 105 m . When compared to the hydrological data, the NO₃ ⁻ concentrations recorded coincided with high turbidity values (22.29 NTU), low potential temperature and high absolute salinity (Θ = 0.19°C, SA < 34.83 g/kg) that is associated with the WCW (Figure 7). These elevated turbidity values could possibly be attributed to the sinking of suspended solids of glacier origin from the surface as a turbidity value of 22.29 NTU was recorded at a depth of 2 m at the same station, where a water mass was physically identifiable from the ΘS as SW (Figure 7).

In the outer region of the fjord at the station 001, samples also show an increase in NO₂ ⁻, NO₃ ⁻ and PO₄ ³⁻ concentrations with increasing depth from <0.5 m (4.14 μM NO₂ ^–, 0.17 NO₃ ⁻ μM, 1.37 μM PO₄ ³⁻) to 150.83 m (4.14 μM NO₂ ⁻, 5.10 NO₃ ⁻ μM, 22.63 μM PO₄ ³⁻) and 290 m (6.43 μM NO₂ ⁻, 8.48 NO₃ ⁻ μM, 25.83 μM PO₄ ³⁻) (Figure 7). To further investigate the change in nutrient concentration and hydrological data with depth, Figure 8 shows a transect along the coast from the outer fjord (001) to the inner fjord (008). ODV maps show that waters close to the Kronebreen-Kongsvegen glaciers are characterized by lower temperature and salinity, as would be expected. Towards the outer section of the fjord, increasing salinity is most likely due to the influence of Atlantic waters. As depth increases, so does NO₃ ⁻ concentration, with maximum values occurring close to the bottom from the inner to the outer fjord, while temperature decreases. Concentrations of NO₂ ⁻ and PO₄ ³⁻ do not follow the same trend and show a uniform distribution in the fjord except for station 001 where both concentrations were significantly elevated (22.63 μM PO₄ ³⁻ compared to the average of all the other stations of 2.68 μM; and 4.14 μM NO₂ ⁻ compared to the average of 0.36 μM). This increase of phosphate and nitrites at 001 may arise from the anticlockwise summer circulation described by (Svendsen et al., 2002; Aliani et al., 2016; Poll et al., 2021).