Estimation of Channel Bankfull Occurrence from Instantaneous Discharge Data

Bankfull discharge $\left(Q_{\mathrm{bf}}\right)$ occurrence has predominantly been described with the annual maximum exceedance random variable, $p_a$ , however the wide availability of daily peak streamflow (DPS) data allows for the description of occurrence with two alternative random variables. The two alternative random variables are: (1) flow duration, $p_f$ , the probability that daily discharge is at, or exceeds, $Q_{\mathrm{bf}}$ and (2) partial duration, $p_b$ , the probability that an independent daily discharge is at, or exceeds, $Q_{\mathrm{bf}}$ . In this research, statistics for $p_f$ and $p_b$ use DPS data, whereas $p_a$ is estimated with four statistics using either DPS or annual peak streamflow (APS) data. Analyses of approximately $15\text{years}$ of DPS and $45\text{years}$ of APS at 46 bankfull surveyed gauges in Maryland, North Carolina, and New York were performed to illustrate differences between the four $p_a$ statistics and the three $Q_{\mathrm{bf}}$ random variables, $p_a$ , $p_f$ , and $p_b$ . For the $p_a$ analysis, non-log Pearson type-III (non-LPT3) statistics generated median recurrence intervals (RI) for $Q_{\mathrm{bf}}$ that were 9% longer than LPT3 statistics. Processing the DPS data with the partial duration estimator shows $Q_{\mathrm{bf}}$ events arrive 1.5 times per year on average $(\mathrm{RI}=0.70\text{years})$ and 8 times per year maximum $(\mathrm{RI}=0.12\text{years})$ . The flow duration estimator shows $Q_{\mathrm{bf}}$ has durations of $2.1\text{days}\text{per}\text{year}$ on average and $13.2\text{days}\text{per}\text{year}$ maximum. Although the DPS record is approximately $30\text{years}$ shorter than the APS record, the DPS data capture the $1.4\text{year}$ $Q_{\mathrm{bf}}$ frequency of the longer APS record. Statistics for $p_a$ , $p_f$ , and $p_b$ inform the river manager to the likelihood of three distinct $Q_{\mathrm{bf}}$ phenomena: Annual maximum occurrence, daily duration of events, and daily time between events.