The application of reweighted Markov chains in water-sediment prediction in inland river basins—A case study of the Shiyang River Basin.

• Traditional Markov chain prediction method is mostly limited to state prediction. In this paper, the reweighted Markov chain prediction method is capable of numerical prediction, which is a key improvement of the Markov chain prediction method, not only improving the prediction accuracy, but also extending the application of the method. Although the runoff of the Shiyang River basin is relatively small and the amount of sediment transferred is very low, the prediction results still will not be affected. • Compared with other Markov chain improvement methods, the reweighted Markov chain can exploit the information more comprehensively from the random sequences. The example of water-sediment prediction of Shiyang River has shown that the established reweighted Markov chain prediction method is scientifically clear and easy to calculate, which is suitable for hydrological observation. • For the prediction when a peak occurs, there may be a relatively large error from the actual prediction. In this paper, a 37-year analysis period has been chosen since the short sequence of analysis period also contributes to the possibility of errors being present. The Markov prediction steps are initially identified as a random sequence, and the error in the numerical prediction may be due to small disruptions. River runoff and sediment play a crucial role in shaping river morphology, providing water resources and maintaining the water ecological balance of the watershed. In this paper, the runoff and sediment transport from 1975 to 2011 at the Jiutiao Ridge hydrological station in the Shiyang River basin are selected as the experimental period, and the runoff and sediment transport from 2011 to 2016 are selected as the validation period. Using a reweighted Markov chain, the correlation analysis of the annual runoff and the overhanging mass sediment transport in the Jiutiao Ridge hydrological station in the Shiyang River basin has been performed to predict the future changes of runoff and sediment transport in the Shiyang River. From the prediction results, it can be seen that there is limited error between the predicted value and the actual value, which can validate the application of Markov prediction method in hydrological water and sand, as well as provide a basis for future water resource management and optimization of the Shiyang River basin. [ABSTRACT FROM AUTHOR]