Using Time-Series Analysis to Assess the Extent of Climate Variability and Climate Change in Bayelsa State, Nigeria

Abstract

Time series analysis is a useful statistical tool in the assessment of climate variability and climate change. This study applied a time series analysis to rainfall and temperature data in the Bayelsa State of Nigeria. Since NIMET has only one gauging station in the state, Climate Research (CRU 0.5× 0.5) gridded data for 28 locations from 1956 to 2016 were used. They were sorted, validated with NiMet data, and utilized for analyses of various time series techniques such as Mann-Kendal, Spearman’s Rho, Linear Regression, Thei-Sen Slope Cumulative sum, Cumulative Deviation, Rank Sum, Student’s (t-test) and spectral analysis. The results obtained revealed that there had been increasing temperature and abrupt climatic changes in the state, especially in the 1976-1985 decade, with 1980 as the most probable year of abrupt change. The hottest decade was 1986-1995, with an average temperature change of 0.14856 ^oC/decade, while the coolest decade was 1976-1985 with an average Temperature change of -0.01723 ^oC/decade. Also, there had been some changes in rainfall, with the wettest decade occurring in 1986-1995 with an average rainfall change of 61mm/decade, while the driest decade occurred in 1976-1985 with an average rainfall change of 14.08 mm/decade. The output of spectral analysis showed that the most Significant Periodicity for Rainfall and Temperature was 15 years. The result further revealed that there was high rainfall variability with a coefficient of variability of 62.74%. These rainfall fluctuations have implications for coastal flooding, quality, and quantity of available groundwater in the state. These results are useful to planners and policymakers in creating awareness of climate change's impact on rainfall in the study area