This project received an MICDE Catalyst Grant in summer 2018.


The problem of flood prediction incurs multiple uncertainties (a) and is of high computational complexity (b). New remote sensing data from the CYGNSS mission (c) can inform complex physically-based simulations (d) but in order to achieve feasibility of real-time solutions and uncertainty quantification, novel approaches are required. This project will develop reduced-order modeling tools (e) as innovative, parsimonious representation of rigorous hydrologic and hydrodynamic model formulations to efficiently obtain probability density distributions of one or many quantities of interest (f).

Despite substantial advances in civil engineering in the modern era, urbanized areas are vulnerable to extreme flooding, since urban drainage and stormwater management infrastructure are generally designed for flows of 10-20 year return periods. Flooding in densely populated areas has remained the costliest natural hazard of all weather-related events in terms of fatalities and material costs.

The emerging need is both to understand how heterogeneity of urban environments impact extreme floods and engineer a comprehensive modeling capacity relevant to decision making in the critical times of flooding. Understanding and predicting floods across a range of space-time scales at the relevant level of detail and with uncertainty assessment remains a poorly addressed challenge. None of the existing flood operational tools communicate uncertainty quantification.

The science of flood modeling has been steadily changing from a data-scarce to data-rich environment because of the growing availability of geospatial, remote sensing datasets, as well as real-time sensor data for hydrologic systems. Finer spatial resolutions of relevant data sets on topography, building layout, and civil infrastructure describe heterogeneity of the real world and have the potential to advance the knowledge on urban hydrology and hydraulics, but also come at an extreme computational cost. Both code optimization and alternative, lower computational cost (i.e., ‘reduced-order’) solutions are needed to achieve simulation feasibility and uncertainty quantification at larger spatial scales. Making these developments sufficiently general, location/case independent will be crucial to enable real-time flood forecasting and informed decision-making.

This project’s research team will enhance urban flood monitoring and prediction using NASA Cyclone Global Navigation Satellite System (CYGNSS) data, taking advantage of state-of-the-science uncertainty quantification tools in a proof-of-concept urban flooding problem of high complexity.

Principal Investigators

Valeriy Ivanov, Professor of Civil and Environmental Engineering, U-M

Nick Katopodes, Professor Emeritus of Civil and Environmental Engineering, U-M

Khachik Sargsyan, Principal Member of Technical Staff, Sandia National Labs