Catchment Tomography - A Catchment Tomography Approach for Estimating Spatially Distributed Manning’s coefficient

In catchment hydrology, distributed parameter identification has been declared infeasible because of non-uniqueness and ill-posedness of the inverse problem. This is due to the ubiquitous issue of data scarcity, which arises when one seeks to invert, for example, spatial heterogeneity in the hydraulic conductivity from integrated observations, such as stream discharge measurements and in-situ soil moisture time series.

In this study, we develop the novel paradigm of Catchment Tomography, where the principle of imaging structures by decomposing a measured integrated signal via a moving transmitter-receiver concept is applied to the catchment hydrologic response to rainfall. Here, precipitation along low pressure fronts and localized, short-lived rain events due to convective precipitation represents the moving transmitter, while stream gauges serve as receivers. Using long-term rain radar information for the transmitter in connection with discharge time series, we will be able to invert the integrated signal using classical inverse theory and derive high-resolution parameter distributions. As a matter of fact, this approach is very similar to data assimilation including rain radar information and will result in "intelligent" catchment models, which improve automatically as more orthogonal forcing information from radar observations is added to the system. The feasibility of the approach will be demonstrated utilizing a hypothetical catchment model in connection with real-world rain radar information from the twin, dual-polarized rain radar setup provided by the Meteorological Institute of University of Bonn and the Forschungszentrum Jülich. In ensuing steps, the approach will be applied to a real-world catchment in the region.