Determine location of changepoints after performing a Pettitt's change point test

I have created a series of three ARIMA models (daily precipitation, daily discharge, and daily increase in discharge) and used a Pettitt change point test to determine the presence of change points. The tests show that the precipitation model has three probable changepoints, the discharge model has two probable changepoints, and the increase in discharge model has 1 probable changepoint. I now know that changepoints exist, but would like to be able to determine where these changepoints are in the time series as well as plot the locations of the changepoints. Does anyone have any advice on how to proceed. TIA.