Monitoring shrub disturbance in the Qinghai–Tibet Plateau from 1990 to 2022 using the LandTrendr algorithm

Abstract

Background

This study addresses the degradation of shrub ecosystems and emphasizes the essential role that shrubs play within ecological systems. The use of advanced technological methods to swiftly and accurately capture information on shrub disturbance is crucial for preserving ecological security.

Methods

Utilizing the LandTrendr temporal segmentation algorithm on the Google Earth Engine cloud platform, and grounded in land cover data, we conducted dynamic monitoring of shrubland change across the Qinghai–Tibet Plateau from 1990 to 2022.

Results

From 1990 to 2022, the cumulative total area of shrub disturbance in the Qinghai–Tibet Plateau amounted to 372.23 km², primarily concentrated in the eastern and southeastern regions, with an overall decreasing trend observed. The duration of shrub disturbance was predominantly concentrated within a 1–2-year period, covering approximately 80.43% of the total disturbed area. Pixel-scale validation indicated an overall accuracy of 95.71%, with a Kappa coefficient of 0.93. User's accuracy for each year surpassed 73.82% and producer's accuracy was above 70.08%. Shrub disturbance on the Tibetan Plateau is mainly concentrated in areas with an altitude of 2000–4000 m, a slope gradient of 15°−40°, and a shady slope aspect. Shrub disturbance shows a moderately significant negative correlation with temperature (r = −0.436, p < 0.05) and a weakly significant positive correlation with precipitation (r = 0.124, p < 0.05), respectively.

Conclusions

Incorporating contextual data, the study identified climate, and topography as primary factors driving shrub disturbance. This study offers valuable scientific evidence and methodological references for monitoring large-scale shrub dynamics.