The physiological and biochemical parameters of forest vegetation change over time, leading to alterations in their spectral characteristics and resulting in the phenomenon of "spectral variability of the same object". To analyze the temporal spectral variation patterns of forest vegetation, this study simulates the temporal spectra of canopy of typical tree species during typical phenological periods of the year based on the PROSAIL model. The Height of a Single Triangle (HST) is constructed, and the Heights of Double Triangles (HDT) is proposed. They are applied to Sentinel-2 A/B multispectral remote sensing images and the causes of temporal spectral differences are explored. Taking Langya Mountain and Laojiashan in Anhui Province, as well as Tieshansi Forest Farm in Jiangsu Province as examples, the study analyzes and extracts the temporal spectral variation patterns and characteristics of forest vegetation. The research results indicate that the HST index can represent the temporal variation characteristics of forest vegetation: it increases rapidly from the bud burst stage (March) to the leaf expansion stage (May) and decreases slowly from the peak stage (June) to the browning stage (November). The HDT feature space can represent the distribution pattern of forest vegetation samples forming an obtuse-angled triangle shape over time, providing a new methodological and technical approach for extracting the temporal spectral variation patterns of forest vegetation.