In order to identify the variation characteristics of extreme precipitation and quantitatively reveal its impact on the encounter probability of rainstorm and tidal level in coastal areas, using Shenzhen City in the coastal area of South China as the study area, the variation of extreme precipitation from 1965 to 2016 was investigated with the Pettitt method. A joint distribution model of rainstorm and tidal level was constructed based on the Copula function with the study period divided into different stages by the change year. The changes of the joint risks of encounter of rainstorm and tidal level at different stages were compared and analyzed. The results show that 1985 was the abrupt change point of the maximum one-day precipitation in Shenzhen City. The risk of encounter of rainstorm and tidal level with a same return period after 1985 was significantly higher than that before 1985 due to changes in extreme precipitation characteristics. The Copula function, which is easy to solve and has strong adaptability, can be used to quantitatively solve the joint risk of rainstorm and tidal level in coastal cities.