Journal of the Irrigation and Drainage Division · 1978 · 17 citations · 0 references
Natural SciencesHydrologyMeteorologyHydrologic ModelingStochastic ModelingWeather ForecastingDiscrete-event System ModelingHydrological ModelingMarkov Chain Monte CarloMarkov ProcessesPrecipitationRainfall-runoff ModelingDiscrete Probability DistributionsGamma Distribution ModelsRainfall ModelingState Transition ProbabilitiesState-dependent Transition ProbabilitiesTransition Probabilities
A modified Markov Chain model is used to generate synthetic traces of daily rainfall amounts at a point. Mixed discrete-continuous distributions are used to model the state transition probabilities. The use of the modified model affords a considerable reduction in historical data requirements for parameter estimation. A two-parameter gamma distribution was used in the model with data from seven weather stations in Kentucky and provided a good representation of the daily point rainfall process.