ryan1799In-depth exploration of the host spectral response mechanism is a necessary basis for promoting the rapid monitoring and early warning of Pinus massoniana caterpillar remote sensing. The spectral curve data of 46 masson pine forest canopy curves collected in Changting County and Jianyang District of Nanping City, Fujian Province were set as rule groups, and the one-way analysis of variance was used to select the wavelengths of different hazard levels. The results showed that the canopy spectral data of P. massoniana forest under different hazard levels showed extremely significant differences (P<0.01), among which the spectral discriminability of the canopy of P. massoniana forest with moderate-severe damage was significantly different at 516.51~598.99 nm and 700.68~706.18 nm (P<0.05), and at 708.92~810.62 nm (P). <0.01)。 To this end, the spectral reflectance at the wavelength bands of 519.20 nm, 540.72 nm, 758.40 nm and 785.88 nm was combined, and the spectral spectral data of the canopy of P. massoniana forest in the healthy state were used as standard samples, and the quantitative determination rules of the damage level of pine caterpillars were established based on the spatial distance method, correlation coefficient method and spectral angle mapping method, and the spectral curve data of 34 verification groups in Jiangle County, Yanping District of Nanping City and Hua'an County were used to verify this rule. The results show that the judgment effect of the spatial distance method is much better than that of the correlation coefficient method and the AND spectral angle mapping method. The spatial distance determination rules for pine caterpillar hazards, mild hazards, moderate hazards and severe hazards were as follows: <0.3553, [0.3553, 0.7425], [0.7425, 0.9631) and ≥0.9631, with a judgment accuracy of 88.24% and an accuracy of 97.06%.