Underwater Wireless Sensor Networks (UWSNs) play an essential role in aquatic environment monitoring, supporting applications such as oceanographic data collection, underwater resource management and disaster prevention. However, accurate localization in underwater remains a significant challenge due to the unique features of underwater environments, including the reliance on acoustic communication, mobility of sensor nodes and the complexity of three-dimensional topology. Traditional localization techniques, like Received Signal Strength Indicator (RSSI) and Angle of Arrival (AOA) methods, suffer from several limitations, including inaccuracies due to time-varying sound speeds affected by salinity, temperature, and pressure. Additionally, they often exhibit high energy consumption, slow convergence, and poor adaptability to dynamic underwater environment. Existing optimization-based localization approaches, face trade-offs between exploration and exploitation, limiting their effectiveness in achieving optimal position estimates. The primary challenge in UWSN localization is achieving high accuracy while minimizing energy consumption and computational complexity. Many existing methods struggle with adaptability in dynamic underwater conditions, where sensor nodes are mobile and environmental factors significantly affect signal propagation. There is a need for an advanced localization approach that can effectively balance accuracy, efficiency, and robustness in complex underwater environments. This paper presents a novel localization approach utilizing the Multi-Verse Optimization (MVO) algorithm, a physics-inspired metaheuristic technique. MVO enhances RSSI and AOA-based localization by maintaining a balance between exploration and exploitation, leading to improved position estimation. Through extensive simulations, we evaluate performance of MVO in terms of localization accuracy, convergence speed, energy efficiency and resilience to anchor node distribution variations. The results demonstrate that MVO significantly outperforms conventional methods by achieving higher localization accuracy while reducing computational overhead. While AOA-based localization is more precise under ideal conditions, RSSI-based methods offer lower complexity, making them suitable for resource-constrained deployments. By overcoming key limitations such as sensitivity to environmental fluctuations and high computational costs, this work establishes MVO as a robust and efficient localization solution for UWSNs operating in challenging underwater environments.