Scottish Coastal Observatory Data

Nutrient data: TOx N, DIP, DSi and Ammonia are collected using the methodology below.Water samples for nutrients at four depths (1 m, 5 m, 10 m & 45 m at Stonehaven and 1 m, 5 m, 10 m and 35 m at Loch Ewe) were collected using a Niskin Sampling bottle, which was also fitted with a digital reversing thermometer. At all other sites surface water samples only were collected. Samples for nitrate and phosphate were stored in glass bottles at -20oC and allowed to thaw for 24 hours before analysis. Samples for silicate analysis were stored in plastic bottles and either stored in a refrigerator maintained between 0 and 8oC (1997 - 2010) or at -20oC (2011 onwards). Refrigerated samples were allowed to come to room temperature before analysis while frozen samples were thawed in the dark for 48 hours to allow for depolymerisation.Water samples were analysed for DIP, TOxN and DSi by colorimetry using a continuous flow analysis (CFA) system based on the techniques of Armstrong et al., (1967) for oxidised nitrogen (nitrate plus nitrite), Murphy and Riley (1962) for PO4 and Koroleff (1971) for DSi. Samples collected prior to 2002 were analysed using a Skalar system. From 2002-2006 a Bran and Luebbe Analyser (AA3) and from 2006 a Bran and Luebbe QuAAtro (SEAL Analytical, UK) was used. Aliquots of each sample (5 ml) were aspirated using an auto-sampler and transmitted through the complete system by a peristaltic pump. The sample stream was split with a pump continuously adding the reagents required to each sample stream using tubes of a specific internal diameter. Air bubbles are evenly pumped through each stream to reduce inter-sample dispersion and improve sample throughput. Once the chemical reaction to allow detection was complete each stream flowed through a specific detector. Concentrations were determined with the wavelengths set at 520 nm for TOxN and at 880 nm and 820 nm for PO4 and DSi, respectively. Dual-beam operation with same-wavelength compensation gives low drift, even at very high sensitivities. Data was collected as an analogue voltage that was fed through a digitising interface to a computer and analysed using either the Bran & Luebbe data processing package Versions 5.31 – 6.02 or with the Skalar data processing package Version 6. Further details of the methodology can be found in Webster et al., (2007), Rose et al., (2009) and Smith et al., (2014). Determination of Ammonia: Samples were analysed for ammonia using a manual method. Buffer solution was added to the samples to prevent the precipitation of calcium and magnesium hydroxides, followed by the phenol catalyst and an oxidising agent (dichloro-s-2,4,6-trione). Samples were then incubated and exposed to UV light for at least 40 minutes. In the presence of the phenol catalyst and excess chlorine, ammonia reacts with phenol to form indophenol blue, which was determined calorimetrically at 630 nm using a spectrophotometer. Further details of the methodology can be found in Webster et al., (2007) and Rose et al., (2009).Since 2012 all nutrient data and the sample information from the Oceanography Environmental Record sheets is held on LIMS. Prior to 2012 nutrient data was saved by analytical Batch number on a shared folder on the internal computer system and in-house developed data base (Chemdat). Nutrient data was sent to the MSS Oceanography Group who would add the nutrient data on to their own databases along with the sample information. Quality Control: A system suitability standard (highest calibration standard) was analysed each day using the CFA. Calibration standards covering the range 0.2-19.3 µM for TOxN, 0.2-13.3 µM for DSi, 0.05 - 5 µM for PO4 were analysed at the start and the highest concentration analysed at the end of each batch, to account for drift. In both cases correlation coefficients of at least 0.999 were achieved for all nutrients. The limits of detection were dependent on the instrument used and were based on 4.65 times the standard deviation of the mean value from repeat analysis of ten low standards and can be found in Webster et al., (2007), Rose et al., (2009) and Smith et al., (2014).Calibration standards, covering the range 0.2 to 10.00 µM for ammonia, were prepared in low nutrient sea water (LNSW) and analysed with each batch. The absorbance readings were used to compute the calibration curve. Correlation figures of at least 0.995 were achieved. The limit of detection (LoD), based on 4.65 times the standard deviation of the mean value from repeat analysis of ten low standards was 0.19 µM.Prior to 2010, QUASIMEME sea water samples were used as reference materials, whilst from 2010 onwards reference standards were prepared from standards procured from OSIL (Havant, Hampshire, UK). The data obtained from the reference samples were plotted on Shewhart charts with warning and action limits drawn at ± 2 x and ± 3 x the standard deviation of the mean. Further quality control was assured thr

This data set comprises the monitoring data collected as part of the Scottish Coastal Observatory and represents the raw data summarised in the data file http://dx.doi.org/10.7489/1761-1.