Chaos Theory and Strange Attractors: Implications for Weather Pattern Analysis in the Guinea Savanna Zone

This study explores the application of chaos theory and the concept of strange attractors to the analysis of weather pattern variability in the Guinea Savanna ecological zone of Nigeria, encompassing parts of Kwara, Niger, Kogi, Nassarawa, and Plateau states in the North Central geopolitical zone. The Guinea Savanna zone is characterised by a strongly bimodal rainfall regime and significant interannual variability in precipitation onset and cessation, patterns that are of critical importance for rain-fed agriculture, which remains the primary livelihood of the majority of rural households in the zone. The study reconstructs phase space trajectories from historical daily rainfall, temperature, and humidity data sourced from the Nigerian Meteorological Agency for a 30-year period from 1993 to 2023, applying Takens' delay embedding theorem to convert univariate time series into multi-dimensional phase portraits. Lyapunov exponent estimation is conducted to quantify the degree of chaotic sensitivity present in the reconstructed trajectories, and the correlation dimension algorithm is applied to characterise the fractal structure of the resulting attractors. Results indicate the presence of low-dimensional chaos in seasonal rainfall onset timing, with positive Lyapunov exponents confirming sensitive dependence on initial conditions in the 15-to-45-day forecasting horizon. The study discusses implications for medium-range agricultural weather advisories. Keywords: chaos theory, strange attractors, Lyapunov exponents, Guinea Savanna, rainfall variability