Knowledge of the seasonal properties of annual peak streamflows is important for planning flood control and response operations and is a key element for the attribution of flood generating mechanisms. A multi-phase Transportation Pooled Fund study is currently underway to evaluate potential nonstationarity in annual peak-flow records in the Midwest United States due to hydroclimatic variability. Based on the results of the detected nonstationarities, there is a need to evaluate changes in seasonality for later attributions. Characterizing changes in seasonality is important for understanding the effects of climatic change on flood flows, particularly in temperate regions, where floods may be caused by snow, rain, or combinations of the two. Circular statistics are used to characterize the seasonal properties of annual maximum series (AMS) and peaks-over-threshold (POT) peak streamflows for qualifying U.S. Geological Survey streamgages without regulation or diversion among common 75-, 50- and 30-year trend periods through water year 2020. By using common trend periods, we address mismatched comparisons that may occur as a result of multi-decadal persistence arising from hydroclimatic variability that has been found in the nine-state study region. Further, changes in seasonality among the three common trend periods are also evaluated. Cohesive spatial patterns are found among the three trend periods across the study region, and pronounced changes in the mean and dispersion of peak-flow dates from the 50- and 30-year trend periods are found. Lastly, AMS peak flow magnitude time series with detected change points (abrupt changes) in the median and/or scale are analyzed on either side of the change point to evaluate changes in their circular statistics due to hydroclimatic variability.
