Detecting and Investigating Frontal Dynamics in the Skin Layer of the North Sea Using Autonomous Surface Vehicle Observations

The shallow North Sea is characterized by complex hydrography and environmental variability, which affects the air-sea interactions and the skin layer (< 1mm). The skin layer plays a crucial role in air-sea interactions, and understanding its physical dynamics is essential for advancing knowledge in this field. Ocean fronts are predominantly narrow horizontal gradients of oceanic properties separating water masses and can be hotspots for marine biodiversity. Satellites have been used to observe large scale fronts however frontal features exist even on the sub-mesoscale. By leveraging seasonal data from an autonomous surface vehicle, we investigate the spatial and temporal variability of the skin layer at these small scales, with a focus on the role of oceanic fronts in shaping surface-layer dynamics. We use high-resolution measurements (0.1 Hz) of temperature, salinity, GPS data and weather-related variables at discrete depths in the air and water. These measurements are from our self-designed autonomous surface vehicle known as HALOBATES, which is equipped with multiple sensors and rotating glass discs to sample and measure the skin layer. Primarily, we applied a gradient-based algorithm to identify oceanic fronts, characterized by sharp horizontal gradients in temperature and salinity within the skin layer and bulk water (1-meter depth). To ensure that frontal features are accurately isolated, the effect of diurnal warming or cooling were taken into account. We isolated fronts solely on temperature or salinity, and by both parameters to understand the nature and origin of the fronts. Secondarily, we provide results of front detection from a Machine Learning perspective, showing results of deep learning models like Convolutional Neural Networks and an unsupervised clustering algorithm. We also present the results of the underlying bulk water (1-meter depth) to understand the pre-eminence of the observed front. We also investigate possible forcing factors due to sudden changes in the wind speed and direction as well as heat flux components such as solar radiation, precipitation and evaporation. By using continuous data from an autonomous platform over an extended period, our findings will highlight the dynamics of sub-mesoscale fronts in the North Sea and their role in shaping the physical characteristics of the skin layer. The insights from this study are relevant for ongoing research on topics such as heat and gas exchange at the ocean-atmosphere boundary layer which is a good indicator of the climate dynamics of the North Sea.