Baseflow Separation and Low-Flow Permit on a Regulated River
A 480 km2 catchment has a 40-year daily streamflow record. A hydropower operator proposes a new diversion of 0.9 m3/s. The gauged mean annual flow is 6.2 m3/s, and the 7Q10 low flow is estimated at 0.75 m3/s. A recursive digital filter applied to the record gives a baseflow index of 0.42, but the filter is sensitive to its turning point and the catchment has 18 percent impervious urban area upstream of the gauge. Recent summers show lower low flows than the long record suggests, and a reservoir 12 km upstream regulates flow with daily peaking. The regulator must decide whether the diversion can be permitted with a minimum environmental flow condition.
1. Critique the use of the long-record 7Q10 and the filter-derived baseflow index as the basis for the environmental flow condition in this catchment.
Read the worked answer
The 7Q10 is a statistical low-flow index from the full record; it assumes stationarity and a natural flow regime. Here two assumptions fail: urban imperviousness and upstream regulation alter both the magnitude and timing of low flows, and recent summers are drier than the long record. The 7Q10 of 0.75 m3/s may therefore overstate reliably available low flow.
The baseflow index of 0.42 is filter-dependent; recursive filters have no unique physical baseflow, and the turning-point parameter materially changes the split. Reporting a single BFI without sensitivity analysis overstates precision.
Daily peaking from the reservoir further decouples instantaneous low flow from the daily statistic. A defensible condition should use recent-period, regulation-aware low-flow statistics, an explicit filter sensitivity range, and a flow-duration or habitat-based criterion rather than a single 7Q10.
2. The operator argues that 0.9 m3/s is only 15 percent of mean annual flow, well within the baseflow. Construct the counter-argument and propose a defensible permit condition.
Read the worked answer
Percentage of mean annual flow is the wrong metric at low flow. The relevant comparison is the diversion against the low-flow regime: 0.9 m3/s exceeds the estimated 7Q10 of 0.75 m3/s, so in a dry summer the diversion could consume the entire low flow plus more.
The baseflow argument also fails because BFI is an annual average partition, not a guarantee of dry-season availability, and reservoir peaking means instantaneous flow can fall below the daily mean.
A defensible condition: cap abstraction as a percentage of instantaneous flow (for example, cease-to-pump at a specified low-flow threshold above the regulation-adjusted 7Q10), require real-time gauging and metering, set seasonal limits, and review against recent dry-year data. The permit should state the low-flow threshold, the measurement method, and a compliance trigger, not a single annual volume.