Deep neural networks for active wave breaking classification

Abstract

Wave breaking is an important process for energy dissipation in the open ocean and coastal seas. It drives beach morphodynamics, controls air-sea interactions, determines when ship and offshore structure operations can occur safely, and influences on the retrieval of ocean properties from satellites. Still, wave breaking lacks a proper physical understanding mainly due to scarce observational field data. Consequently, new methods and data are required to improve our current understanding of this process. In this paper we present a novel machine learning method to detect active wave breaking, that is, waves that are actively generating visible bubble entrainment in video imagery data. The present method is based on classical machine learning and deep learning techniques and is made freely available to the community alongside this publication. The results indicate that our best performing model had a balanced classification accuracy score of $$\approx$$ ≈ 90% when classifying active wave breaking in the test dataset. An example of a direct application of the method includes a statistical description of geometrical and kinematic properties of breaking waves. We expect that the present method and the associated dataset will be crucial for future research related to wave breaking in several areas of research, which include but are not limited to: improving operational forecast models, developing risk assessment and coastal management tools, and refining the retrieval of remotely sensed ocean properties.

The Coastal Imaging Research Network (CIRN)

“…In the five years since CIRN's inception, the use of machine learning algorithms to exploit coastal imagery has grown. Data-driven approaches have been used to extract nearshore parameters from imagery including bathymetry [78,79], wave heights [80], wave breaking type [81] and occurrence [82][83][84][85], shoreline position [86], and land cover [87]; to classify nearshore morphology [88]; and to identify dangerous flows, like rip currents [89]. These data-driven approaches can increase analysis capabilities, enabling rapid extraction of data and improving ease of use in engineering and science applications.…”

Remote Sensing Observations of Dominant Breaking Waves in Intermediate to Deep Water from a Lighthouse During Storm Conditions

“…The wave breaking detection technique used in this paper is an extension from 156 the method developed in Stringari et al [22]. These authors developed different neural the vast majority of the pixels but small errors, particularly for small scale breakers, 176 remains.…”

“…To track the the spatio temporal evolution of the waves, a combination of data . Finally, wave breaking crest lengths were calculated using the α-shape algorithm [28] (red lines in Figure 3-f) and ellipses were fitted to each 189 detected event as per Stringari et al [22]. While these data were not analysed in detail 190 here, it allows, for example, to obtain Phillips' Λ(c)dc distribution [29] from which wave 191 breaking probabilities can be derived.…”

Nonlinear wave evolution with data-driven breaking

“…The key ingredient is to include the memory of previous time steps in a recurrent neural network (RNN), allowing non-local representations by which the missing physical processes or “closure terms” can be parameterized. While ML studies have been performed in the context of wave breaking, these have focused on the detection and classification of breaking waves 33 – 35 , rather than their prediction and simulation.…”