Climate change estimation and prediction over eastern Afromontane biodiversity hotspot areas

Abstract

The Eastern Afromontane Biodiversity Hotspot Area (EABHA) is among the world’s most biologically rich regions. Yet these highland systems are increasingly threatened by climate change. In this study, we assess long-term climate variability and project future trends to 2050, aligning our analysis with the United Nations Sustainable Development Goals (SDGs). Using Principal Component Analysis (PCA), Empirical Orthogonal Function (EOF), and an artificial neural network model (ANNM), we examined precipitation, temperature, rainy-day frequency, and wind dynamics from satellite observations, in-situ measurements, and reanalysis climate models. The ANNM was trained on detrended EOF components and station data (1981–2022) and validated using a data split into training (80%), validation (10%), and test (10%) sets, ensuring representation across regions and time periods. Results reveal a marked decline in the number of rainy days over the past four decades, a trend expected to persist through 2050. The annual mean temperature has increased and is projected to rise by ~0.040 K yr -1 for 2023-2050 relative to the 1981–2022 baseline (0.030 K yr - ). Reductions in the number of rainy days are the primary drivers of drought events in ecosystems located above 2,800 m a.s.l. in the study area. These findings advance SDG 13 (Climate Action) by quantifying climate risks, SDG 15 (Life on Land) by identifying threats to biodiversity, and SDG 6 (Clean Water and Sanitation) by highlighting freshwater security challenges. By demonstrating that drought risk is more closely linked to declining rain-day frequency than annual rainfall totals, this research provides actionable insights for adaptation, conservation, and resilience planning across the EABHA. • Climate data patterns were examined via EOF and PCA for space-time variability. • High-altitude areas (>2800 m a.s.l.) face faster climate change impacts than lower areas. • Eastern Afromontane drought is linked to rainy days, not total annual precipitation.