Modeling climate change impacts and predicting future vulnerability in the Mount Kenya forest ecosystem using remote...

Area of Science:

• Ecology
• Climate Science
• Remote Sensing

Background:

• Mount Kenya forest ecosystem (MKFE) is a vital biodiversity hotspot and water tower.
• Climate change, characterized by rising temperatures and erratic rainfall, poses a significant threat to MKFE's ecological integrity and services.
• Altered vegetation patterns, reduced resilience, and threats to biodiversity and water security necessitate urgent conservation planning.

Purpose of the Study:

• To assess historical vegetation changes in MKFE using remote sensing data.
• To predict future vegetation dynamics and identify areas at risk under climate change scenarios.
• To understand the relationship between climate extremes and forest health.

Main Methods:

• Utilized Landsat imagery (2000-2020) to derive vegetation indices (NDVI, EVI, SAVI, BSI).
• Extracted climate variables (extreme precipitation, temperature) from CHIRPS and ERA5 datasets.
• Employed machine learning models (Random Forest, XGBoost, SVM) to analyze climate-vegetation relationships and project future changes using CMIP6 data under SSP245.

Main Results:

• The Random Forest model demonstrated high accuracy (R²=0.82, RMSE=0.15) in predicting vegetation dynamics.
• Projections indicate a 49-55% decline in Enhanced Vegetation Index (EVI) across forest areas by 2040.
• Lower montane zones are identified as most vulnerable to climate-induced vegetation stress, with precipitation being a critical factor for forest health.

Conclusions:

• Urgent adaptive management strategies are required, including afforestation and sustainable land-use planning.
• Policy-driven conservation efforts are crucial for mitigating climate change impacts on MKFE.
• The study offers a scalable framework for global forest ecosystem impact modeling and provides actionable insights for policymakers.