Assessment of long-term ecological impact in lower Levisa Watershed: A remote sensing-based approach

This study assesses the long-term ecological impact in the Lower Levisa watershed, Eastern Kentucky, using an entropy-based Remote Sensing Ecological Index (entropy-RSEI) from 1986 to 2022. Eastern Kentucky has a long history of coal mining, especially mountaintop removal mining, which has severely impacted the ecological quality of the region. The reclamation strategy has been employed to tackle the environmental issues caused by these mining activities. Assessing the spatiotemporal evolution of eco-environment quality is crucial for protecting ecological environments and enhancing sustainable development strategies. The entropy weight method was used to integrate greenness (NDVI), wetness (Tasseled cap), dryness (NDBSI), and heat (LST) to develop the entropy-RSEI. The overall ecological quality of the Lower Levisa watershed increased over time. Trend analysis showed improvement in ecological quality in the majority of the study area. 6.79% of the area showed a degradation trend. Future trends are likely to follow the same trend, with improvement still dominant (82.77% of the total area), while 10.01% of the area would reverse to degradation from improvements. The slope was the dominant factor affecting the ecological quality, indicating the higher the slope, the higher the ecological quality. Precipitation had negative impacts, suggesting that higher rainfall would negatively affect eco-environment quality. Other factors such as elevation, temperature, and net primary productivity (NPP) exerted a positive impact in most of the Lower Levisa watershed. These assessments are crucial for effectively assessing ecological environmental quality. In addition, these findings provided the basis for assessing different reclamation strategies for reclamation of mining in this region. • An entropy based remote sensing ecological index (entropy-RSEI) was developed to study ecological changes. • Entropy based RSEI is more effective than PCA based RSEI based on image entropy and contrast. • The ecological quality showed significant recovery with a notable shift towards the excellent class of entropy-RSEI. • CatBoost regression model exhibited higher efficiency in identifying most influencing factors affecting entropy RSEI. • Slope was the most influential positive factor affecting cological quality whereas precipitation had an adverse effect.