The demand for water withdrawal from rivers continues to increase world‐wide, and these water withdrawals can affect fish habitat and aquatic life in general. The study of river hydrology and environmental flows often require the calculation of flow metrics (e.g., mean annual flow, flow duration statistics, low flow frequency analyses, etc.); however, uncertainties in these flow metrics have not been fully studied in the literature. The present study looks at uncertainties related to four flow metrics used in environmental flow studies using bootstrapping techniques for five hydrometric stations (located in the Miramichi River basin, Canada). The objectives of the study are: (a) to calculate and compare different flow metrics, (b) to carry out bootstrap samples in order to assess uncertainties as a function of sample size and (c) to determine which flow metrics are associated with the least uncertainty when calculating environmental flows using hydrological based methods. Results showed that uncertainties rapidly decreased from a 5‐ to a 15‐year sample size for each flow metric. The uncertainties from a 15‐ to a 50‐year sample size decreased, but not as rapidly as from a 5‐ to a 15‐year sample. The mean annual flow metric showed the least variability. For monthly flow metrics, sample sizes of 25–30 years were needed to obtain similar coefficients of variation as the 5‐year sample size of the mean annual flow. Drainage area did not impact the mean annual flow metric; however, uncertainty was observed to vary more significantly for smaller rivers on a monthly basis.

No abstract