1.1 Developing Updated PMP Estimates for the Third and Fifth Largest Reservoirs in the United States

Monday, 8 January 2018: 8:45 AM
Room 18A (ACC) (Austin, Texas)
Charles D. McWilliams, U.S. Army Corps of Engineers, Omaha, NE; and C. Pathak

Fort Peck Dam is located in northeastern Montana within the upper portion of the Missouri River basin approximately 17 miles southeast of Glasgow and 10 miles southwest of Nashua. The Missouri River is formed at the confluence of the Jefferson and Madison Rivers near Three Forks, Montana and is joined approximately a mile downstream by the Gallatin River. Major tributaries of the Missouri River upstream of Fort Peck Dam include the Sun, Teton, Marias, and Musselshell Rivers. Garrison Dam is located near Riverdale in western North Dakota and is approximately 55 miles northwest of Bismarck. Major tributaries of the Missouri River downstream of Fort Peck Dam include the Yellowstone, Milk, Poplar, and Little Missouri Rivers, with the Yellowstone River having the greatest amount of drainage contribution to Lake Sakakawea.

Fort Peck Lake has a total upstream drainage of 57,500 square miles and a storage capacity of 18,463,000 acre-feet, making it the 5th largest reservoir in the United States and the first of a series of 6 USACE dams located along the upper Missouri River Basin. Lake Sakakawea is the second of a series of 6 USACE reservoirs located along the upper Missouri River Basin and has a total upstream drainage of 123,900 square miles and storage capacity of 23,821,000 acre-feet, making it the 3rd largest in the United States. The other projects working downstream are Oahe, Big Bend, Fort Randall, and Gavins Point. Lake Oahe is the 4th largest reservoir in the United States. The record flows for Fort Peck and Garrison Dams are, respectively, 65,900 cfs and 150,600 cfs in June 2011.

As part of an update to the Inflow Design Flood for the Garrison project, a Probable Maximum Precipitation (PMP) study was performed in order to quantify the appropriate rainfall to use in the hydrologic modeling of the updated Probable Maximum Flood (PMF). Since the watershed is too large for the simplified generalized PMP methods of the Hydrometeorological Reports (HMRs), a modified approach was used based upon guidance from the HMRs and the World Meteorological Organization Manual on the Estimation of PMP.

The enormous size of each watershed required for several separate storm centerings to be considered to take into account the higher elevations as well as the High Plains region closer to Fort Peck Lake and Lake Sakakawea. In addition, Probable Maximum Storms (PMS) were developed for both early spring and late spring flood scenarios. These were assumed to occur on 1 April and 1 June, respectively, with an adjustment based upon moisture climatology included for further future refinement by the hydrologic modelers.

The calculations within this study included the extreme storm events that have been observed in the region of interest over the past century with the results accounting for the significant orographic effects of the watershed. The PMP storm for the upper portion of the Fort Peck basin is a maximized (by 170%) recreation of the June 1964 Gibson Dam storm, which is the key PMP driver for the Northern Rockies and occurred within the Fort Peck watershed. This produced a maximum rainfall estimate of 12.03 inches for 1 April and 24.33 inches for 1 June. Transposition of the June 1964 Gibson Dam storm into the upper Garrison watershed was performed using the isopercental method. This produced a maximum rainfall estimate of 9.59 inches for 1 April and 21.29 inches for 1 June. A review of the temporal pattern of the storm was also used in order to maintain consistency with the observations.

In the development of the PMP for the lower portion of the Fort Peck basin, the HMR 52 elliptical storm pattern was used as the base spatial distribution and then modified for orographic effect through the normalization of precipitation frequency data. These steps were performed in order to compensate for the impact of rainfall intensity within the region as a result of a series of mountain ranges located in central Montana. The maximization and transposition of extreme storms into the basin centroid produced maximum rainfall estimates of 8.24 inches for 1 April and 18.55 inches for 1 June.

For the lower portion of the Garrison basin, the HMR 52 elliptical storm pattern was also used as the base spatial distribution and then placed with two distinct storm centers and orientations. One storm center selected was in the lower Yellowstone River and oriented at 222°. The maximization and transposition of extreme storms into that location produced maximum rainfall estimates of 10.50 inches for 1 April and 19.93 inches for 1 June. The additional storm center selected for the lower Garrison watershed was centered near the confluence of the Missouri and Yellowstone Rivers and oriented at 279°. The maximization and transposition of extreme storms there produced maximum rainfall estimates of 10.53 inches for 1 April and 20.46 inches for 1 June. The extreme storm of June 1921 at Springbrook, MT was observed to be the PMP controlling storm for much of the Northern High Plains. The spatial pattern of the observed rainfall analysis is consistent with the elliptical pattern in this study and this storm was utilized in the development of a temporal pattern for each PMS storm center in the lower basin.

A comparison of the 96-hour results of this study with the previous analysis in HMR 20 for the Garrison watershed indicates a 6% increase in PMP depths for late spring (June 1) and a 42% decrease for early spring (April 1) at a storm area size of 10,000 mi2. These results are not unexpected as the PMP values quantified in HMR 20 were for mid-May, which is a season where the moisture availability is much higher (or lower) compared to June 1 (or April 1). Follow-on development of the PMF requires the incorporation of extreme snowpack (from both mountain and plains sources) as well as the development of an appropriate temperature sequence for use in hydrologic models.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner