Accurately modelling rainfall events is crucial for flood risk assessment and stormwater infrastructure design. However, transforming statistical characteristics of events into relevant rainfall patterns is challenging due to the natural variability of rainfall. Two commonly used methods to schematize rainfall events have limitations: the nested storm profile overestimates the resulting flow by assuming complete dependence between different durations, while determining the critical event duration by simulating each duration separately assumes independence and underestimates the flow. To overcome these limitations, this study presents a method that models the dependence between different rainfall durations using a Gaussian copula and combines this with marginal rain statistics to create a probabilistic model for the rain event. The SCS Curve Number approach is used to model the resulting flow, and a first-order reliability method (FORM) is applied to determine the critical combination of durations within an event. The findings of this study show that the rainfall events generated using the proposed method result in comparable flows to those produced by conventional design events. While this may not make the model a preferred choice for standard applications, it can still be valuable for flood risk assessments as it provides a probabilistic model that better captures critical rainfall patterns.