The current version of HEC-SSP allows the user to perform a frequency analysis using a peaks over threshold approach. The Peaks Over Threshold Analysis allows users to select and filter a data set to a partial duration series, fit a Generalized Pareto distribution, convert the partial duration probabilities to annual exceedance probabilities, and compute confidence limits and expected probability. As in the Data Filter, the Peaks Over Threshold Analysis allows users to filter a dataset based on a threshold value, time separation, and magnitude differential criterion. Users can investigate threshold selection using diagnostic plots of number of exceedances, rate of exceedances, mean excess, Generalized Pareto distribution shape parameter, and the right tail weighted Anderson Darling test statistic. The analysis converts the partial duration probabilities to annual exceedance probabilities empirically or analytically. Additionally, the annualized expected probability curve and confidence limits are computed.
Examples of this type of analysis can be found here:
