Prevención de riesgos laborales 1 Lactancia

This study demonstrates the capability of estimating N using measurements obtained by an autonomous glider. Both c(O2) and c(DIC) were used to create mass budgets, taking into

account physical fluxes relating to horizontal advection, air-sea exchange, and mixing. It was clear that on timescales of less than a month, physical fluxes can be significant and should not be ignored when estimating N. Daily mean N fluxes estimated using mass

3.9 Conclusion 69

Table 3.3: The associated mean errors of the fluxes, and individual components, used to calculate N using c(O2) (blue), and c(DIC) (green), and the errors associated with N

calculated using buoy measurements (yellow).

budgets were significantly affected by advection. N fluxes estimated using buoy c(DIC) with/ without including advection were useful for comparing with N estimated using glider measurements during certain time periods, although calcification and entrainment effects were not explored. The spring bloom started in this area of the northwestern Mediterranean Sea on or around March 19 2016. This date was identified using trends in physical and biogeochemical measurements obtained by the glider and the buoy, and N estimated using all three methods. This date in March was expected, when considering that phytoplankton blooms occurred in either March or April in the past. However, this was the first time

that high-resolution vertical profiles covering the wider DyFAMed area were used, offering insights into the biogeochemical and physical processes before, during, and after, the bloom.

4

Physical and Biogeochemical Scales of

Variability in the Northwestern

Mediterranean Sea Using Glider

Measurements

4.1

Summary

Underwater gliders are more advantageous than traditional methods of ocean data collection, as high-resolution spatial and temporal information can be obtained remotely. Physical and biogeochemical spatial characteristics in the northwestern Mediterranean Sea were quantified using semivariograms derived from underwater glider measurements. The horizontal spatial scales obtained using semivariograms highlight the influence of certain processes (e.g. biology, weather) on key physical and biogeochemical parameters. Spatial scales are generally small when affected by photosynthesis (i.e. where chlorophyll patches form), and large when dominated by large-scale processes, such as atmospheric forcing. This was however not universal to every orientation and dataset, and in some cases the

opposite could be seen. There is some anisotropy (i.e. different scales in different

directions), as ranges calculated using zonal distances were larger than those calculated using meridional distances. This anisotropy may be related to ocean currents, or in the case of REP14 - MED, may be related to the limited meridional measurements collected by the gliders. The semivariogram models used to extract the spatial characteristics did not perform as well as in other studies (lower coefficient of determination (r2) values). Obtaining spatial scales in the northwestern Mediterranean Sea is useful for designing

future observational glider campaigns in the region.

4.2

Introduction

High-resolution measurements of physical and biogeochemical parameters were obtained by underwater gliders at two locations in the northwestern Mediterranean Sea: the Sardinian Sea in June 2014, and close to the BOUSSOLE (translated from French as ‘buoy for the acquisition of a long-term optical time series’, (Antoine et al., 2008)) biogeochemical buoy in March - April 2016 off the coast of France (Fig. 4.1). Underwater gliders are relatively cheap, costing a fraction of the cost of a full ship survey or mooring deployment. Gliders are more manoeuvrable, and energy efficient than ship and mooring measurements (Eriksen et al., 2001). Ship surveys normally last just a few weeks or months at a time. Measurements obtained by ship are often higher resolution in the vertical (1 m to 2 m) than in the horizontal (some kilometres), and data distributions are often discrete and not repeated over time. Measurements obtained by moored buoys are isolated spatially, both horizontally and vertically, but temporal resolution is good (Hemsley, 2003). Therefore, measurements obtained by ship or moored buoys cannot resolve all

scales of temporal and spatial variability. Underwater glider spatial and temporal

information can be controlled. High-resolution spatial and temporal measurements can be repeated over periods lasting months at a time (Eriksen et al., 2001; Rudnick, 2016). With an expected increase in underwater glider use in future, quantifying scales of spatial variability is useful in the design of underwater glider observational campaigns that aim to capture important features and processes. Determining the spatial characteristics of water masses in a particular region aid the development of ocean models by establishing the dominant processes determining variability, and improve data assimilation techniques through the determination of observation spatial footprints (Schaeffer et al., 2016).

