A mathematical model of monthly flow sequences / Un modèle mathématique des séquences débits mensuels

Abstract A general mathematical model for the generation of synthetic monthly flows is developed. The relevance of the model is its simultaneous preservation of lag-zero cross-correlations between successive months together with the first-order serial correlation of each month. It is shown that the conventional lag-one Markov model and the Thomas-Fiering model are special cases of the model presented. All of the necessary analytical expressions of the correlation structure of the model are derived. Finally, the fitting of the model to monthly flow data and the generation of synthetic sequences are discussed.