Trend Analysis and Time-Series Forecasting of Climate Variability in Urban and Rural Settlements of Niger State, Nigeria
Abstract
Climate variability has become a major environmental concern due to its implications for water resources, agriculture, ecosystem sustainability, and human livelihoods. This study analysed historical climate trends and forecasted future climate variability in selected urban and rural settlements of Niger State, Nigeria, using statistical and time-series techniques. Annual rainfall and temperature data covering the period 1994–2024 were analysed using the Mann–Kendall trend test and Sen's Slope estimator, while future climate conditions were projected using Autoregressive Integrated Moving Average (ARIMA) models. The results revealed significant warming trends across the study area. Temperature increased by approximately 16.0°C in Mokwa and Muwo, 18.2°C in Minna and Garatu, 17.1°C in Kontagora, and 2.9°C in Rafin Gora between 1994 and 2024. Rainfall exhibited spatial variability, with substantial declines in some settlements and slight increases in others. Mann–Kendall and Sen's Slope analyses confirmed increasing temperature trends and varying rainfall patterns. The ARIMA model identification process indicated that ARIMA(1,2,1) was suitable for rainfall forecasting, while ARIMA(1,1,1) adequately modelled temperature variations. Diagnostic statistics confirmed the adequacy of the selected models, with non-significant Ljung–Box statistics indicating white-noise residuals. Forecasts suggest continued temperature increases and increasing climate variability across both urban and rural settlements up to 2054. The study concludes that climate variability in Niger State is intensifying and may pose significant environmental and socio-economic challenges if adaptation measures are not implemented. The study recommends the integration of climate forecasting into regional planning, strengthening of climate monitoring systems, and implementation of adaptation strategies to enhance resilience.