Little Bear River Floodplain Storm Routing and Damage Analysis
By Joe Hawkes
Term Paper Geographical Information Systems in Water Resources 2009 Utah State University
Created For
Dr. David G. Tarboton, Dr. David R. Maidment, and Dr. Ayse Irmak
December 4, 2009
Contents
Proposal
Data
Area Calculation
Land Cover
Snow Contribution
Runoff Coefficients
Storm and Hydrograph
Baseflow
Practicality of Flows
Conclusion
References
Proposal
Prime real estate in Cache Valley is often found at the mouth of canyons and in floodplains. Recent development in the southern part of Cache Valley lies in the floodplain of the Little Bear River between Porcupine Reservoir and Hyrum Reservoir. The probabilities of flooding and the levels associated with it will be analyzed in this report. GIS tools will be used to create a model upon which a 100 year one hour storm can be simulated and flood hydrographs created. River flows during most of the year are relatively low, and a 100 year storm for several hours would not have the effect that is seen yearly during the runoff. However, a 100-year storm for just one hour at the peak runoff times would be the maximum flow seen in the river. Data from United States Geological Survey (USGS), National Oceanic and Atmospheric Administration (NOAA), ArcGIS Online, Google Earth, and information about the reservoir operations of Porcupine Dam gathered from the owners and operators of the reservoir. This goal of this report is provide understandable data about potential flooding near the Little Bear River. This report will not assess reservoir failure, but will take into account the possible flows from the reservoir during flood situations.
Data
Data was first found from several sources in order to create a map of the region in Arcmap. The USGS was the source for the majority of the information that was implemented in the project. The National Elevation Dataset (NED) was downloaded from the USGS website, as well as the National Hydrography Dataset (NHD), and the stream gage data for USGS stream gage 10105900 Little Bear River at Paradise, Utah. The National Land Cover Dataset (NLCD) was also downloaded from the USGS.
Google Earth was used throughout the project, but only qualitatively, not quantitatively.
Visual and reference images were implemented in the Little Bear GIS map from ArcGIS Online. The World Topographic Map and World Imagery were the maps added to the project map.
SNOTEL data was downloaded for the region from station 11H25S, the Little Bear Pillow. The data contained temperature for the last 20 years, but did not have any snow depth measurements which were needed to calculate the potential moisture holding capabilities of the snowpack.
Storm data for the area was found from the resource available from the NOAA. Point precipitation frequencies used in the project correspond to those estimations at 41.610 N 111.847 W and 4665 feet above mean sea level.
The NED was added to the map and to get a reference the World Topographic Map was also added. The NHD flowlines, subbasins, watersheds, and subwatersheds for the region 16 were also added. The amount of information was cumbersome however, so the amount of data was cut down by exporting selected data that was within HUC 8 16010203, the Little Bear Subbasin. The NHD flowlines were burned into the digital elevation model (DEM), and then yet another elevation raster was created which has the elevation sinks filled.
Figure 1. Digital Elevation Model of HUC 10 1601020301
The flow accumulation raster was used to define the streams in the subbasin, and the defined streams were segmented and then used to create catchment rasters. The catchments were converted to polygons. As can be seen, the catchments and defined streams match well with the subwatersheds and NHD flowlines respectively.
Figure 2. Random Screenshot Showing Flowlines, Flow Accumulation (facc), Watershed, and Catchments
Area Calculation
Cell accumulation values were found using the attribute table for the flow accumulation raster at the mouth of Hyrum Dam, and multiplied by the cell area found in the flow accumulation raster properties. This gives a total catchment area of 622 million square meters, 622 square kilometers, and 243 square miles. The cell accumulation was also found at the mouth of Porcupine using the same method. The drainage area above porcupine reservoir is approximately 213 square kilometers or 83 square miles. This gives a net area of the Hyrum Dam drainage excluding the Porcupine Dam drainage equal to 410 square kilometers or 158 square miles.
Land Cover
Land cover classes were found using the added NLCD tagged image formatted layer. The layer was exported and added again to the map as a grid. Land coverage percentages were then found using the statistics in the attribute table for the selected region.
Table 1. Land Cover Classes and Percent Coverage of Hyrum Dam Drainage Basin
|
Value |
Percent Coverage |
Group |
|
11 |
0.26% |
Water |
|
21 |
1.68% |
Developed |
|
22 |
0.99% |
Developed |
|
23 |
0.24% |
Developed |
|
24 |
0.04% |
Developed |
|
31 |
0.26% |
Range |
|
41 |
38.51% |
Forest |
|
42 |
13.63% |
Forest |
|
43 |
0.76% |
Forest |
|
52 |
34.66% |
Range |
|
71 |
0.16% |
Range |
|
81 |
5.33% |
Agriculture |
|
82 |
3.17% |
Agriculture |
|
90 |
0.29% |
Water |
|
95 |
0.02% |
Water |
The land cover classes were assimilated into five different groups to simplify our runoff analysis. The groups consist of water bodies, developed land, open range land, agricultural land, and forested land. As seen in table 2, the vast majority of the land in the drainage basin is forested. Only three percent is developed, so we can assume that our runoff would be significantly less than the actual precipitation.
