Download files
Abstract
This project aimed to study dynamical forcing for seasonal variability of the East Australian Current (EAC) and mode water in the East coast of Australia using numerical models. The current thesis also reports on an attempt to investigate the impact of initial data errors in model initialization and analysis of the role of the bottom boundary layer in the evolution of Joint Effect of Baroclinicity and Relief (JEBAR) using Bluelink Reanalysis (BRAN) and Ocean Forecasting for Earth Simulator (OFES) datasets. The Princeton Ocean Model (POM) has been used to investigate impact of the bottom boundary layer on evolution of JEBAR and model initialization errors on prognostic simulations. The vorticity has been used as a tool to identify the contribution of various components driving seasonal variability of the EAC. The results obtained from the BRAN and the OFES data suggest the seasonal variation of local baroclinic forcing contributes significantly in comparison to wind forcing. Furthermore, the difference in JEBAR evolution in the case of BRAN and the OFES strongly correlates with the variation in the seasonal cycle of the EAC. Using the POM model, the role of the bottom boundary layer (BBL) has been investigated. The development of BBL considerably improves the simulation of the ocean transport. However, the impact of the BBL on evolution of JEBAR remains inconclusive. The seasonal variability of the subtropical mode water (STMW) has been investigated using the BRAN data. The model simulations show the highest (lowest) STMW volume in the winter (summer) months. Further investigations of inter-annual variability of the STMW reveal that the upper ocean temperature in the summer months provides essential preconditioning for the production of the STMW. The study also demonstrates that an anomalous increase in the EAC provides a less favorable precondition for the STMW volume. The impact of interpolation errors on the model initialization has been investigated using the POM model. It has been observed that noise due to interpolation of the initial data over the model grid may contribute up to 80% in the steady state obtained by diagnostic initialization. The noise introduced by diagnostic initialization has been analyzed using thermo-haline source/sinks. The study reveals an imaginary mean state of thermo-haline source/sinks that enables robust model initialization. Deviation from the mean state induces abnormally strong initial velocities by diagnostic initialization.