Abstract
<title>Abstract</title> <p> <bold>Background</bold> : Urbanization is a transformative global trend, serving as a primary driver of economic and social change. In Somalia, rapid urban expansion—driven by environmental pressures, migration, and demographic shifts—presents significant challenges for sustainable development and infrastructure planning. Accurate population projections are essential for informed policymaking and resource allocation. <bold>Objectives</bold> : This study aimed to model the historical trajectory of Somalia's urban population from 1960 to 2024 and compare the performance of various time-series models to project growth through 2050. <bold>Methods</bold> : Using longitudinal data from 1960 to 2024, the study employed a comparative analytical framework splitting data into training and testing sets (sample out = 26). Six forecasting models were evaluated: Autoregressive Integrated Moving Average (ARIMA), Exponential Smoothing (ETS), BATS, Trigonometric Box-Cox ARMA Trend Seasonal (TBATS), Theta, and Autoregressive Fractionally Integrated Moving Average (ARFIMA). Model adequacy was verified via Ljung-Box tests, and performance was measured using Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and symmetric Mean Absolute Percentage Error (sMAPE). <bold>Results</bold> : Descriptive analysis revealed that Somalia's urban population grew from 508,138 in 1960 to 10,382,351 in 2024, reflecting a significant upward trend with a skewness of 1.23. Diagnostic tests indicated that the series was non-stationary, requiring second-order differencing. Model comparison identified TBATS/BATS as having the highest accuracy on the test set (MAPE = 10.53%, sMAPE = 0.11), while the best-fitting overall model for long-term projection was ARIMA (0,2,2) (MAPE = 14.62%). Projections indicate that Somalia’s urban population will reach approximately 13.1 million by 2030 and is expected to double to 22.2 million by 2050. <bold>Conclusion</bold> : The findings suggest that Somalia is on a path of unprecedented urban expansion, with the urban population expected to increase by approximately 113% over the next 25 years. TBATS and ARIMA models proved to be robust tools for demographic forecasting in this context. Policymakers must implement proactive urban planning strategies to manage this growth, mitigate the intensification of urban heat islands, and address the socioeconomic disparities inherent in rapid, uncontrolled urbanization. </p>