The presence of spiral waves in cardiac tissue causes ventricular fibrillation (VF). Clinical studies have demonstrated that reducing the blood temperature to hypothermia can improve the success rate of electrical defibrillation when VF occurs. In this paper, the Hodgkin-Huxley-style model is used to simulate the effects of temperature and electric shock on the spiral waves present in mammalian tissues, and the phase singularity (PS) identification methods are used to track the wave behaviors in network, finally the Luo-Rudy model is used to test the applicability of our results for cardiac tissues. The results show that temperature modulates the optimal parameters for defibrillation: (1) Arnold's tongue structure shows that an alternating electric field close to the tissue natural frequency (e.g.: cardiac sinus rhythm) allows the nodes to be phase-locked or PS directionally drift at a smaller intensity, thus eliminating spiral waves more effectively than a constant electric field. (2) The average node frequency increases with temperature whether the network is synchronized or present spiral waves, and the statistics show that the robustness of electric defibrillation can be enhanced when the tissue is at moderate hypothermia. (3) More interestingly, we define the defibrillation score (DFS) and find that moderate hypothermia is a balance to improves defibrillation success, which may provide some insights for practical defibrillation studies.