Abstract: To address the problem of localization accuracy degradation in underwater acoustic sensor networks caused by sensor motion effects and sound speed stratification effects, a joint time delay and Doppler shift localization method is proposed to enhance the adaptability in complex environments. The method establishes a joint measurement model of time delay and Doppler frequency shift under motion and stratification effects, constructs a maximum likelihood estimation objective function, and solves the target position using the Gauss-Newton iterative method. To ensure iterative convergence, a simplified model based on the assumption of straight-line acoustic propagation is developed, with initial solutions obtained by the least squares method serving as iterative starting points. Simulation results demonstrate that neglecting motion and stratification effects leads to significant increases in localization errors. The proposed method effectively improves localization accuracy by considering both effects in the measurement model. This method enables simultaneous compensation for motion and stratification effects. Through a strategy of layered modeling-joint estimation-optimization iteration, it achieves high-precision localization, providing an effective solution for underwater target localization.