Stochastic rainfall occurrence modeling in Benin

Document Type

Conference Proceeding

Publication Date



Hydrologic modeling often requires long sequences of daily rainfall observations, which are rare in sparsely gauged basins. Multiple-order Markov chain occurrence models are common stochastic rainfall generators used to provide extensive daily rainfall sequences in areas where such observation record lengths are not available, such as Benin in Western Africa. Using objective information criterion methods, the optimum stochastic occurrence model(s) are determined for multiple gauges in Benin. Overdispersion is also measured to determine the model's representation of interannual variability. © 2007 ASCE.

Publication Title

Restoring Our Natural Habitat - Proceedings of the 2007 World Environmental and Water Resources Congress