Mineral suspended solids are a strong source of ackscattered light within coastal shelf seas. Backscattered light has a direct relationship to remotely-sensed reflectance, measured by satellites. For four decades, remote-sensing algorithms have been used to estimate the concentration of suspended particulate matter (SPM). Present algorithms for SPMassume that the mass-specific backscattering coefficient (b⇤bp(667nm) = bbp(667nm)/SPM) is constant. Chapter III aimed to explore how observations of the influence of particle compositon upon backscattering could be incorporated into remote-sensing algorithms to improve our representation of b⇤bp(667nm). Field observations showed that bbp ⇤ (667nm) varies with the size, cross-sectional surface area and mineral content of particles, in addition to the total concentration of suspended solids. Observations showed that the increase in concentration related to an increase in the mineral content of the particles, presumed to be causing the relationship between concentration and b⇤bp(667nm). b⇤bp was shown to vary by a factor of 4, therefore using the latter relationship, iterative estimation of b⇤bp(667nm) suggested that it could be possible to reduce the assumption made in future algorithms for the remotely-sensed SPM, by a similar factor. Using the ocean colour satellite archive of remotely-sensed SPM concentration from MODIS (Moderate Resolution Imaging Spectroradiometer) on the Aqua satellite, it is possible to produce statistical models based upon dominant forcings of tides, wind-driven waves and chlorophyll-a. Chlorophyll-a is used as an estimation of phytoplankton biomass, with increases in concentration expected to relate to increased particle size and thus increases in the particle settling speed (ws) of suspended sediments. These have the potential to predict the concentration when atmospheric correction fails. Atmospheric correction failure is most commonly due to the presence of cloud (in addition to cloud shadowing, land, high concentrations of coccoliths or coastal adjacency), which lead to a lack of data in a given pixel (Ardanuy et al., 1991). Rivier et al. (2012) used average observed values of SPM concentration, a tidal coefficient, the significant wave height and chlorophyll-a concentration to predict the SPM concentration in the English Channel. Chapter IV aimed to improve upon Rivier et al. (2012), looking at a wider subject area and improving representation of tides and ws. Tidal inclusion was improved through the use of localised model predictions of tide for every pixel in comparison to a standardised single-location tidal coefficient based upon the tidal amplitude in Brest for the entire English Channel . This study expanded the region to the entire European shelf, included data from a global tidal model and used higher frequency wind speed observations to improve upon the results of Rivier et al. (2012). Furthermore, a crucial shortcoming of Rivier et al. (2012) is the use of single yearlong regression coefficients. The coefficients in this study were allowed to vary seasonally with a sinusoidal form, where the influence of forcings upon the subsequent concentrations varies seasonally. A sinusoidal coefficient was used upon the forcings to represent the e↵ect of flocculation and the influence upon ws. These improvements demonstrate a substantial improvement upon the framework of the present statistical model for suspended particulate matter concentration developed by Rivier et al. (2012). Chapter V aimed to test the influence of ws upon SPM concentration and how well a simple model could be used to represent remotely-sensed observations of SPM. To examine the influence of ws upon the SPM concentration, a simple numerical turbulent kinetic energy (TKE) based resuspension model was used to predict the surface concentration on the northwest European shelf (Elliot & Clarke, 1991; Bowers, 2003). Using a seasonally-imposed ws, the model was shown to perform well when compared to satellite observations. The model identified the importance of ws on the influence of the hydrodynamic forcings of the wind and tides. Bowers (2003) predicted that spring-neap variation in SPM would vary seasonally due to changes in ws. Chapter VI aimed to test this assumption using the remote-sensing archive and examine whether it was possible to quantify ws using the spring-neap variability. Spring-neap variation in the remotely-sensed SPM concentration was shown to be modulated seasonally. In the winter, the range of concentration observed was greater than in the summer. Following results from the numerical model, this variability was proposed to be due to seasonal changes in ws. Quantifying this variation, the bulk ws was estimated through remote-sensing using a novel method. Remotely-sensed observations of the range of concentrations observed over the spring-neap tidal cycle were related to the seasonal variability in ws. Estimates of bulk ws were comparable in magnitude to field observations. ws was shown to peak in the summer, with some locations experiencing additional peaks in the spring and autumn, corresponding with localised blooms. This was proposed to be due to increased phytoplankton productivity, which leads to increases in EPS (extracellular polymeric substances), promoting flocculation and therefore increasing the speed of particle settling.