In this article, we focus on the event-triggered observer-based fuzzy positioning control and $L_{2}$-gain analysis of nonlinear unmanned marine vehicle (UMV) systems in network environments, which is subject to a hybrid attack consisting of denial-of-service attack and deception attack. First, based on the Tagaki-Sugeno fuzzy modeling method, a unified switched system model is established for the positioning control of nonlinear UMV systems with actuator saturation, multiplicative gain uncertainties, external disturbance, and hybrid attacks. Then, by resorting to introducing the integral-type state error, a novel adaptive memory-based event-triggered mechanism is proposed to lighten the communication load in UMV systems. Furthermore, a sufficient criterion is proposed to make the mean square exponential stability of UMV systems with a nonweighted $L_{2}$-gain level be guaranteed, and gain parameters of the designed controller and observer are also solved efficiently. Finally, the simulation example demonstrates the effectiveness of the proposed nonfragile control strategy.