Changing Urban Forest Cover and Climate Variability in Bamenda, Cameroon: A Spatio-temporal Assessment
Tita Walters Ade *
Department of Geography and Planning, The University of Bamenda, Bambili, Cameroon.
Suiven John Paul Tume
Department of Geography and Planning, The University of Bamenda, Bambili, Cameroon.
*Author to whom correspondence should be addressed.
Abstract
This paper examines how urban forest change and climate variability jointly shape hazard exposure in Bamenda. Using a mixed-methods design, it combines Landsat classification (1984, 2004, 2025), a 75-year rainfall series (1951-2025), a temperature window (1984-2023), a 400-household survey, key-informant interviews and field observation, organised within an analytical framework where regional climate provides the trigger, urban forest and built form modify exposure, and household capacity conditions sensitivity. City-wide forest fell from 11,170.8 ha in 1984 to 10,856.3 ha in 2025 (-2.8%), but municipal trajectories diverged: Bamenda I lost 38.8%, and Bamenda III lost 22.3%, while Bamenda II gained 51.6%. Mean monthly rainfall intensity declined from 213.4 mm (1951-1970) to 139.4 mm (2011-2025), with a highly significant trend of -1.17 mm yr⁻¹ (R²=0.581; p=2.04×10⁻¹⁵). Mean temperature rose from 18.56°C (1984-1990) to 22.97°C (2013-2023). Standardized Precipitation Index (SPI) shifted from wet extremes in the 1950s-1960s to severe-to-extreme drought after 2007, including 2016 (SPI=-2.49) and a six-year severe-drought run from 2020 to 2025. Visual analytics, including graphs, heat maps, box plots, scatter plots, correlation matrices and spider charts, make co-movements visible. Flooding in Mulang, Ndamukong, Ngomgham and Ntarinkon and slope instability in Abangoh, Sisia and Mbatu are interpreted as joint outcomes of a drier, more variable climate and fragmented forest. Forest loss is thus not mere vegetation loss but a component of the climate-hazard-vulnerability nexus, amplifying flood, erosion, landslide and heat risks when combined with impervious-surface growth and settlement on floodplains and steep slopes. The following recommendations are drawn: reconstructing a quality-controlled monthly archive for Walsh-Lawler SI, SPI-3/6 and onset; adopting 2008, 2014, 2016 as design drought years while retaining flood/slope preparedness for wet extremes like 2013; sizing storage for multi-year drought; governing forest as hazard infrastructure, enforcing floodplain/slope regulation with a unified risk map and embedding annual SPI heat-maps and forest-stock visuals in climate communication.
Keywords: Urban forest, climate variability, land-cover change, Standardised Precipitation Index, rainfall intensity, remote sensing, drought, urban flooding, landslide risk