Rainfall potential analysis plays a critical role in the management of air resources, mitigation of hydrometeorological disasters, and agricultural activity planning. Accurate estimation of rainfall patterns is essential to ensure effective decision-making in irrigation systems, water resource management, and disaster risk reduction strategies. This study aims to model the probability of rainfall occurrence using a statistical approach based on historical data obtained from the Bureau of Meteorology. The data spans a multi-year period and captures seasonal and regional variability in rainfall events. To characterize rainfall patterns, various probability distributions are tested, including the exponential distribution and the Weibull distribution, which are commonly applied in hydrological studies. Furthermore, the Markov chain method is employed to assess the likelihood of rainfall occurrence on a given day based on the conditions of the preceding day, thereby capturing temporal dependencies. Parameter estimation is conducted using Maximum Likelihood Estimation (MLE), a robust statistical method that enhances the precision of the model. The suitability of each probability distribution in representing the observed rainfall data is evaluated through goodness-of-fit tests such as the Kolmogorov-Smirnov test. The findings reveal that certain distributions align more closely with the local rainfall characteristics, demonstrating the importance of regional analysis in climate modeling. The combination of probabilistic modeling, Markov analysis, and rigorous statistical testing provides a reliable framework for forecasting rainfall. These results are expected to serve as a scientific basis for stakeholders in agriculture, environmental planning, and disaster preparedness, offering insights that support sustainable water resource utilization and risk management.