Vegetation change of giant panda habitats in Qionglai Mountains through dense Landsat Data

Aims Understanding the processes and drivers of vegetation change in giant panda habitats plays an important role in their conservation and management. Methods Based on the long-term normalized difference vegetation index ( NDVI ) time series that was constru-cted by all available Landsat TM/ETM/OLI images from 1986 to 2018, we employed the BFAST (Breaks For Additive Seasonal and Trend) method and harmonic model to monitor the vegetation change during the period of 1986–2018. Three types of NDVI changes (i.e. vegetation accumulated abrupt change, accumulated gradual change, and total change) were built to reveal the spatial distribution characteristics of vegetation change. The effects of different factors (i.e. mean annual precipitation, mean annual air temperature, elevation, slope, aspect, distance to rivers, soil type, land cover type, distance to roads and distance to engineering disturbance area) on the spatial distribution of the three types of vegetation change were evaluated by Geodetector. accumulated gradual change, and total change were mean annual precipitation, elevation, and soil type, respec-tively. The interactions between driving factors were mutually enhanced and nonlinearly enhanced.