L-moment estimation using annual maximum (AM) and peak over threshold (POT) series in regional analysis of flood frequencies

This article is a contribution to the development of robust approaches for frequency analysis in hydrology based on regional datasets. It focuses on finding relations between flood estimates based on Annual Maximum Series (AMS) and Peak Over Threshold Series (POTS). A consistent rule is elaborated for changing between the AMS and POTS representations for the individual observed values, for the choice of index flood and for parameter estimation. The point of departure for the analysis is the Langbein formula. The median related to AM as index flood is recommended, because, besides being robust, this characteristic can be estimated from both AMS and POTS. The methodology allows a direct comparison between L-moments of the annual maximum based on AMS and POTS, respectively, when the PWM method for estimating L-moments is applied. A conclusion of the analysis is that AM and POT data give identical results for estimation of AM L-moments but the latter has the smallest estimation errors.