In extremely cold regions, OPGW optical cables are prone to anomalies due to icing, especially when their internal fiber length is easily depleted, which then seriously affects the normal operation of the entire power communication system. Therefore, this paper proposes a nonlinear inversion analysis method that couples catenary theory with biaxial statics to monitor the variation of the FEL under extreme icing conditions. First, the quantitative mapping relationship between BFS and the dynamic evolution of FEL has been established, and a three-dimensional response model, encompassing tension, ice thickness, and FEL has been constructed. Subsequently, an inversion algorithm for ice thickness based on BOTDR frequency shift signals has been developed by introducing horizontal tension as a key intermediate physical quantity. When the FEL decreases from 0.2% to 0.1%, the corresponding EIT can evolve from 0.08 m to 0.16 m approximately, indicating that the system has entered a warning zone characterized by a sharp increase in strain. The proposed model has successfully achieved the transition of distributed monitoring theory from physical perception to quantitative inversion, thereby providing theoretical support for early ice warning and disaster prevention assessments of in-service power optical cables.