Analysis of Spatiotemporal Variations and Driving Mechanisms of Ecological Environment Quality in Yantai City Based on the XGBoost-SHAP Model

DOI: 10.22161/ijeab.113.3

Abstract:
Conducting long-term spatiotemporal monitoring of ecological environment quality and identifying its driving mechanisms are of great significance for coordinating ecological protection with high-quality regional economic development. Based on Landsat remote sensing data from 2005 to 2025 on the Google Earth Engine (GEE) platform, this study employed Principal Component Analysis (PCA) to construct the Remote Sensing Ecological Index (RSEI), analyzing the spatiotemporal evolution patterns of RSEI in Yantai City. The study further utilized the explainable extreme gradient boosting model (XGBoost-SHAP) to reveal key natural and anthropogenic drivers of ecological environment quality, along with their nonlinear and interactive effects. The results indicate that: (1) On a temporal scale, the RSEI of Yantai City exhibited a fluctuating trend of initial decline followed by an increase, then another decline and subsequent rise, with a multi-year average of 0.438, indicating an overall moderate level of ecological environment quality; (2) Spatially, the RSEI in the study area showed significant heterogeneity, with high-value areas concentrated in central Yantai and low-value areas distributed near the coast; (3) Ecological quality was primarily driven by land use type and precipitation, with anthropogenic factors also playing a significant role. This research provides methodological references and scientific foundations for ecological monitoring, risk assessment, and territorial spatial optimization in coastal cities.

Keywords:
Remote Sensing Ecological Index (RSEI), XGBoost-SHAP Model, Google Earth Engine (GEE), Principal Component Analysis (PCA), Ecological Environment Quality (EEQ), Yantai City

Article Info:
Received: 30 Mar 2026; Received in revised form: 27 Apr 2026; Accepted: 03 May 2026; Available online: 13 May 2026

Cite this Article: