Abstract: To address the limitations in parameter dimensionality, the single damage evaluation metric, and the lack of generalizable quantitative prediction models in the lightning protection research of carbon fiber-reinforced polymer (CFRP) composites, this study employs thermo-electrical coupled finite element analysis to investigate multi-parameter quantitative correlations within CFRP protective layers. First, for the widely used aluminum coating protection system, the evolution of lightning-induced damage area and depth in CFRP under various lightning current parameters and coating thicknesses is elucidated. The results show that increasing the coating thickness significantly suppresses lightning damage: when the thickness reaches 0.25 mm, damage is completely mitigated under low peak currents (20 kA), and under high peak currents (100 kA), the damage area and depth are reduced by approximately 90.7% and 86.4%, respectively. Correspondingly, a Power-Law model for damage area and a Gaussian process regression (GPR) model for damage depth are developed, enabling quantitative characterization of CFRP lightning damage with aluminum coating protection. Subsequently, for protective materials with arbitrary thermo-electrical properties, the quantitative coupling relationships between material parameters and damage responses under the most severe lightning waveform are investigated. Based on parameter sensitivity analysis, generalized damage prediction models are constructed. The results indicate that electrical conductivity, volumetric heat capacity, and layer thickness are the key parameters governing protection performance. The logarithm-corrected power-law model and GPR model incorporating these parameters exhibit excellent predictive capability for both damage area and depth, achieving average R² values of 0.986 and 0.924, respectively. These findings provide theoretical foundations and quantitative guidance for the structural design and performance optimization of CFRP lightning protection layers.