Bayesian Time-Series Modelling of Cotton Yield under Climatic Variability in Burundi: A Comparison of ARIMA and ARIMAX Models
Abstract
Background: Cotton is an important cash crop in Burundi, but its productivity has declined over recent decades under the combined influence of climatic variability, agronomic and socio-economic constraints. This study aimed to analyze the temporal dynamics of cotton yield and generate probabilistic forecasts using a times-series modelling framework. Methods: Annual cotton yield and climatic data (temperature, precipitation and relative humidity) for the period 1995-2025 were analyzed. Stationarity was assessed using the Kwiatkowski-Phillips-Schmidt-Shin (KPSS) test. Bayesian ARIMA and ARIMAX models were estimated using Markov chain Monte Carlo (MCMC) methods in Stan. Model performance was evaluated using the Leave-One-Out Information Criterion (LOOIC) and the Widely Applicable Information Criterion (WAIC). Probabilistic forecasts were generated for 2026-2030. Results: Cotton yield showed marked interannual variability and an overall declining trend during the study period. Although precipitation exhibited greater variability than temperature and relative humidity, none of the climatic variables showed statistically significant posterior effects after accounting for temporal dependance. The Bayesian ARIMA (2,1,0) model outperformed the ARIMAX specifications, while residual diagnostics confirmed an adequate model fit. Forecasts suggested a gradual decline in cotton yield from 2026 to 2030, with progressively wider credible intervals, indicating increasing uncertainty over longer forecasting horizons. Conclusion: Historical yield persistence explained cotton yield dynamics better than annual climatic averages alone. The findings suggest that climate effects are likely nonlinear and mediated by seasonal and agronomic factors not captured by annual climatic indicators. Bayesian time-series modelling provides a reliable framework for probabilistic agricultural forecasting and can support climate adaptation strategies and evidence-based decision-making for sustainable cotton production in Burundi.
Downloads
References
Adebiyi, A. A., Adewumi, A. O., & Ayo, C. K. (2014). Comparison of ARIMA and Artificial Neural Networks Models for Stock Price Prediction. Journal of Applied Mathematics, 2014, 1–7. https://doi.org/10.1155/2014/614342
Alam, W., Ray, M., Kumar, R. R., Sinha, K., Rathod, S., & Singh, K. N. (2018). Improved ARIMAX modal based on ANN and SVM approaches for forecasting rice yield using weather variables. The Indian Journal of Agricultural Sciences, 88(12), 1909–1913. https://doi.org/10.56093/ijas.v88i12.85446
Bado, V. B., & Bationo, A. (2018). Integrated Management of Soil Fertility and Land Resources in Sub-Saharan Africa: Involving Local Communities. In Advances in Agronomy (Vol. 150, pp. 1–33). Elsevier. https://doi.org/10.1016/bs.agron.2018.02.001
Baumeister, C. J. S., & Hamilton, J. D. (2018). Structural Interpretation of Vector Autoregressions with Incomplete Identification: Revisiting the Role of Oil Supply and Demand Shocks. (24167).
Blanc, E., Quirion, P., & Strobl, E. (2008). The climatic determinants of cotton yields: Evidence from a plot in West Africa. Agricultural and Forest Meteorology, 148(6), 1093–1100. https://doi.org/10.1016/j.agrformet.2008.02.005
Box, G. E. P., Jenkins, G. M., & Reinsel, G. C. (2015). Time Series Analysis Forecasting and Control.
Brooks, S., Gelman, A., Jones, G., & Meng, X.-L. (2011). Handbook of Markov Chain Monte Carlo. CRC Press.
CARTER, C. K., & KOHN, R. (1994). On Gibbs sampling for state space models. Biometrika, 81(3), 541–553. https://doi.org/10.1093/biomet/81.3.541
Forecasting: Principles and Practice (3rd ed). (n.d.). Retrieved May 25, 2026, from https://otexts.com/fpp3/
Gelman, A. (2006). Prior distributions for variance parameters in hierarchical models (comment on article by Browne and Draper). Bayesian Analysis, 1(3). https://doi.org/10.1214/06-BA117A
Gelman, A. (2008). Scaling regression inputs by dividing by two standard deviations. Statistics in Medicine, 27(15), 2865–2873. https://doi.org/10.1002/sim.3107
Gelman, A., Carlin, J. B., Stern, H. S., Dunson, D. B., Vehtari, A., & Rubin, D. B. (2025). Bayesian Data Analysis Third edition (with errors fixed as of 20 February 2025).
Goyal, M., Agarwal, S., Ghalawat, S., & Malik, J. S. (2024). ARIMA and ARIMAX Analysis on the Effect of Variability of Rainfall, Temperature on Wheat Yield in Haryana. Indian Journal of Extension Education, 60(1), 95–99. https://doi.org/10.48165/IJEE.2024.60118
Hyndman, R. J., & Athanasopoulos, G. (2018). Forecasting: Principles and practice. OTexts.
Kaushik, S. (2024). Forecasting Mustard Yield Using Weather Parameters: An Application of the ARIMAX Model. Journal of Computational Analysis and Applications, 33(8).
Link, W. A., & Eaton, M. J. (2012). On thinning of chains in MCMC. Methods in Ecology and Evolution, 3(1), 112–115. https://doi.org/10.1111/j.2041-210X.2011.00131.x
Muhammed, N., & Athar, A. K. (2023, June). Bayesian Analysis of ARMA and BSTS models for COVID-19 data using R and Stan. RT&A, 18(2(73)).
Paul, R. K., & Ghosh, H. (2013). Statistical modelling for forecasting of wheat yield based on weather variables. Indian Journal of Agricultural Sciences, 83(2), 180–183.
Shumway, R. H., & Stoffer, D. S. (2016). Time Series Analysis and Its Applications.
Sultan, B., Baron, C., Dingkuhn, M., Sarr, B., & Janicot, S. (2005). Agricultural impacts of large-scale variability of the West African monsoon. Agricultural and Forest Meteorology, 128(1), 93–110. https://doi.org/10.1016/j.agrformet.2004.08.005
Thornton, P. K., Ericksen, P. J., Herrero, M., & Challinor, A. J. (2014). Climate variability and vulnerability to climate change: A review. Global Change Biology, 20(11), 3313–3328. https://doi.org/10.1111/gcb.12581
Vehtari, A., Gelman, A., & Gabry, J. (2017). Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC. Statistics and Computing, 27(5), 1413–1432. https://doi.org/10.1007/s11222-016-9696-4
Wesner, J. S., & Pomeranz, J. P. F. (2021). Choosing priors in Bayesian ecological models by simulating from the prior predictive distribution. Ecosphere, 12(9), e03739. https://doi.org/10.1002/ecs2.3739
Copyright (c) 2026 Samuel Hatungimana, Emmanuel Barankanira, Edouard Musabanganji, Menus Nkurunziza, Jean de Dieu Nkurunziza
This work is licensed under a Creative Commons Attribution 4.0 International License.
