Management of Runoff Flood Waters through McPhee Reservoir
Term Project Final Report
GIS in Water Resources
December 16, 2009
Josh Mortensen
Introduction
McPhee Reservoir is the second largest man-made reservoir in Colorado and provides water to users in the South West region of Colorado. While the water from McPhee is used for municipal and industrial uses, hydropower, recreation, and flood control most of the water is used for irrigational purposes. This reservoir provides irrigation water for approximately 62,000 acres of farmland downstream. Due to the dry climate of South West Colorado and the droughts of recent years water conservation is a major concern. Historical data shown in Figure 1 has shown that since the reservoir and dam’s completion it has reached capacity 13 times in the last 24 years (USBR, 2009). While this is encouraging for a reservoir in a dry climate, the reservoir is quickly drawn down each summer due to irrigation. Knowledge of water flow into the reservoir is essential to efficiently operate the reservoir and conserve the greatest amount of water.
Currently inflow to McPhee is measured by a single USGS stream gage on the Dolores River which enters on the South East side of the lake as shown in Figure 2. This stream gage provides the only inflow data to assist in regulating releases to the irrigation canals on the west side of the lake (USGS, 2009). Figure 2 shows there are at least four other inlets on the North and East side of the reservoir which may contribute to the inflow which are not measured (CDWR, 2009). While water may not run through these inlets all year round, the amount of runoff that enters the reservoir at these locations may be significant during springtime as well as during a storm event. Knowledge of the amount of water that enters the lake due to a storm event in the upstream watersheds may be helpful in operating the reservoir to conserve as much water as possible by regulating the pool elevation and release flows.
The main objective of this project is to predict the inflow at remote inlet points where stream gage data are unavailable and determine if it is significant in changing the pool elevation of the reservoir. Tools within ArcGIS software are used to analyze the upstream watersheds that drain to these points by predicting their specific runoff coefficients and topographic attributes. Future work that would need to be done for a complete analysis would include a model of a 100 year – 2 hour storm event over the watershed. This method could then be verified by simulating a past storm event over the watershed that drains to the main inlet at the Dolores River and comparing the stream flow results at the Dolores River with the stream gage data at the same location for the same time period. However, this would be a long iterative process that would require more time than was allotted for this project. However, accurate results would be a valuable resource for the reservoir and could possibly replace any future stream gages at these remote locations if the stream flow there were significant.
Figure 1: Historical pool elevation and volume data of McPhee Reservoir (USBR, 2009)
Figure 2: Map showing inlets and outlets of McPhee Reservoir (CDWR, 2009)
Procedure
Four different remote inlet locations are shown and the sub watersheds that drain to those points are organized into four separate watersheds (Fig. 2). Tools within ArcGIS are utilized to represent the drainage area, slope, elevation, and land cover attributes of each organized watershed. Arc Hydro Tools are used to determine the drainage lines and areas to each inlet. The Spatial Analyst tools use data from a digital elevation model (DEM) to represent the slope and elevation of the upstream watersheds. Tools from within the arc toolbox are used with land cover data to determine runoff coefficients of each watershed.
Data were downloaded from various databases on the World Wide Web and imported into GIS for analysis. Watershed and boundary data were downloaded from the National Hydrography Dataset as well as NHD flow line data (NHD, 2009). The DEM was taken from the National Elevation Dataset (NED, 2009) and land cover data was acquired from the National Land Cover Dataset (NLCD, 2009).
Results
Figure 3 shows how the original sub watersheds were organized into four watersheds that drain to each of the four remote inlets (red dots). NHD flow lines were imported to help determine which sub watersheds drained to each of the remote inlets. Each of the new watersheds were analyzed to determine drainage area, elevation, slope and land cover results. These results will be useful in predicting runoff into the four remote inlet locations.
Figure 3: Watersheds of each remote inlet formed from sub watersheds.
Drainage area results are shown in Table 1 in acres and square miles. Watershed 3 has the greatest drainage area as expected as it is a combination of three separate sub watersheds. Watershed 4 has the smallest area as it is a single sub watershed of small size. Knowledge of the total drainage area of these watersheds will be useful in determining if the stream flow at the remote inlets is significant in changing the pool elevations of the reservoir.
Table 1: Drainage Area of each watershed
|
Watershed |
Drainage Area |
|
|
# |
acres |
sq. miles |
|
1 |
24,326 |
38.0 |
|
2 |
46,117 |
72.1 |
|
3 |
53,305 |
83.3 |
|
4 |
11,990 |
18.7 |
|
Total |
135,738 |
212 |
The elevation of each watershed is shown in Table 2. These results were obtained by using the DEM (Fig. 4) and Spatial Analyst tools. Watershed 2 has the highest elevation which correlates to its furthest location from the lake. Elevation increases to the North and East of the reservoir. Elevation ranges were fairly consistent among the first three watersheds. Watershed 4, which is directly north of the McPhee reservoir, showed a much smaller elevation range which also correlates to its smaller area.
