Study on the driving mechanisms of spatiotemporal nonstationarity of vegetation dynamics in Heilongjiang Province
Abstract Heilongjiang Province, a key ecological barrier in Northeast China, is crucial for regional ecosystem stability. Previous vegetation index research in this region primarily focused on annual or growing-season scales, without comprehensive comparisons of seasonal and interannual variations....
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-08-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-025-14182-x |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract Heilongjiang Province, a key ecological barrier in Northeast China, is crucial for regional ecosystem stability. Previous vegetation index research in this region primarily focused on annual or growing-season scales, without comprehensive comparisons of seasonal and interannual variations. This study addresses this gap by analyzing spatiotemporal vegetation dynamics and their driving forces in Heilongjiang Province using MODIS data (2000–2021). The findings reveal: (1) Analysis of MODIS-derived Fractional Vegetation Cover (FVC) from 2000 to 2021 revealed decreasing trends in spring, autumn, and winter, alongside an increasing summer trend. Spatially, FVC was higher in the northwest, central, and southeast regions, indicating significant heterogeneity. (2) Theil-Sen trend and Hurst exponent analyses indicated a declining annual FVC trend in 61.8% of the area, with 54.7% projected for continued future decline. A centroid shift model showed an overall westward FVC movement, except in spring. Coefficient of variation analysis demonstrated highest FVC stability in summer and lowest in winter. The global Moran’s I index indicates that FVC exhibits a highly spatially concentrated distribution. Local Moran’s I analysis primarily reveals two clustering patterns: “high-high” and “low-low” aggregations.(3) Random Forest SHAP analysis identified altitude, land cover type, evapotranspiration (ET), and slope as primary factors influencing FVC. Furthermore. The geographical detector analysis demonstrates that the interactions among factors strengthen their overall impact on FVC. |
|---|---|
| ISSN: | 2045-2322 |