Irrigation and Water Engineering

Irrigation and Water Engineering

Application of Advanced Hidden Markov Model in Hourly Rainfall Simulation

Document Type : Original Article

Authors
Zeynab Afsharipour 1
ALI Salajegheh 2
Behzad Moshiri 3
Javad Bazrafshan 4
Hamid reza Keshtkar 5
1 1. Phd student, Department of Reclamation of Arid and Mountainous Regions, Faculty of Natural Resources, University of Tehran,Iran
2 2. Professor, Department of Reclamation of Arid and Mountainous Regions, Faculty of Natural Resources, University of Tehran,Iran
3 Professor, School of Elec. & Comp. Engineering, College of Engineering, University of Tehran,Iran
4 Lecturer, Department of Water Resources, Faculty of Agriculture, , University of Tehran,Iran
5 5. Assistant professor, Department of Reclamation of Arid and Mountainous Regions, Faculty of Natural Resources, University of Tehran,Iran
10.22125/iwe.2025.508700.1866
Abstract
Artificial rainfall simulations by using stochastic models provide a wider range of potential meteorological scenarios compared to historical observational data, enabling a more comprehensive assessment of potential water resource challenges. The hidden Markov model framework for simulating hourly rainfall is capable to capture essential characteristics of daily precipitation, including dry periods and droughts, seasonal and temporal variations in occurrence and intensity, as well as tendencies towards extreme values. This model incorporates several innovations compared to conventional methods, including three(Dry, Wet& Wetter) and clone states for dry periods and temporal non-homogeneity in the transition matrix. It is set up in a Bayesian framework that allows for quantification of parametric and predictive uncertainty, allowing for full model evaluation through posterior predictive analyses. For this research, eight years of hourly rainfall data (from 2015 to 2022) from the Amir Kabir station in Alborz province were utilized. The results of the model are interpretable and allow for the examination of seasonal and annual variations in hourly rainfall occurrence and intensity. Considering the various complex aspects of rainfall patterns at an hourly time scale, this model can serve as a valuable tool for meteorologists, hydrologists, and water resource planners.
Keywords
Bayesian
hidden Markov model
posterior predictive
predictive uncertainty
rainfall
Subjects
سلاجقه،ع.خسروی،م.1393. مدل­های حوزهء آبخیز.انتشارات دانشگاه تهران(3605)
Brooks, S. P., & Gelman, A. (1998). General methods for monitoring convergence of iterative simulations. Journal of computational and graphical statistics, 7(4), 434-455.‏
Chandler, R. E., Isham, V. S., Northrop, P. J., Wheater, H. S., Onof, C. J., & Leith, N. A. (2014). Uncertainty in rainfall inputs. Applied Uncertainty Analysis for Flood Risk Management, edited by: Beven, KJ and Hall, JW, Imperial College Press: London, 101-152.‏
Dawkins, L. C., Osborne, J. M., Economou, T., Darch, G. J., & Stoner, O. R. (2022). The advanced meteorology explorer: a novel stochastic, gridded daily rainfall generator. Journal of Hydrology, 607, 127478.‏
de Valpine, P., Turek, D., Paciorek, C. J., Anderson-Bergman, C., Lang, D. T., & Bodik, R. (2017). Programming with models: writing statistical algorithms for general model structures with NIMBLE. Journal of Computational and Graphical Statistics, 26(2), 403-413.‏
Economou, T., Bailey, T. C., & Kapelan, Z. (2014). MCMC implementation for Bayesian hidden semi-Markov models with illustrative applications. Statistics and Computing, 24, 739-752.‏
Guo, L., Jiang, Z., Li, L., & Wang, H. (2022). Increase of future summer rainfall in the middle and lower reach of the Yangtze River basin projected with a nonhomogeneous hidden Markov model. Geophysical Research Letters, 49(7), e2021GL097325.‏
Maruotti, A., & Rocci, R. (2012). A mixed non‐homogeneous hidden Markov model for categorical data, with application to alcohol consumption. Statistics in Medicine, 31(9), 871-886.‏
Meligkotsidou, L., & Dellaportas, P. (2011). Forecasting with non-homogeneous hidden Markov models. Statistics and Computing, 21, 439-449.‏
Ramadhan, R., & Devianto, D. (2020). A Hidden Markov Model for Forecasting Rainfall Data Availability at the Weather Station in West Sumatra. Science and Technology Indonesia, 5(2), 34-40.‏
Segond, M. L., Wheater, H. S., & Onof, C. (2007). The significance of spatial rainfall representation for flood runoff estimation: A numerical evaluation based on the Lee catchment, UK. Journal of Hydrology, 347(1-2), 116-131.‏
Stoner, O., & Economou, T. (2020). An advanced hidden Markov model for hourly rainfall time series. Computational Statistics & Data Analysis, 152, 107045.
Scott, S. L. (2002). Bayesian methods for hidden Markov models: Recursive computing in the 21st century. Journal of the American statistical Association, 97(457), 337-351.‏
Team, R. C. (2019). RA language and environment for statistical computing, R Foundation for Statistical. Computing.‏
Wood, S. N. (2016). Just another gibbs additive modeller: interfacing JAGS and mgcv. arXiv preprint arXiv:1602.02539.‏
Xie, M., Jiang, Z., & Sainju, A. M. (2018, July). Geographical hidden markov tree for flood extent mapping. In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (pp. 2545-2554).‏
Zucchini, W., & MacDonald, I. L. (2009). Hidden Markov models for time series: an introduction using R. Chapman and Hall/CRC.‏