Objective  Based on remote sensing data, the characteristics of canopy spectral changes in different succession stages of broadleaved Korean pine forest in Changbai Mountain of northeastern China were analyzed to provide theoretical basis for revealing the interspecies change and the response mechanism of vegetation productivity to climate factors in Changbai Mountain.

  Method  Through the Google Earth Engine platform, Landsat and Sentinel series of remote sensing images were used to extract multi-temporal canopy spectrum data for the broadleaved Korean pine forest (primary forest) and birch-aspen forest (secondary forest), both were in a same succession series in Changbai Mountain. Also we analyzed the seasonal variations of the canopy spectrum characteristics of the two, the seasonal and inter-annual variation of vegetation greenness, and calculated the Pearson correlation coefficient between the inter-annual vegetation greenness variation and the monthly average temperature of the same period from 1985 to 2019.

  Result  (1) For canopy spectral reflectance of the primary forest, the visible light was higher in leaf-off season than in growing season, while the near-infrared reflectance showed an opposite pattern. In the vigorous growth season (from the end of May to the end of October), the canopy reflectivity of the primary forest and the secondary forest was similar in the visible light band, but the near-infrared band was significantly different, and the secondary forest canopy reflectivity was higher. The phenomenon of “red valley”, “green peak”, “blue valley” and “red edge”, the curve form of spectral reflectance in the two vegetation types were evident, and the interannual fluctuation was weaker than that of the secondary forest. (2) The greenness of primary forest and secondary forest showed the same changing trend. It exhibited growth during leaf development in spring and attenuation during leaf fall in autumn. In the non-growing season, the degree of change in vegetation index of the primary forest was relatively stable and greater than that of the secondary forest, indicating that the understory of the secondary forest had high light transmittance. In vigorous growing season, the EVI and S2REP of the secondary forest were larger than those of the original forest, and the physiological activities of the vegetation canopy were more vigorous. Different satellite image data showed consistent performance, and the EVI peak of the secondary forest appeared slightly earlier than the original forest. (3) During the 35-year period from 1985 to 2019, the temperature in the study region had been on the rise, resulting in the increase in both vegetation greenness and the length of growing season; EVI of the primary forest was increasing, with the rate greater in summer than in other seasons. The interannual difference between spring and autumn for enhanced vegetation index was significant. (4) Compared with the primary forest, the interannual variation in EVI of the secondary forest was more correlated with spring temperature. At the beginning of growing season, both forests presented the same pattern that EVI and temperature were positively correlated. During the entire growing season, EVI increased steadily prior to the period when temperature reached a high level.

  Conclusion  Long-term continuous monitoring and analysis of canopy spectrum changes can effectively reflect the difference in vegetation phenology between the primary forest and the secondary forest. Temperature rise may be one of the important factors causing the greenness of the broadleaved Korean pine forest in Changbai Mountain of northeastern China.