Figure 4: DEM used to determine elevation information of organized watersheds
Table 2: Elevation results of each watershed
|
Watershed |
Mean Elevation |
Elevation Range |
|
# |
ft |
ft |
|
1 |
7,686 |
1,809 |
|
2 |
8,409 |
2,209 |
|
3 |
8,152 |
2,202 |
|
4 |
7,705 |
1,315 |
The DEM was projected to raster data using the tools within the Arc Tool Box and then Spatial Analyst tools were used to obtain the slope results in Figure 5 and Table 3. The image from the projected DEM shows deep canyons on three of the four remote inlets. This is reflected in the results that show very steep slopes in each watershed. Watershed 4 has the steepest mean slope which may be due to the canyons within its small area. The steeper slopes in Watersheds 3 – 4 may produce more runoff to those inlets than the one on Watershed 1. This seems likely from the results in the table as well as the image.
Figure 5: Image of slopes over watersheds
Table 3: Slope results of each watershed
|
Watershed |
Mean Slope |
Min. Slope |
Max. Slope |
|
# |
% |
% |
% |
|
1 |
9.92 |
0 |
112.2 |
|
2 |
14.13 |
0 |
139.7 |
|
3 |
17.35 |
0 |
166.1 |
|
4 |
22.39 |
0 |
125.6 |
As shown in Figure 6 and Table 4 most of the land cover in each watershed is Forest and Range. Spatial Analyst tools were again used along with the Tabulate Area tool to determine land cover percentages for each watershed. For each watershed Forest land cover accounted for at least 50% of the total area while Range land cover accounted for most of the remainder. Runoff coefficients were calculated using assumed coefficients for each land cover type (Table 5) and an area weighted average. Due to the assumed runoff coefficients that were used as well as Forest and Range land cover dominating the watershed areas, calculated runoff coefficients were all close to 0.3. These lower runoff coefficients in the watersheds may cause lower stream flow the remote inlet locations. This may be why there is not flow all year round at these points.
Figure 6: Land cover over watersheds
Table 4: Land cover percentages and average weighted runoff coefficients
|
Watershed |
Land Cover Type |
Runoff |
|||
|
# |
Open Water |
Range |
Forest |
Agriculture |
Coefficient |
|
1 |
3.09% |
33.45% |
62.47% |
0.99% |
0.29 |
|
2 |
0.03% |
25.11% |
73.04% |
1.83% |
0.25 |
|
3 |
0.04% |
46.75% |
50.91% |
2.31% |
0.30 |
|
4 |
7.80% |
22.18% |
68.07% |
1.95% |
0.31 |
Table 5: Assumed runoff coefficients used in calculations
|
Land Cover |
Assumed Runoff Coefficient |
|
Water |
1 |
|
Range |
0.4 |
|
Forest |
0.2 |
|
Agriculture |
0.3 |
Summary and Conclusions
ArcGIS was used to analyze the topographical and land cover attributes of four organized watersheds that provide runoff into McPhee reservoir. Results showed that lower runoff coefficients resulted from the type of land cover in these four watersheds. These lower coefficients may be part of the reason that stream flow only occurs part of the year at these remote inlet locations. Again, since stream flow is currently measured only at the main inlet to the reservoir, the purpose of this project was to determine if runoff to the lake at the four remote inlets was significant enough to influence the pool elevation of the reservoir. Additional work would be required to answer this question such as modeling a storm event or analyzing precipitation data from a past storm event. This process could be verified by making a similar analysis over the watershed that drains to the main inlet of McPhee and comparing results with the stream gage results at that point. After this analysis if the stream flow at these locations does influence the pool elevations of the lake significantly the results may be used to optimize reservoir operations in order to conserve more water. The timing of regulating water releases and the incoming runoff from a storm event would then need to be worked out as well. GIS proves to be a useful tool in predicting the significance of results before actual stream gages are installed, possibly saving a great deal of time and money.
References
Colorado Division of Water Resources (CDWR) “Colorado Decision Support Systems Map Server” October 2009 <http://water.state.co.us/>
National Elevation Dataset (NED) “The Seamless Data Distribution System” December 2009
National Hydrography Dataset (NHD) “NHD Geodatabase” October 2009
National Land Cover Dataset (NLCD) “Get Land Cover Data” December 2009
< http://landcover.usgs.gov/index.php>
United States Bureau of Reclamation (USBR) “Historic Data – McPhee Reservoir” October 2009 <http://www.usbr.gov/>
United States Geological Survey (USGS) “Water – Historical Data” October 2009