DEC’S Water Resources team is using a two-dimensional (2D) hydrodynamic simulation tool to visualize rainfall-runoff processes and better manage flood risks, especially in urban watersheds.
By Biswajit Mukhopadhyay, Ph.D., P.E., CFM, D. WRW; Nellie Yang, EIT; Haiying Cui, EIT; Mounika Jella, P.E., CFM, ENV SP; and Amy Dziuk, P.E., CFM
Flooding is the most common natural disaster in Texas. The eastern region of the state that parallels the coastal arc from Louisiana to Mexico is particularly vulnerable to disastrous flooding during tropical storms and hurricanes. Though the exact hydrological impacts of climate change cannot be adequately quantified, we can make a qualitative prediction of climate change effects. With the current rate of global warming, heavy precipitation events are very likely to be more frequent, increasing flood risks in urban areas in the coming decades.
1-D Modeling Limitations
Water resources engineers have traditionally used one-dimensional hydraulic modeling of streams and rivers as a flood risk management tool. However, this method suffers from a severe limitation. The tool cannot identify the risks of flooding in overland areas at elevations above the floodplains of the well-defined channels.
Furthermore, the classical approach of riverine flood modeling decouples the rainfall and flood events by determining the runoff hydrographs via modeling the watershed hydrology independently and then using the calculated hydrographs as input boundary conditions in the hydrodynamic calculations of the flood processes. However, the latest versions of The Hydrologic Engineering Center’s River Analysis System (HEC-RAS) offer the platform for rain-on-grid two-dimensional (2D) hydrodynamic simulations where an integrated approach to visualization of rainfall-runoff processes can be used to better manage flood risks, especially in highly urbanized watersheds.
DEC, one of the most experienced water resource management engineering firms in Texas, has developed flood management solutions for public and private clients for decades. DEC’s hydrologists and hydraulic engineers have expertise in flood risk management using 2D hydraulics.
Trailblazing Developments
For example, Dr. Matthew Garcia, one of our team members, in his doctoral work at Rice University pushed the boundaries of what was possible with 2D HEC-RAS by working on applications of the model in surrogate Machine Learning methods. Through his trailblazing, Dr. Garcia developed methods for modularizing large-scale 2D HEC-RAS models, allowing for local applications of regional scale models and the automation to accompany it. The automation, using python, included all aspects of 2D HEC-RAS, from realistic synthetic radar rainfall, to specified geometry and structural inputs, to output exports for high-dimensional calibration methods (calibrating all metrics and gauges simultaneously as shown in the figure).
Our 2D HEC-RAS rain-on-grid model of Whites Bayou watershed, with an area of 143 square miles, in Chambers and Liberty counties (see Figure), reveals that realistic stage hydrograph responses can be simulated when areal effective precipitation is used for 2D HEC-RAS rain-on-grid forcing’s during saturated antecedent soil moisture conditions surface digital elevation models adequately describe overland flow paths. Rain-on-grid modeling is implemented by means of nonlinear partial differential equations of simultaneous balance of mass and momentum in unsteady fluid flow in response to a source or sink flux to account for precipitation, evaporation, and infiltration. The nonlinear equations are discretized using an iterative scheme for the temporal integration and finite element volume for the spatial component. Thus, HEC-RAS has implemented a finite volume method and a sparse parallel linear solver running on a unique processor for which further developments are necessary for large-scale watersheds.
