Estimation of seasonal representation of the seawater temperature profile using machine learning and its effect on the prediction of underwater acoustic detection performance

Abstract

Seawater temperature and salinity profiles are important physical properties that represent oceanic environments and affect underwater acoustic detection prediction performance. In an environment in which it is difficult to obtain real-time ocean data or predict the SONAR detection area immediately, the SONAR detection area can be predicted from the average ocean data. However, it can yield distorted results. In this study, the K-means clustering, an unsupervised machine learning technique, was applied to obtain representative temperature profiles reflecting the characteristics of the vertical structure at various temperatures in the studying area. The K-means clustering was applied to the seawater temperature profiles obtained from the three stations of the Ulleung Basin in the East Sea, where the interannual variations are large, and clustering was performed. Additionally, the physical characteristics of the representative profiles obtained were compared, and the representativeness of the acoustic detection area obtained from the representative profiles could represent the acoustic detection area was evaluated. In summer when the thickness of the mixed layer was thin, each cluster was classified according to the vertical temperature gradient of the thermocline. In winter, on the other hand, the clusters were classified according to the mixed layer depths and thermocline depths rather than the vertical temperature gradient of the thermocline. In addition, for each obtained cluster, the acoustic detection area was calculated using all the profiles and displayed as a histogram. The acoustic detection area calculated from the representative profile of the cluster was found to be generally close to the average of the acoustic detection area. In conclusion, K-means clustering effectively classified temperature profiles physically and acoustically. It is expected that it can be applied in various fields to classify and analyze the characteristics of seawater temperature and salinity profiles in the future.