Assessing the Ship Motion Prediction Capabilities of the Open-Source Model NEMOH Against Field Observations

Abstract

Accurate ship motion prediction is critical for safe and efficient maritime operations, particularly in open ocean environments. This study evaluates the capability of NEMOH, an open-source potential flow boundary element solver, as an example of a ship motion prediction tool for real-world open ocean conditions. A linear model, known as the Response Amplitude Operator (RAO), is obtained using NEMOH, and is combined with the wave directional spectrum derived from the WaMoS-II marine radar on the research vessel Akademik Tryoshnikov during the Antarctic Circumnavigation Expedition (ACE) to predict ship motion responses in the frequency domain. Predictions of heave, pitch and roll are benchmarked against concurrent ship motion observations recorded by an onboard inertial measurement unit (IMU). The comparisons, based on the zeroth order moment of the ship motion spectrum, demonstrate a reliable heave prediction (Pearson correlation coefficient r=0.89, scatter index SI=0.41), a reasonable pitch prediction (r=0.80, SI=0.47), and an acceptable roll prediction (r=0.63, SI=0.84). More significant discrepancies for pitch and roll are identified under specific extreme sea conditions. The results demonstrate the capability of NEMOH, offering insights into its applicability for real-world maritime operations.