Copula-Based First-Order Markov Process for Forecasting and Analyzing Risk of Water Quality

Abstract:

This study applied the copula-based first-order Markov process to predict and analyze the risk for univariate water quality time series at four water quality assessment stations located on Snohomish River in the state of Washington. The copula-based first-order Markov process was applied to forecast the future values of the time series. Then, the Value at Risk (VaR) values with were determined for all of the time series to analyze risk for the study parameters. Results showed that the copula-based first-order Markov process was able to capture the temporal dependence structure of the time series and was capable of satisfactorily predicting future values and VaRs. Overall, the copula-based Markov process may be an efficient tool in the assessment of water quality and the associated risk in river systems.