Cottonwood Creek Drainage Basin Planning Study
Matrix was hired by the City of Colorado Springs to update the Cottonwood Creek Drainage Basin Planning Study (DBPS). This watershed is approximately 85% developed and has several existing stormwater-related issues. The main stem of Cottonwood Creek is approximately 12 miles long and has a drainage area of approximately 19 square miles at its confluence with Monument Creek. Previous studies have been completed with a focus on flood attenuation and conveyance, which has left the watershed in a state of instability. Current concerns within the watershed consist of the release of higher than historic peak flows downstream, channel instability, habitat destruction, and water quality issues.
The process for completing the Cottonwood Creek DBPS was unique in that everything from the model inputs to the calibration of the output was physically based. In this DBPS four unique hydrologic modeling techniques were applied to create a hydrologic model that is much more accurate than the industry standard.
First, detailed survey was performed at four key cross sections on the Cottonwood Creek main stem as well as an additional 20 rough tape surveyed cross sections which were performed along the Cottonwood Creek main stem and tributaries. Additionally, regression analysis of USGS cross section and current meter data was completed for flow vs. cross sectional area, depth, width, and velocity to aid in identifying channel features in the field. These 24 cross sections allowed for all reach parameter assignments to be physically based which lead to an accurate representation of channel storage during routing.
Second, impervious cover data was obtained from the same dataset collected by the Colorado Springs Stormwater Enterprise for billing purposes. This dataset represented all roofs, driveways, sidewalks, and streets within the basin. Within an ArcMap® GIS environment, discrete grid cell combinations of hydrologic soil group and underlying land uses were developed. For existing conditions, runoff CNs were determined for each unique soil/land use combination based on the presence or absence of impervious cover in the grid cell. Future CNs were developed by applying the CNs calculated for existing conditions to areas of similar use according to available master planning data. This application allowed for a physically based CN assignment as opposed to assuming average impervious conditions.
Third, Manning’s n values were calculated at four USGS gage locations using flow frequency analysis for select recurrence intervals and the regression equations for channel features described above. These four physically-based Manning’s n values provided guidance in the assignment of Manning’s n values for all other reaches.
Fourth, flow frequency analysis was performed at three USGS gage locations to determine the two-year recurrence interval (mean) of the peak flows. The existing two-year modeled flow was calibrated to three USGS gages by adjusting the initial abstraction value for each subbasin. Multiple scenarios were modeled and the optimal initial abstraction value was selected by using the sum of least squares to compare the modeled flows versus the two-year peak flows at the three USGS gage locations calculated from the flow frequency analysis. This method of calibration allowed for the hydrologic model to accurately represent the physical response the watershed has to rainfall.
The goal of this DBPS was to create a sustainable watershed plan that: