Watershed Analysis and Detention Basin Flood Routing
of Rudd Creek Drainage
Prepared By:
Cristina Nelson
Prepared For:
Dr. Christopher Neale
Term Project Report
CEE 6440
December 6, 2007
Table of Contents
Page
1.0 Introduction .1
2.0 GIS Base Map Construction ............2
3.0 Watershed Delineation . ....3
4.0 Peak Canyon Discharge . ......3
4.1 Rainfall Intensity . .4
4.2 Runoff Coefficient . ..5
5.0 Detention Basin Hydraulic Modeling ..6
6.0 Outflow Path Analysis .........7
7.0 Results and Recommendations 8
8.0 Discussion and Conclusion ..9
References ..10
List of Figures
Page
Figure 1: East Farmington Bench Near Rudd Creek ..1
Figure 2: Hill Shade of the Reconditioned DEM to Conform to Existing Drainage Streets ..2
Figure 3: Rudd Creek Delineated Catchments and Streams ...3
Figure 4: IDF Curve for Farmington, Utah 4
Figure 5: Slope Class Histogram . 5
Figure 6: Influent hydrograph and Flood Routing Model Generated Outflow hydrograph ...6
Figure 7: Flow Path Tracing from Basin Outlet .7
Figure 8: FEMA Flood Map for Farmington Creek in Undeveloped Land Parcel 8
1.0
Introduction
The Rudd Creek Drainage area is located in the
Wasatch Mountains, East of Farmington City, Utah. The Rudd Creek canyon outlet is in a
residential area on the East bench of
Figure 1: East Farmington Bench near Rudd Creek
This report details a study of the Rudd Creek
watershed and addresses the concerns of
2.0
GIS Base Map
Construction
A base map of the area was created with a 5 meter resolution Digital Elevation Model (DEM) and 1 ft resolution Ortho-photographic images (AGRC, 2007). The 5 meter resolution of the DEM was too weak to accurately depict the micro relief of the streets that were modified after 1983 to channel Rudd Creek floods to Farmington Creek. In order to account for these important drainage features, the DEM was reconditioned with AGREE DEM software to burn in the known street cross-sections where the existing berms and retailing walls are located. A hill shade surface depiction of the reconditioned DEM as shown below in figure two, depicts the drainage paths created by the channeled streets.
Figure 2: Hill Shade of the Reconditioned DEM to Conform to Existing Drainage Streets
A terrain analysis using the ArcHrdo 9 tools was then conducted using the reconditioned DEM to map the streams and catchments. The DEM was first used to calculate a flow direction grid of the area. This grid was used to calculate a flow accumulation grid. The flow accumulation grid was used to define streams using an 80,000 ft2 stream definition threshold. This threshold value was selected because the streams that were defined closely correlated with stream beds that could be seen in the ortho-photographic images of the area. The streams were then segmented and catchments were defined for each stream segment. The streams and catchments were then converted from raster data to vector lines and polygons to be visualized on the base map as shown in figure three.
Figure 3: Rudd Creek Delineated Catchments and Streams
3.0
Watershed
Delineation
The watershed boundary defining the area draining to the Rudd Creek detention basin was generated by using a batchpoint watershed delineation of the base map. The attribute table statistics were used to find the area of the watershed. Network analyst tools were then used determine the length and average slope of the longest drainage path within the watershed.
4.0
Rudd Creek has no base flow, thus the discharge is a function of storm runoff events only. The 100 year storm peak discharge for Rudd Creek Canyon was calculated using the Rational Method as found in Equation 1 (Kaluarachchi, 2007).
Q = AIC Rational Method (1)
Where A is the drainage area, I is the rainfall intensity and C is the runoff coefficient. The drainage area was found through the watershed delineation process.
4.1
Rainfall
Intensity
The intensity of the 100 year storm was determined from Intensity-Duration-Frequency (IDF) curves provided by the Farmington City Engineer as found below in figure four (CRS, 2007).
Figure 4: IDF Curve
for
The duration of the storm was set equal to the time of concentration for the watershed. The time of concentration (Tc) was calculated as the sum of travel times (Tt) for overland flow and channel flow for the longest drainage path in the watershed. The overland flow travel time was calculated from Equation 2 (Zomoradi, 2005).
Overland
Flow Travel Time (2)
Where P24 is the rainfall intensity for a 24 hour duration 100 year frequency, n is the estimated Mannings roughness coefficient, L is the estimated length of sheet flow, S is the average slope. The n value was estimated from tables for sparsely vegetated wooded areas (Rahmeyer, 2007). The average velocity for channel flow was calculated from the Mannings equation as found below (Finnemore, 2002)
Mannings
Equation (3)
Where V is flow velocity, n is the Mannings roughness coefficient, A is the average channel area, P is the average wetted perimeter and s is the channel slope. The values of A and P were estimated based on field observations. The n value was estimated from tables for mountain streams (Rahmeyer, 2007). The channel slope was found with GIS network analyst tools. The channel flow travel time was then found by dividing the length of the longest drainage path by the channel velocity. The overland flow and channel flow travel times were then summed to yield the time of concentration. The rainfall intensity was then determined from the IDF curve for the 100 year frequency storm.
4.2
Runoff
Coefficient
Runoff Coefficients (C) are usually found from published land use tables. These C values are only valid for slopes less than 10%. This approach does not work well for mountain watersheds like Rudd Creek. As the slope of a surface increases, the C value is increasingly a function of slope. The GIS spatial analysis slope function was used to determine the slope of every point in the watershed. The statistics function was used to create a histogram of the areas of each slope class within the watershed as show in figure 5.
Figure 5: Slope Class Histogram
This histogram was used to determine that the weighted average slope of the watershed is 42%. The runoff coefficient for the Rudd Creek watershed was then found by applying an equation for determining rational coefficients in steeply sloped watersheds as a function of watershed slope found below in Equation 4 (Cutter, 2007).
Equation (4)
Where Co is the runoff coefficient based of land use tables, n is the estimated Mannings roughness coefficient. Roughness was again estimated for a sparsely wooded area, S is the average watershed slope.
5.0
A basin flood routing model of the Rudd Creek detention basin was created using the step-method in a spread sheet in order to predict the performance of the basin during the 100 year flood event. With the peak discharge calculated for the watershed, an influent hydrograph for the basin was estimated assuming linear flow increases up to the time of concentration, that the storm would produce the peak discharge for the storm duration and also that the discharge out of the canyon would gradually taper to zero after the full storm duration. With the area and storage capacity of the basin known, a relationship of depth as a function of water storage was derived. The outlet structures of the basin were then modeled as a function of the water depth. The low level outlet pipe was modeled as full pipe flow, and the emergency overflow spill way was modeled as a broad crested weir with the known spillway dimensions (Dames, 1984). Thus the flood routing model produced a basin outflow hydrograph based on the influent hydrograph and the resultant change in water storage and corresponding depth within the basin. The detention basin inflow hydrograph and the resultant flood routing outflow hydrograph are shown below in figure 6.
Figure 6: Influent hydrograph and Flood Routing Model Generated Outflow hydrograph.
The peak discharge on the outflow hydrograph was
compared to the known capacity of the basins spillway. This comparison was
used to determine if the basin would be overtopped in the 100 year flood.
6.0
Outflow Path
Analysis
In order to determine the effectiveness of the
channeled streets to direct flood flows, a flow path tracing analysis was
conducted at the basin spillway outlet. The flow path tracing tool in the Arc
Hydro GIS extension uses the flow direction grid generated in the previous
terrain analysis process. The flow path
tracing for water flowing out of the
Figure 7: Flow Path Tracing from Basin Outlet
From the figure it can be seen that the berms built after the 1983 flood do not fully protect the undeveloped land parcels from a 100 year flood. Berms should be built along the road adjacent to the undeveloped land. These berms should be similar in size to the berms along the roads of the existing residential areas. These Modifications would redirect flow into the streets and down safely into Farmington Creek. The flow path tracing analysis is only a simplistic analysis. A full flood map should be generated for Rudd Creek using a DEM of greater resolution and more surveyed cross sections than are currently available for this project. Without a complete flood map, the certainty of the 100 year flood protection from Rudd Creek for the undeveloped land cannot be reasonably assumed. As suggested by classmates in the presentation of this project, the flood map levels for Farmington Creek were examined and found to have very little impact on the undeveloped land. The cross-section of Farmington Creek in the area of the undeveloped land is able to contain the 100 year flood as shown in figure 8. The FEMA 100 year flood plain is marked in dark grey (FEMA, 2007). Zone X in the figure is the 500 year flood plain and is not a restricting factor to development.
Figure 8: FEMA Flood Map for Farmington Creek in Undeveloped Land Parcel
7.0
Results and
Recommendations
The terrain analysis and watershed delineation process resulted in a drainage area of 0.77 square miles. The network analysis revealed a longest drainage path of 13,287 feet at an average slope of 46.6%. The sum of calculated travel times for overland and channel flow resulted in a time of concentration of 15 minutes. Using a storm duration of 15 minutes, the 100 year rainfall intensity was found to be 3.68 in/hr from the IDF curve. The statistical slope analysis determined an average watershed slope of 42%. The resultant composite runoff coefficient for the watershed was calculated to be 0.67 based on the average watershed slope. The rational method calculations returned a peak creek discharge of 1210 cfs. The flood routing model predicted a total peak basin outflow of 930 cfs, with a maximum pipe flow of 50 cfs, and a peak spillway outflow of 880 cfs. Thus the spillway capacity of 1000 cfs was not exceeded, indicating that the detention basin would not overtop during a 100 year flood event. Flow path tracing indicated that the undeveloped land may be vulnerable to the 880 cfs peak basin spillway outflow. The construction of small berms and retaining walls are recommended along the eastern edge of the undeveloped land parcel to contain flood flows to the streets. These drainage structures should be sized to provide similar street cross-sections as constructed in residential areas after the 1983 floods.
8.0
Discussion
and Conclusions
In summary,
References
Automated Geographic Reference Center of Utah (AGRC), (2007), State Geographic Information Database, Raster GIS Downloads, 5m DEM and Aerial Images, (accessed on 10/15/2007) http://agrc.its.state.ut.us.
Caldwell Richards Sorenson (CRS), 2007, Farmington City Rainfall Curves, Farmington City Engineers File #06106, Salt Lake City, Utah
Cutter, Gillian, McCuen, Richard, (2007), Rational Coefficients for Steeply Sloped Watersheds, Journal of Irrigation and Drainage Engineers, March/April issue 2007, p188-191
Dames & Moore Consultation Services, (1984), Davis County Flood Control Report, Proposed Rudd Creek Debris Basin, Farmington City Engineers File # 06106.
Federal Emergency Management Agency, (FEMA), (2007) Map Service Center Catalog, Davis County, (accessed 12/5/2007), http://msc.fema.gov
Finnemore, E. John & Fanzini, (2002), Joseph, B., Fluid Mechanics with Engineering Applications, Tenth Ed., McGraw-Hill Publ. Comp., New York, NY
Kaluarachchi,
Jagath J, (2007), CEE 3430 Engineering Hydrology, Class Notes (accessed on
11/19/2007) http://www.engineering.usu.edu/uwrl/CEE3430/3430.htm
Rahmeyer, William, (2007), Flow Resistance for Utah Floodplains, (accessed on 11/22/2007), http://www.engineering.usu.edu/classes/cee/6500/flow_resistance-revised%20rev3.pdf
Utah Division of Comprehensive Emergency Management (UDCEM), (1984), Hazard Mitigation Plan, Utah Department of Public Safety, Salt Lake City, Utah 84114
Zomoradi, Kaveh, (2005), Revising the NRCS Sheet Flow Travel Time Equation, AWRA Annual Water Resources Conference, Seattle, Washington