Please join us for a Woods Seminar with Ciaran Harman, Johns Hopkins University.
Catchment hydrology has struggled to develop a unified approach to understanding and predicting the flow of water and the transport of solutes at catchment scales. The observation of relatively old stream water in small watersheds, even at high discharge, has demonstrated that precipitation inputs can mobilize large stores of water that integrate previous inputs over many months to years (or more). These subsurface storages and flow paths are challenging to observe directly even in intensively-studied watersheds where the latest geophysical tools can be deployed. The ability of a model to reproduce the catchment water balance or the shape of a hydrograph has proved to be a poor, even uninformative, indicator of its ability to reveal the internal organization and process dynamics that apparently control the delivery of water and solutes to streams.
Our current difficulty answering basic questions about flow and transport through shallow and deeper pathways, particularly in ungauged watersheds, arguably limits our ability to deliver the kind of predictions needed for water resource planning and regulation the over the long term, and across spatial scales. In this talk I will discuss this long-standing challenge, and two approaches to addressing it. One is to seek physically-meaningful ways of conceptualizing and mathematically representing flow and transport at catchment and large-grid scales. The other is to reframe the question entirely: perhaps in addition to asking “what is the hydrologic structure of the landscape?” we should also be asking “why is it so?”.
New process representations are needed that a) capture the integrated effect of un-observable internal structure without requiring us to resolve that structure explicitly, b) are physically meaningful, yet c) are parsimonious enough that they stand some chance of being inferable from available data. StorAgeSelection (SAS) theory has emerged as a powerful tool for both interpreting active and passive tracer data, and making predictions about the movement of solutes in a variety of hydrologic settings. SAS is a generalization of transit time distributions that allows for time-variable fluxes through the system, and similarly provides a framework for capturing the integrated effect of internal structure and heterogeneity on transport. When used to interpret water tracer data (e.g. stable water isotopes), SAS functions have revealed details of the volume of internal storage turning over in the landscape, how rates of turnover differ between younger and older stores of water, and how that turnover changes as hydrologic conditions change. I will discuss advances we have made toward empirically observing SAS functions, connecting their properties to the physical structure of hillslopes, stream reaches, and small watersheds, and using them to anticipate the effects of climate change on agricultural nitrate delivery to streams.
Ultimately, however, this approach faces the same challenge that has plagued catchment hydrology: much of the most important physical structure controlling flow and transport is hidden in the subsurface. Perhaps this hidden structure can be better understood by asking: how do landscapes develop the hydrologic properties they have? The potential and promise of ‘co-evolution’ or ‘Darwinian hydrology’ has been widely discussed, but there have been limited attempts to address the question directly. Here I will present an attempt to understand how the internal structure of chemically-weathered hillslopes could both control and be controlled by the lateral flow of meteoric water towards adjacent streams. I will present a model that couples hillslope hydraulics and solute transport with a parsimonious model of geochemical weathering and its effect on porosity and permeability. The resulting approximate analytical solutions yield realistic hillslope internal architectures when supplied with realistic geomorphic, geochemical and hydrologic parameter values derived from literature.