Academic Affiliation: University of Idaho
Northern California was always a well-suited place for Amy’s development into a geologist. The majestic valleys, centuries old redwood trees, and enormous mountain chains paved the path for her natural science passion. Now, Amy is a senior at the University of Idaho and studying flash floods in semi-arid environments for her RESESS research. Her project takes place in West Bijou Creek in east-central Colorado, where, armed with tipping-bucket rain gauges and sonic ranging sensors, she analyzes antecedent conditions for flash floods.
2013- Determining the recipe for a Flash Flood: An east-central Colorado analysis of high-resolution rainfall, flow, and infiltration in a semi-arid grassland
Flash floods commonly cause rapid gully erosion, creating headwalls that erode into previously stable surfaces, thereby reducing arable land and releasing sediment that can contaminate water supplies. The flash floods associated with gully networks present a hazard because they occur with little warning. In this study, I analyze the necessary and sufficient conditions for flash flood generation in grassland gully networks. An ideal example of a dryland zone impacted by gully erosion is West Bijou Creek, located in the Denver Basin of east-central Colorado. To determine the conditions that promote flash flood generation, I analyzed six years of data from a monitoring site in a 7.9 ha gully basin. The monitoring station consists of a tipping-bucket rain gauge and a sonic water-level sensor positioned above the channel. During rainstorms, rainfall and water level were recorded at 20-second intervals, providing a high-resolution picture of rainfall intensity and flood hydrographs. Analysis of rainfall and stage records revealed a clear intensity-duration threshold for flood generation. I found that rainstorms with a peak 15-minute intensity greater than 30 mm/hour always produced flash floods. Less intense but longer-lasting events were also capable of producing flash floods; for example, storms lasting 5 hours or more produced flash floods when the 15-minute intensity exceeded 5 mm/hour. In addition, to estimate the soil infiltration capacity, I conducted a sprinkler experiment on two 0.6 m2 – one on a grass-mantled surface and one on a bare surface. The sprinkler experiments revealed a dramatic contrast in soil infiltration capacity between bare and vegetated surfaces. The vegetated plot was able to sustain infiltration rates as high as 43 mm/hour, whereas the bare plot exhibited an infiltration rate of only about 6 mm/hr. These findings imply that high-intensity rainstorms associated with summertime moist convective systems drive gully incision, and that a self-enhancing feedback may exist in which initial incision creates steep and relatively bare slopes that tend to generate more runoff, leading to more aggressive gully incision. Therefore, gullying should be more active generally in monsoonal climates that provide episodic, intense rainfall.
2014- Mechanisms of flash-flood generation in a gullied high-plains grassland: evidence for partial contributing area runoff
Flash floods commonly cause rapid gully erosion, creating headwalls that erode into previously stable surfaces, thereby reducing arable land and releasing sediment that can contaminate water supplies. In semi-arid landscapes, gully erosion tends to be driven by flash floods. Here, we study the mechanisms for flash-flood generation, seeking to answer two questions: (1) how spatially variable is runoff production, and (2) what combination of rainfall intensity and duration is required to produce runoff? To answer these questions, we combine field data from a study site on the Colorado High Plains, USA, with numerical modeling. The site is characterized by patchy, dryland shrub vegetation dispersed throughout the otherwise bare slopes and gullies. Analysis of six years of rainfall and runoff data indicate that flash flood generation requires a 15-minute intensity of approximately 38 mm/hr. Sprinkler experiments on isolated bare and vegetated plots revealed a large contrast between infiltration capacities: bare areas can produce runoff when the rainfall exceeds 10-15 mm/hr, whereas vegetated areas permit infiltration of at least 45 mm/hr during relatively brief, intense events. These findings imply that high-intensity rainstorms associated with summertime moist convective systems drive gully incision. They also suggest that a self-enhancing feedback may exist in which initial incision creates steep and relatively bare slopes that tend to generate more runoff, leading to more aggressive gully incision.