A DAILY STOCHASTIC PRECIPITATION GENERATOR: CASE STUDY IN OHIO

Stochastic generation of precipitation are probabilistic models that produce synthetic time series data mimicking observed weather while keeping the statistical signature. These weather generators have been widely used for decades for water resource management, numerical modeling, hydrology, and agricultural applications as they emulate realistic sequences.

In this study, a simple stochastic precipitation generator based on a first-order two-state Markov chain model was developed using multi-decadal data to predict daily rainfall scenarios at a single site for future climate change scenarios. The model was initially developed for the city of Athens but later updated and applied to multiple larger cities across the state of Ohio.

The model was reasonably successful in generating rainfall statistics and preserving daily characteristics of historical observations for most of the sites used in the study.