Results of Channel Transmission Loss Studies during Ephemeral Flow Events

Transmission losses along ephemeral channels are a critically important, yet poorly understood, aspect of rainfall-runoff prediction. Losses occur as flow infiltrates channel beds, banks, and floodplains. Two channel transmission loss field experiments were performed on the U.S. Department of Energy/National Nuclear Security Administration's Nevada Test Site: the first was conducted in the ER-5-3 channel, and the second was conducted in the Cambric Ditch. Both studies quantified transmission losses using water discharged from unrelated drilling activities during well development and aquifer pump tests. Discharge measurements at several flumes located along the channels were used to directly measure transmission losses. These measured values were compared with three different analysis methods: Lane's Method for estimating flood magnitude in ephemeral channels; applying a finite element-based flow and transport code that solves governing flow equations in partially saturated media, to estimate infiltration rate from soil temperature data; and using hydraulic gradient and water content in a Darcy's Law approach to calculate one-dimensional flow rates. In the ER-5-3 channel experiment, using the discharge measurements from the flumes, up to 70 percent of the flow was lost in reaches of 1,000 to 2,000 m (3,280 to 6,560 ft). In Cambric Ditch, approximately 40 percent of the initial upstream flow was lost throughout the 1,000 m (3,300 ft) experimental channel. As inflow-outflow data for each reach was known from flume measurements, Lane's Method was used to predict outflow. Only two reaches, ER-5-3 reach 3 and Cambric reach 2, produced parameters for Lane's Method that satisfied the method. In both reaches, Lane's Method very slightly over estimates the measured outflow for each reach in two of the three flow events. For the remaining flow event in each reach, the method slightly under estimates the measured outflow. HYDRUS-2D was used to model measured temperature data from the channels. The model was able to fit both the measured temperature and water content data for ER-5-3 and Cambric channels relatively well. Models for the two experimental channels produced cumulative infiltration estimates ranging from about 3.0 to 740.0 m³/m (32.3 to 7,961.2 ft³/ft), with corresponding infiltration flux, ranging from 0.3 to 2.0 m/d (1.0 to 6.6 ft/d). However, the HYDRUS-2D analyses significantly under or over estimated the measured transmission losses per reach. Darcy's Law calculated one-dimensional flow rates using hydraulic gradient and water content data. Infiltration estimates in the ER-5-3 channel reach 3, using the Darcy analysis, ranged from 7,572.6 to 12,420.1 m³ (267,423.8 to 438,611.7 ft³) — 6 to 8 times greater than the measured losses.