Journal article · 2026 · Open access

A Comparative Statistical Time-Series Framework for Long-Term Rainfall Forecasting in Bangladesh Using ARIMA, SARIMA, ARCH, and VAR Models

Hira Farman, Muhammad Fahad, Muhammad Suleman Soomro, Sharaf Hussain, Alisha Farman

Pakistan Journal of Engineering Technology & Science, 14(1), 52–82 · Published 16 June 2026 · DOI 10.22555/pjets.v14i1.1571

Open PDFView at the publisherCite this paper

In brief

“A Comparative Statistical Time-Series Framework for Long-Term Rainfall Forecasting in Bangladesh Using ARIMA, SARIMA, ARCH, and VAR Models” is a 2026 journal article by Hira Farman and 4 co-authors, published in Pakistan Journal of Engineering Technology & Science. OpenAlex counts 0 citations as of 20 November 2026. It is open access under a CC BY licence: the full text is free to read at Pakistan Journal of Engineering Technology & Science.

Citations
0
References
54
Authors
5

Abstract

Forecasting rainfall is a complicated process and is the product of several climatic variables. Rainfall prediction is essential for optimal crop planning, management of water resources, and disaster management in Bangladesh, where agriculture is a significant contributor to the economy. The present study is devoted to the investigation and comparison of four different models of statistical time series forecasting, namely ARIMA, SARIMA, ARCH, and VAR, for the monthly rainfall and temperature of Bangladesh for 1901-2019. To measure the performance of the models, the following data preprocessing, stationarity test, seasonality analysis, volatility modelling, and multivariate dependency analysis were developed. Evaluation of forecasting models is done through Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), Coefficient of Determination (R²), the Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), and Percentage of Correct Directional Change (POCID). The results show that the SARIMA (1, 1, 1, 12) model has the lowest RMSE of 89.3049, MAE of 56.0834, and R2 of 0.7851 among other models, which proves that the model is best capable of forecasting the rainfall data of Bangladesh with good accuracy and explains the seasonal characteristics of the data very well. The highest value of direction of forecasting accuracy (POCID is 80.00% and was obtained by the VAR model, which showed that the model could be used to capture interdependence between climatic variables and predict the trend of rainfall. Compared to this, ARIMA and ARCH had lower forecasting accuracy and higher prediction error. The novelty of the study is that the four model types, such as ARIMA, SARIMA, ARCH, and VAR models, are compared with a common evaluation framework using a century-long dataset of Bangladesh precipitation data. The results reveal that SARIMA is the most appropriate model for seasonal rainfall forecasting, and the VAR model is useful for gaining insight into directional trend forecasting. The proposed framework can be used to inform agricultural planning, flood preparedness, and water resource management in climate-sensitive areas.

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Farman, H., Fahad, M., Soomro, M. S., Hussain, S., & Farman, A. (2026). A Comparative Statistical Time-Series Framework for Long-Term Rainfall Forecasting in Bangladesh Using ARIMA, SARIMA, ARCH, and VAR Models. Pakistan Journal of Engineering Technology & Science, 14(1), 52–82. https://doi.org/10.22555/pjets.v14i1.1571

MLA 9

Farman, Hira, et al. “A Comparative Statistical Time-Series Framework for Long-Term Rainfall Forecasting in Bangladesh Using ARIMA, SARIMA, ARCH, and VAR Models.” Pakistan Journal of Engineering Technology & Science, vol. 14, no. 1, 2026, pp. 52–82. https://doi.org/10.22555/pjets.v14i1.1571.

Chicago

Farman, Hira, Muhammad Fahad, Muhammad Suleman Soomro, Sharaf Hussain, and Alisha Farman. 2026. “A Comparative Statistical Time-Series Framework for Long-Term Rainfall Forecasting in Bangladesh Using ARIMA, SARIMA, ARCH, and VAR Models.” Pakistan Journal of Engineering Technology & Science 14 (1): 52–82. https://doi.org/10.22555/pjets.v14i1.1571.

Harvard

Farman, H. et al. (2026) ‘A Comparative Statistical Time-Series Framework for Long-Term Rainfall Forecasting in Bangladesh Using ARIMA, SARIMA, ARCH, and VAR Models’, Pakistan Journal of Engineering Technology & Science, 14(1), pp. 52–82. doi:10.22555/pjets.v14i1.1571.

IEEE

H. Farman, M. Fahad, M. S. Soomro, S. Hussain, and A. Farman, “A Comparative Statistical Time-Series Framework for Long-Term Rainfall Forecasting in Bangladesh Using ARIMA, SARIMA, ARCH, and VAR Models,” Pakistan Journal of Engineering Technology & Science, vol. 14, no. 1, pp. 52–82, 2026, doi: 10.22555/pjets.v14i1.1571.

BibTeX

@article{farman2026comparative,
  title     = {{A Comparative Statistical Time-Series Framework for Long-Term Rainfall Forecasting in Bangladesh Using ARIMA, SARIMA, ARCH, and VAR Models}},
  author    = {Farman, Hira and Fahad, Muhammad and Soomro, Muhammad Suleman and Hussain, Sharaf and Farman, Alisha},
  journal   = {Pakistan Journal of Engineering Technology \& Science},
  year      = {2026},
  volume    = {14},
  number    = {1},
  pages     = {52--82},
  doi       = {10.22555/pjets.v14i1.1571},
  url       = {https://doi.org/10.22555/pjets.v14i1.1571}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Topic and keywords

Social Sciences › Decision Sciences › Management Science and Operations Research › Forecasting Techniques and Applications

  • precipitation forecasting
  • seasonal rainfall
  • monthly rainfall
  • monthly temperature
  • ARIMA
  • SARIMA
  • ARCH models
  • vector autoregression

Work with this paper

Most-cited papers in Decision Sciences

All highly cited Decision Sciences papers

Questions about this paper

How do I cite this paper in APA style?

Farman, H., Fahad, M., Soomro, M. S., Hussain, S., & Farman, A. (2026). A Comparative Statistical Time-Series Framework for Long-Term Rainfall Forecasting in Bangladesh Using ARIMA, SARIMA, ARCH, and VAR Models. Pakistan Journal of Engineering Technology & Science, 14(1), 52–82. https://doi.org/10.22555/pjets.v14i1.1571

How many times has this paper been cited?

0 times, according to OpenAlex on 20 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access at Pakistan Journal of Engineering Technology & Science: sign in and open the PDF from this page.

Who wrote this paper?

Hira Farman, Muhammad Fahad, Muhammad Suleman Soomro, Sharaf Hussain, Alisha Farman.

What is the DOI of this paper?

10.22555/pjets.v14i1.1571, permanently at https://doi.org/10.22555/pjets.v14i1.1571. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 20 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.