Detecting Vegetation Variations and Main Drivers over the Agropastoral Ecotone of Northern China through the Ensemble Empirical Mode Decomposition Method

Yayong Xue,Baoqing Zhang,Chansheng He,Rui Shao

doi:10.3390/rs11161860

Abstract

Vegetation is the major component of the terrestrial ecosystem. Understanding both climate change and anthropogenically induced vegetation variation is essential for ecosystem management. In this study, we used an ensemble empirical mode decomposition (EEMD) method and a linear regression model to investigate spatiotemporal variations in the normalized difference vegetation index (NDVI) over the agropastoral ecotone of northern China (APENC) during the 1982–2015 period. A quantitative approach was proposed based on the residual trend (RESTREND) method to distinguish the effects of climatic (i.e., temperature (TEM), precipitation (PRE), total downward solar radiation (RAD), and near surface wind speed (SWS)) and anthropogenic effects on vegetation variations. The results showed that the NDVI exhibited a significant greening trend of 0.002 year−1 over the entire study period of 1982–2015 and that areas with monotonous greening dominated the entire APENC, occupying 40.97% of the region. A browning trend was also found in the central and northern parts of the APENC. PRE presented the highest spatial correlation with the NDVI and climate factors, suggesting that PRE was the most important factor affecting NDVI changes in the study area. In addition, the RESTREND results indicated that anthropogenic contributions dominated the vegetation variations in the APENC. Therefore, reusing farmland for grass and tree planting made a positive contribution to vegetation restoration, while deforestation, overgrazing, and the reclamation of grasslands were the opposite. In addition, with the continuous implementation of national ecological engineering programs such as the Grain to Green Program, positive human activity contributions to vegetation greening significantly increased. These results will support decision- and policy-making in the assessment and rehabilitation of ecosystems in the study region.

Highlights

Agropastoral ecotones refer to transition zones interlaced by croplands and grasslands [1], and they are widely distributed in the arid and semiarid regions of the world [2]
The results suggested that anthropogenic contributions dominated the vegetation variations, which was consistent with the results of Liu et al [8]
We used the ensemble empirical mode decomposition (EEMD) method to assess the spatiotemporal dynamics of vegetation variation in the agropastoral ecotone of northern China during 1982–2015

Summary

Introduction

Agropastoral ecotones (hereafter APEs) refer to transition zones interlaced by croplands and grasslands [1], and they are widely distributed in the arid and semiarid regions of the world [2]. The arid and semiarid agropastoral ecotone of northern China (hereafter APENC) is a typical ecologically transitional zone (largest in area and the longest in spatial scale), which effectively prevents the desert in the northwest from spreading south and east [4]. Vegetation serves as a sensitive indicator of an ecosystem’s status and could reflect regional ecological security [7,8]. Systematic monitoring of vegetation dynamics and quantitative analyses of factors affecting vegetation variations at spatiotemporal scales are essential to understand the sustainability of terrestrial ecosystems, especially in areas with fragile ecosystems and high sensitivity to external environmental changes

Methods

Results

Discussion

Conclusion