Flooding in ephemeral streams: incorporating transmission losses

Authors: PACHECO-GUERRERO, A. - Universidad Autónoma De Guerrero; Goodrich, David - Dave; GONZÁLEZ-TRINIDAD, J. - Universidad Autónoma De Guerrero; JÚNEZ-FERREIRA, H.E. - Universidad Autónoma De Guerrero; BAUTISTA-CAPETILLO, C.F. - Universidad Autónoma De Guerrero

Submitted to: Journal of Maps
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 3/9/2017
Publication Date: 4/10/2017
Citation: Pacheco-Guerrero, A., Goodrich, D.C., González-Trinidad, J., Júnez-Ferreira, H., Bautista-Capetillo, C. 2017. Flooding in ephemeral streams: incorporating transmission losses. Journal of Maps. 13(2):350-357.

Interpretive Summary: Stream flow in semiarid lands commonly occurs as a form of flash floods in dry ephemeral stream beds. While these flood events may be of short duration they still cause considerable damage to people and property. The goal of this research was to predict the flood inundation levels around an ephemeral stream where infiltration losses through the dry channel bed can be large. The research was conducted with data collected from two runoff measuring stations in the USDA-ARS Walnut Gulch Experimental Watershed (WGEW) near Tombstone, Arizona for the storm and runoff event from August 27, 1982. The method developed employs a two-dimensional hydraulic model with dynamic, interactive channel infiltration. Two models formulations were developed; the first does not incorporate channel transmission infiltration losses; and the second one was developed to take into account transmission. The error in volume and peak runoff rate between the observed and simulated model data ranges was from -4.5% to 34.4% for runoff volume and -16.4% to 9.6% peak runoff rate. There are important differences between the two flood inundation mappings due to the way the hydraulic model treats the input hydrograph in each channel sections. There are differences in depth up to approximately 0.68 and 0 meters as minimum. Depending on the flood plain topography a difference in flood stage of 0.68 meters can result in a substantial change in the area flooded. The importance of using models with a more realistic treatment of channel transmission infiltration losses can provide an improved estimate for flood hazard mapping. The biggest difference between using and not using of these features is reflected in the depth of flooding for the topography of the case study.

Technical Abstract: Stream flow in semiarid lands commonly occurs as a form of flash floods in dry ephemeral stream beds. The goal of this research is to couple hydrological and hydraulic models treats channel transmission losses and test the methodology in the USDA-ARS Walnut Gulch Experimental Watershed (WGEW). For hydraulic modeling the reach which is between flume 2 and flume 1 in WGEW was tested. The method developed employs a 2D hydraulic model with dynamic, interactive channel infiltration; two hydraulics models were set up, the first does not incorporate channel transmission and the second one was developed to take into account several hydrographs with transmission losses as boundary conditions. The rainfall event occurred in August 27, 1982 was modeled. The error in volume and peak runoff rate between the observed and simulated data ranges was from -4.5% to 34.4% for runoff volume and -16.4% to 9.6% peak runoff rate. There are important differences between the two flood mappings due to the way the hydraulic model treats the input (upper) hydrograph in each sub-reach. There are differences in depth up to approximately 0.68 m and 0 m as minimum. The importance of using models with dynamic treatment of transmission losses is the ability to provide an improved estimate for flood hazard mapping. The biggest difference between using and not using of these features is reflected in the depth of flooding for the topography of the case study. Greater impacts could be expected in less incised channel networks with flatter, broader flood plains.