Table 2. Group Percent Coverage of Hyrum Dam Drainage Basin
|
Group |
Total Coverage |
|
Water |
0.57% |
|
Developed |
2.95% |
|
Range |
35.08% |
|
Forest |
52.90% |
|
Agriculture |
8.50% |
Snow Contribution
As was mentioned earlier, the project is focusing on a 100 year one hour storm during the peak runoff season. At the peak runoff season snowmelt is contributing to the streamflow, so any storm would have rain on snow effects, especially at the higher elevations. It is also common to have a snowstorm, followed by a rainstorm due to the fluctuating temperatures during the spring season. This requires assumptions and allows for several different scenarios for our storm.
Snow could cover the entire drainage basin, and it could all contribute to the runoff, or snow depths could vary throughout the basin and some or all of the snow, if deep enough, could retain some of the precipitation instead of contributing to the runoff (Linsley, 1979). Snow could partially cover the drainage basin due to different elevations and slopes, and either of these cases could either contribute to the runoff or retain some of the precipitation. In order to forecast a damaging event we need to assume that the conditions are ripe for the maximum runoff possible. Because of this we will assume that a recent snowstorm deposited just enough snow to melt and contribute to the runoff during the whole one hour storm.
Snow melt in relies on several variables; the solar radiation which depends on the albedo, the air temperature, wind speed, and vapor pressure. During rain on snow events however, the rain temperature is yet another factor in calculating snow melt. The rain temperature is not as effective agent of snow melt as the warm air, strong winds and high humidity that often accompany storms. A simple calorimetric equation to estimate snowmelt due to rainfall is as follows.
Equation
1
Where MS is the snowmelt in inches, P is the precipitation in inches, T is the rain temperature in degrees Fahrenheit. This means that we must use the temperature data from april and may. Although this is not necessarily the rain temperature, the average for the second half of April, and the first half of May is 37 degrees Fahrenheit. Assuming rain warmer than the average 24 hour temperature would be impractical. Assuming the rain is equal to the average temp gives us a contribution from the snow melt that is practical. Using Equation 1 gives an additional snowmelt of 0.05 inches, which seems too low.
Runoff Coefficients
Land cover is valuable to the project because of the reliance of runoff coefficients on land cover. There are two methods when using runoff coeffients. The rational method uses the following equation.
Qp = CiA Equation 2
Where Qp is the peak flow of the runoff hydrograph, C is the runoff coefficient which varies with land use, i is the intensity of rainfall for the concentration time, and A is the area of the watershed.
The other method is the curve number method, which incorporates the curve number which is dependent upon the land use and the soil type, to find the initial abstraction, the potential maximum retention, and the runoff. The equations are as follows.
Equation
3
Equation
4
Equation
5
S is the potential maximum retention, CN is the curve number, Q is the runoff per unit area, P is the precipitation, and Ia is the initial abstraction (NCEES, 2008). The curve numbers can be seen in table 3.
Table 3. CN Values (NRCS, 1972)
|
Runoff Curve Numbers* |
|||||
|
|
Hydrologic Soil Type |
||||
|
Land Use, Crop, and Management |
A |
B |
C |
D |
|
|
CULTIVATED, with crop rotations |
|
||||
|
Row Crops, poor management |
72 |
81 |
88 |
91 |
|
|
Row Crops, conservation mgmt |
65 |
75 |
82 |
86 |
|
|
Small Grains, poor management |
65 |
76 |
84 |
88 |
|
|
Small Grains, conservation mgmt |
61 |
73 |
81 |
84 |
|
|
Meadow |
55 |
69 |
78 |
83 |
|
|
PASTURE, permanent w/moderate grazing |
39 |
61 |
74 |
80 |
|
|
Water, Wetlands |
0 |
0 |
0 |
0 |
|
|
WOODS, permanent, mature, no grazing |
25 |
55 |
70 |
77 |
|
|
ROADS, hard surfaces and roof areas |
74 |
84 |
90 |
92 |
|
The region has poor to moderately well-drained, moderately fine to fine texture soil which is characteristic of soil C (NRCS 1972).
Storm and Hydrograph
The point precipitation estimates for the basin were estimated from the Utah 41.610 N 111.847 W 4665 feet site and assumed to be constant over the basin. The 100 year one hour storm estimate is for 1.54 inches of precipitation. With the snowmelt we can assume that the runoff is equal to 1.59 inches. With a basin area equal to 158 square miles we can calculate the volume of runoff using the CN values.
Table 4. Runoff Values for C Soil
|
Runoff |
|
|
Land Use, Crop, and Management |
ft3 |
|
Small Grains, conservation mgmt |
11366421.65 |
|
PASTURE, permanent w/moderate grazing |
25129436.96 |
|
Water, Wetlands |
3222101.84 |
|
WOODS, permanent, mature, no grazing |
20963467.33 |
|
ROADS, hard surfaces and roof areas |
82022.69 |
|
Total |
60763450 |
Using the total value we can then solve for a simple hydrograph. The time to peak of the hydrograph is five times the duration of the storm, so the time to peak is 5 hours later (Fang, 2005). Using the NRCS unitless hydrograph we can then solve for the various flows by summating the flow rates over their time period to obtain the runoff volume. By using a simple numerical solver we can set the flow peak so the runoff volume equals that found in Table 4. The runoff volume back calculations give a peak flow of 2500 cfs. The unit hydrograph is below.
