ABSTRACT:

Urbanization has profound impacts on stream morphology and streamflow patterns. Predicting these impacts in semi-arid areas is limited by the lack of information on pre-development streamflow, as non-perennial streams are not well monitored in semi-arid areas. This research focuses on the ephemeral stream West Stroh Gulch, located south of Denver, Colorado, U.S. This grassland watershed was used for grazing during the pre-development study period of June 2020 – August 2023. During this time, time-lapse photography at five locations along the stream network along with radar-rainfall data were used to determine which storm events did, or did not, lead to a streamflow response. Out of 115 recorded storm events, 32 led to streamflow. A classification tree indicated that streamflow was due to storm events with higher rainfall intensities (maximum 30-minute intensity above 4.3 mm/hr and 60-minute intensity above 6.2 mm/hr) and 7-day antecedent rainfall above 20.3 mm. A two-dimensional hydrodynamic model (SRH-2D) enabled comparison of the impacts of predicted flows through a reach of interest. The model used a digital elevation model developed using structure-from-motion techniques and drone aerial imagery. Storm Water Management Model (SWMM) generated peak flows were used to simulate impacts of different sized storms (Water Quality Capture Event (WQE), 2-, 5-, 10-, 50-, and 100-year storms) and development and stormwater management scenarios (pre-development, developed undetained, and developed with distributed detention). Results predicted that the smallest but most frequent storms (WQE and 2-year storms) would cause increased flow and boundary shear stress post-development. For larger storms, distributed detention facilities reduced post-development flows below pre-development and were well below the undetained development, which consistently had the largest flow and potential for sediment mobilization. The findings contribute to understanding of ephemeral streamflow in rangelands and streamflow regime alterations as drivers of sediment mobilization with urban development.

ABSTRACT:

Precipitation in Colorado is highly variable in space and time, and surface water, which supplies most of the state's water use, faces growing gaps between supply and demand, with municipal shortfalls identified as a major risk in the Colorado Water Plan 2023. Rainwater harvesting coupled with water-wise landscaping offers one conservation strategy to curtail residential irrigation demand. Extending the Denver-focused analysis of Gilliom et al. (2020) to the whole state, this study assesses the variability of rainwater harvesting potential across Colorado's 60 main urban areas under future climate conditions. A daily water balance model using a yield-before-spillage release rule was run for April to October across 9 precipitation scenarios (PRISM historical data plus eight GCM projections under RCP 4.5 and 8.5, downscaled via MACAv2-METDATA), 6 housing configurations, 3 storage capacities (110, 300, and 500 gallons), and 3 irrigation demands (1.5, 2.5, and 3.5 mm/day). System performance was evaluated using the annual rainwater harvesting yield, the fraction of irrigation demand supplied by harvested rainwater. Yield increased with smaller irrigated area, greater storage capacity, and lower irrigation demand, with the irrigated-area effect outweighing housing density. Across urban areas, mean annual yield was highest in Vail (28%) and lowest in Trinidad (18%), with the highest-yielding locations concentrated in the high-elevation Interstate 70 corridor and the lowest in southern and western Colorado. Under favorable conditions rainwater harvesting yielded 18% to 100% (mean 62%), while under Colorado's current 110-gallon legal storage limit it yielded 10% to 65% (mean 32%), indicating a promising supplemental source for offsetting outdoor municipal water use.