Quantifying catchments' nonlinear hydrologic response and solute transport behavior using ensemble unit hydrographs and ensemble hydrograph separation

Abstract:

Catchment processes are nonlinear and nonstationary. As a result, each mm of rain that falls on a catchment may affect streamflow differently, depending on how that individual parcel of rainfall fits into the sequence of past and future precipitation. Characterizing catchments' hydrologic behavior (as distinct from modeling the consequences of individual precipitation scenarios) requires tools for abstracting general patterns from this complex nonstationary rainfall-runoff relationship. Here I present two data-driven, model-independent methods – ensemble unit hydrographs and ensemble hydrograph separation – for characterizing catchments' responses to precipitation inputs.

Ensemble unit hydrographs are thematically related to conventional unit hydrographs, but with key conceptual and mathematical differences. One important difference is that they can separately quantify the characteristic impulse response to different intensities of precipitation, falling under different catchment antecedent conditions. Thus they can be used to map out nonlinear patterns in catchment hydrologic response, directly from data.

Similarly, ensemble hydrograph separation is thematically related to, but distinct from, conventional hydrograph separation. It uses tracer time series to estimate both "backward" transit time distributions (the fraction of streamflow that originated as rainfall at different lag times in the past) and "forward" transit time distributions (the fraction of rainfall that will become future streamflow following different time lags). It can separately quantify catchments' transport behavior under different catchment conditions, and different types and intensities of precipitation. Thus it can be used to quantify nonlinearities in catchment transport behavior.

Using both techniques jointly allows one to quantify the velocity and celerity of transport and runoff generation at the catchment scale, and also to quantify how they respond to catchment characteristics and hydrometeorological forcing. This presentation will briefly introduce these techniques and illustrate them using data from several experimental catchments.