The physical characteristics of water masses are predominantly driven by density changes, wind forcing, and bathymetry, whilst the biogeochemical characteristics of water masses are further determined by locally acting biological processes, such as photosynthesis and respiration, nutrient cycling, and remineralisation (Schaeffer et al., 2016; Ballantyne IV et al., 2011). The mechanisms driving phytoplankton communities are complex, and their distributions range over many spatial scales of variability (Schaeffer et al., 2016; Martin, 2003). A range of theories exist relating to the physical and biological controls on the patchiness of phytoplankton communities (Steele, 1978; Denman and Abbott, 1988). Processes occurring at the interface between the continental shelf and the open ocean are relevant on small spatial scales, as water masses governed by either oceanic, or coastal processes interact here (Yoder et al., 1987).

4.2 Introduction 73

occur on smaller spatial and temporal scales than in other oceans ( ´Alvarez et al., 2014). The Mediterranean Sea is generally considered oligotrophic when compared globally ( ´Alvarez et al., 2014), however phytoplankton blooms are common in spring and autumn, and a deep chlorophyll maximum (DCM) is often present during the summer when waters are stratified (Estrada, 1996). The northwestern Mediterranean Sea is characterised by Modified Atlantic Water (MAW) typically in the top 140 m of the water column, Levantine Intermediate Water (LIW) typically between 80 m and 730 m, Winter Intermediate Water (WIW) typically between 120 m and 400 m, and Western Mediterranean Deep Water (WMDW) deeper than 535 m. These depth ranges are dependent on space and time (Knoll

et al., 2017). Wind-driven vertical mixing in the top 300 m of the water column is

dominant during the winter and autumn periods, whilst surface stratification is common during the summer (Copin-Mont´egut et al., 2004).

The Sardinian Sea includes both open ocean and continental shelf environments. In June 2014, off-shelf geostrophic transport in the Sardinian Sea was northward. Closer to the shelf edge, it was southward near the surface (MAW), but predominantly northward at depths corresponding to LIW across all areas (Knoll et al., 2017). As a result of strong summer stratification, a DCM was present in the Sardinian Sea at depths of 20 m to 90 m in 2014 (Fig. 2.12).

The BOUSSOLE site is considered open ocean, with a bottom depth > 2000 m.

West/ south-westward geostrophic transport is found here at depths corresponding to

MAW and LIW (Millot, 1999; Niewiadomska et al., 2008). In 2016, wind-driven vertical mixing at the BOUSSOLE site was strong during the first half of March, but weaker during the second half, resulting in an increase in primary production after 19 March (Fig 3.7). In this chapter, physical and biogeochemical spatial characteristics in the northwestern Mediterranean Sea are quantified using semivariograms derived from underwater glider

measurements. The purpose of this is to understand the variability of physical and

biogeochemical properties in this region, relating the observed correlation patterns to underlying processes. The glider data sets, including the sensor payloads, calibrations, and data processing techniques are described in Section 4.3.1, and semivariograms are described in Section 4.3.2. The results, including the spatial characteristics, spatial anisotropy, and the unresolved variances are presented and discussed in Section 4.4. Previous studies investigating spatial characteristics at other regions are discussed in Section 4.5, and the results are discussed within the context of designing observational systems in Section 4.6. Finally, the conclusions of the chapter are presented in Section 4.7.

Figure 4.1: (a) The locations of the BOUSSOLE (green) and REP14 - MED (orange) deployments. (b) A close-up of the BOUSSOLE deployment between 7 March and 5 April 2016. (c) A close-up of the REP14 - MED experiment between 6 and 25 June 2014. The satellite image in panel (a) has been taken from GoogleMaps, and has no corresponding colour scale. GEBCO 1 minute resolution bathymetry data (metres) were used in (b) and (c) (http://www.bodc.ac.uk/projects/international/gebco/). Note the difference in bathymetry scales in (b) and (c). Circulation flows in Modified Atlantic Water (MAW, blue), and Levantine Intermediate Water (LIW, pink) are shown, which are adapted from Millot (1999).