Explaining Variations in Streamflow Recession Rate Using Observations of Subcatchment-Scale Heterogeneity in Flow Contribution

Abstract:

For several decades, streamflow recession events has often been evaluated by comparing the rate of change in flow (dQ/dt) versus the concurrent average flow (Q) within so-called dQ/dt-Q graphs. Recent work has evaluated independent streamflow recession events instead of a cloud of all dQ/dt and Q data points plotted irrespective of time of occurrence. When viewed as individual events, the dQ/dt-Q cloud displays meaningful structure, with individual curves having relatively constant slope and displaying seasonal offsets. With this shift to evaluating independent recession curves, recession has also been considered as possibly being dependent on landscape features instead of being dependent on aquifer properties.

From the viewpoint of considering landscape controls on recession, we consider whether spatial heterogeneity in flow at the scale of first-order subcatchments can explain the behavior of recession in mesoscale watersheds. We present continuous flow data collected concurrently in multiple, small first-order subcatchments that drain to a common outlet. To compare among subcatchments, we make use of area-normalized fractional flow [ (Q_sub/Area_sub)/(Q_main/Area_main) where “sub” refers to the respective subcatchment and “main” refers to the main catchment each subcatchment is nested within] . We observe clear variations in area-normalized fraction flow contributions among subcatchments. In particular, we see that some subcatchments contribute flow disproportionately larger than their drainage area early in the baseflow period while others contribute relatively little early on but contribute disproportionately larger amounts during later baseflow periods. This difference in hydrologic behavior of subcatchments is quantified within a model composed of multiple, parallel linear reservoirs of differieng drainage rates. The parallel linear reservoir model is able to reproduce the distinct signatures of observed individual recession curves. Using observations from several watersheds in the Northeastern US, this work offers a novel approach to integrate observable subbasin heterogeneities into a simple model that is able to reproduce distinct streamflow recession behaviors of mesoscale watersheds.