A Runoff Prediction Model Based on Nonhomogeneous Markov Chain

Runoff prediction is one of the important research fields of hydrology. As for the runoff series with unstable, poor periodicity and non-obvious tendency, this paper solves the problem that the general traditional models are not suitable for the short and medium-term prediction of such runoff series. To describe the nonhomogeneous characteristics of runoff series, a novel prediction model is established based on a nonhomogeneous Markov chain (NHMC-RPM). In this model, the probability distribution function of weekly runoff is obtained and the predicted value is calculated using the expected state. Taking the Yellow River as a case, the prediction results show that the NHMC-RPM is more accurate than other traditional models. The model presented in this work may be used to deal with similar runoff or other series data, as well as provide a practical approach for river managers to predict short and medium-term runoff.

Automated summary

This paper presents the application of a nonhomogeneous Markov chain prediction model in runoff series with unstable, poor periodicity, and non-obvious tendency. Using the Yellow River as a case study, the results show that the NHMC-RPM model is more accurate compared to traditional models, providing a practical approach for river managers to predict short and medium-term runoff.