Table 5. Storm Hydrograph and Unit Hydrograph
|
Time |
Flow |
||
|
t/Tp |
q/qp |
hrs |
cfs |
|
0.0 |
0.000 |
0 |
0 |
|
0.2 |
0.100 |
1 |
253 |
|
0.4 |
0.310 |
2 |
784.3 |
|
0.6 |
0.660 |
3 |
1669.8 |
|
0.8 |
0.930 |
4 |
2352.9 |
|
1.0 |
1.000 |
5 |
2530 |
|
1.2 |
0.930 |
6 |
2352.9 |
|
1.4 |
0.780 |
7 |
1973.4 |
|
1.6 |
0.560 |
8 |
1416.8 |
|
1.8 |
0.390 |
9 |
986.7 |
|
2.0 |
0.280 |
10 |
708.4 |
|
2.2 |
0.207 |
11 |
523.71 |
|
2.4 |
0.147 |
12 |
371.91 |
|
2.6 |
0.107 |
13 |
270.71 |
|
2.8 |
0.077 |
14 |
194.81 |
|
3.0 |
0.055 |
15 |
139.15 |
|
3.2 |
0.040 |
16 |
101.2 |
|
3.4 |
0.029 |
17 |
73.37 |
|
3.6 |
0.021 |
18 |
53.13 |
|
3.8 |
0.015 |
19 |
37.95 |
|
4.0 |
0.011 |
20 |
27.83 |
|
4.2 |
0.008 |
21 |
20.24 |
|
4.4 |
0.006 |
22 |
15.18 |
|
4.6 |
0.004 |
23 |
10.12 |
|
4.8 |
0.002 |
24 |
5.06 |
|
5.0 |
0.000 |
25 |
0 |
Figure 3. 100 Year 1 Hour Storm Resulting Hydrograph Without Baseflow
Baseflow
The baseflow cannot be easily predicted. A summer baseflow is equal to approximately 20 cfs, while a springtime baseflow is closer to 400 cfs. When taking these into account a range of values for this particular hydrograph can be found. The resulting peak from assuming a spring runoff baseflow is approximately 3000 cfs, while a summer base flow gives a peak near 2600 cfs. When taking into account all the variables and the data sources a single significant figure should be used when presenting the final flows, therefore, a peak flow of 3000 is the predicted peak flow from the 100 year one hour storm over the basin.
Practicality of Flows
The last 16 years of flow data were analyzed a peak daily flow of 2260 cfs was recorded on April 28, 2005. All other peaks were less than 1000 cfs. This shows that the predicted flood is above the recorded flows. Although a 100 year flood and a 100 year storm are very different, our “worst case scenario” tries to create a maximum flood with the return period storm.
Figure 4. Hydrograph with April 28, 2005 Peak at Stream Gage Near Paradise, UT
Conclusion
Although the factors entering into the computation of a 100 year storm hydrograph are many, this report shows that with assumptions, basic calculations can near an accurate prediction. Further or more detailed data needed for hydrograph prediction include, a GIS raster of soil types, additional stream gage data within the drainage, the humidity records for the area, and specific reservoir operations. Further calculations to be performed to find flooding regions along the Little Bear River include detailed cross section profiles of the riverbed, and routing separate hydrographs from the storm using the Muskingum Method and open channel hydraulics.
References
Bedient, Philip B., Huber, Wayne C., Vieux, Baxter E. (2008) “Hydrology and Floodplain Analysis.” Ed: 4, Prentice-Hall, Inc., Upper Saddle River, NJ.
Fang, Xing. Et al. (2005) “Revisit of NRCS Unit Hydrograph Procedures.” ASCE Conference 2005.
Linsley, Ray K., Franzini, Joseph B. (1979) “Water-Resources Engineering.” Ed: 3, McGraw-Hill Publishing Company, New York, NY.
NCEES (Nation Council of Examiners for Engineering and Surveying). (2008) “Fundamentals of Engineering Supplied-Reference Handbook.” Ed: 8, NCEES Publishing, Clemson, SC.
National Resources Conservation Service (NRCS). (1972) “National Engineering Handbook.” Hydrology, Section 4.
National Resources Conservation Service (NRCS). Utah Snotel Sites. http://www.wcc.nrcs.usda.gov/snotel/Utah/utah.html. (Accessed December 1, 2009)
National Oceanic and Atmospheric Administration (NOAA). (2006) “Precipitation-Frequency Atlas of the United States" NOAA Atlas 14, Volume 1, Version 4.
USGS (United States Geological Survey). The National Map Seamless Server. http://seamless.usgs.gov/website/seamless/viewr.htm. (accessed November 21, 2009)
USGS (United States Geological Survey). MRLC Consortium Viewer. http://gisdata.usgs.net/website/MRLC/viewer.htm. (accessed November 29, 2009)
Here