Skip over navigation

Inverse River Routing

Speaker: Ming Pan, Postdoctoral Research Associate
Series: EEWR Brown Bag Seminars
Location: Engineering Quad E225
Date/Time: Friday, December 14, 2012, 12:00:00 p.m. - 01:00:00 p.m.


As a major component in terrestrial hydrology, runoff (from either surface or subsurface) has not been a directly measureable variable, at least in a spatially distributed sense. What we do measure is the river streamflow at gauging locations, and by the time the runoff water reaches gauging stations it has gone through a number of processes – over the hillslope and in the river channel. So the quantity that is actually measured, i.e., the streamflow, is an integrated response of runoff water in both time and space. And this creates a discrepancy between the runoff and other components in terrestrial hydrology like precipitation, evapotranspiration, and soil moisture, which are all measured at the same time and location as they occur. Such a discrepancy becomes a problem especially when we need to perform a water budget analysis or propagate information from one variable to another. The ultimate way to close this gap at pixel level is to derive spatially distributed runoff field from gauge measurements, or any other forms of river measurements like satellite altimetry using interferometer.

To derive runoff field from streamflow is essentially a downscaling or disaggregation process in time and space. Note that the forward process, i.e., the runoff-to-streamflow integration, has been very thoroughly studied in hydrology. And the in-channel part is usually referred as river routing. We propose to solve the streamflow-to-runoff problem by inverting a linear routing model, and we refer this as “inverse routing”. To prove the concept of inverse routing, we start with a very simple routing model and a simple example study driven by model generated streamflow time series. We show that the inverse routing is able to distribute streamflow water across the contributing watershed and back in time. Based on the preliminary results, we also discuss the potential use of such approach in crosscorrecting or cross-validating measurements from hydrology mission satellites.