Overland flow connectivity states in Lower Triangle Region, East River, Colorado

Abstract:

The Lower Triangle Region (LTR) in East River Watershed, Colorado, is an intensive research site of the DOE-funded Watershed Scientific Focus Area project. The topography in the LTR is rough with steep gradients. Ephemeral streams emerge during spring. The Watershed Scientific Focus Area project aims to understand and predict hydrological and biogeochemical processes across multiple scales ranging from genome- to catchment-scale. Indeed, understanding how these processes (inter)act across different scales has been identified as an unsolved problem in hydrology. From a numerical modeling perspective, accounting for the multiscale nature of these processes makes the simulation most challenging. In this contribution, we relate the overland flow connectivity in the LTR to the qualitative response of the catchment to precipitation, in order to gain insight into the hierarchy of processes at different scales—from microtopography to (sub)catchment scale. Given the highly nonlinear interactions between topography and precipitation, we argue that understanding the hierarchy of processes will enable us to formulate scaling laws. Using a mechanistic flow model, we carry out an ensemble simulation spanning a range of precipitation intensities and durations. We then compute overland flow connectivity metrics for each simulation, and observe how they develop over time. We present results for several connectivity metrics, namely Shannon entropy, Kullback-Leibler divergence, and coefficient of determination. We show that for mountainous catchments such as the LTR, the coefficient of determination is a significant metric of connectivity, which is maximized during the fully connected state. From the temporal evolution of these metrics, we identify distinct states of overland flow connectivity—disconnected, partially connected, and fully connected—and relate them to the precipitation and outflow time series. Understanding how connectivity develops in the system enables us to relate precipitation characteristics to lag and damping in the catchment outflow hydrograph. It further paves the way to identify and model transient hydrological units in the catchment, which may be used to derive a surrogate model for the overland flow dynamics in the catchment.