Seasonality and Fractional Integration in the Sea Level Rise and Surface Temperature Data along the Barrier Coast of Nigeria

This paper examines the time series properties of sea level rise and the surface temperature data along the Barrier Coast of Nigeria. In particular, we focus on the seasonality and the degree of persistence of the series, measured in terms of seasonal and non-seasonal unit roots along with fractionally integrated models. Our results show that the two series are fractionally integrated, with orders of integration of about 0.34 for the sea level rise and about 0.76 for the temperatures. Due to the disparity in the degree of integration of the two series, they do not cointegrate. However, assuming that temperatures are weakly exogenous for the sea level rise, our results indicate that they are positively and significantly correlated in a long memory relationship. Further, while time-domain Granger causality test show that sea surface temperature leads sea-level rise, with frequency domain causality test suggesting that causality is primarily at longer horizons from 18 months and ahead. In other words, it takes year and half for changes in sea-surface temperatures to affect sea-level rise.