Flood-Frequency Data for Six Selected Streamgages Following the Central New York Flood of October 31 – November 3, 2019

This U.S. Geological Survey (USGS) data release contains batch formatted annual peak streamflow data (PkFlows_AllSites.txt) through the 2020 water year for six selected USGS streamgages (01321000, 01342797, 01343060, 01346000, 01347000, and 01348000) that recorded the flood of October 31 – November 3, 2019, which severely affected the Mohawk Valley and southern Adirondack region in central New York State. This data release also contains batch formatted specification (PkFlows_AllSites.psf) and output (PEAKFLOWS_ALLSITES.PRT) files from log-Pearson type III (LPIII) flood-frequency analysis of the annual peak streamflow data in version 7.4 of the USGS PeakFQ software (Flynn and others, 2006), which implements the Bulletin 17C (England and others, 2018) methodology for determining flood flow frequency. A comma separated values file (FloodFrequencyEstimates.csv) with estimates of flood magnitudes for selected annual exceedance probabilities from three different types of flood-frequency analysis (LPIII, regional regression equations, and weighted LPIII) is also included. The basins of the selected streamgages include the Sacandaga River basin (located within the upper Hudson River basin), and the East and West Canada Creek basins (located within the Mohawk River basin). Annual peak streamflow data were obtained from the USGS National Water Information System (U.S. Geological Survey, 2016). Considerations that informed the PeakFQ specifications for all six streamgages (including the effect of regulation, differences in station and regional skew, and other considerations) are discussed in Graziano and others (2024). Regional regression equation estimates were obtained for the locations of the four streamgages where streamflow is unregulated (01321000, 01342797, 01343060, and 01348000) using the StreamStats web application (U.S. Geological Survey, 2019), which presently utilizes equations in Lumia and others (2006). Weighted LPIII estimates (for the same four streamgages where streamflow is unregulated) were computed by weighting the (at-site) LPIII estimates with the regional regression equation estimates (Lumia and others (2006), equation 3). The weighted LPIII estimates are considered the most accurate of the three types of flood-frequency estimates. For this analysis, stationarity in the annual peak streamflow records was assumed after visual inspection of the records indicated no clear trends.