Floods

Overview

The Colorado River is s...

File:FILENAME.png

Figure 1. CAPTION. SOURCE.)

The Colorado River Basin historically experienced periodic flood events, in which the river or its tributaries rose over their banks and inundated adjacent lands. These floods disproportionately contributed to the river’s overall sediment transport and deposition. Many native fish species and riparian plant and tree species depended on flooding for completion of their life cycle.

With increasing Euro-American settlement of the basin in the late 1800s, especially the agricultural development along the river in the Lower Basin, these periodic floods were cast as a destructive menace. Accordingly, Hoover Dam and many dams subsequently built on tributaries have flood control as a primary purpose.

Today, the natural flow regimes on the mainstem and most tributaries have been extensively altered by dams and diversions. These alterations have especially affected the periodic flood events, greatly reducing their frequency and magnitude compared to pre-management conditions. For the mainstem below Glen Canyon Dam (Lees Ferry, AZ gage; Figure 2), the average annual peak streamflow has been reduced by about two-thirds since the closing of the dam (1963), from 85,000 cfs to 30,000 cfs. Annual peak streamflows greater than 100,000 cfs at the Lees Ferry gage occurred 12 times from 1921 to 1962, but have not occurred since (although Lake Powell inflows greater than 100,000 cfs occurred in both 1983 and 1984; Table 1).

The flood of 1983 damaged the outlet works at both Glen Canyon Dam and Hoover Dam and inundated hundreds of homes and business below Lake Mead, demonstrating that, despite the overall attenuation of the river’s flood regime, major impacts from large floods can still occur. infrastructure and property damage from large floods. Yet the peak streamflows in 1983, and in 1984, were not remarkable compared to historical floods in the 19th and 20th century, let alone the even larger paleofloods that have been reconstructed mainly from flood deposits stranded well above more recent high-water marks (Table 2).

The paleoflood records show that many events exceeding the largest historical floods (1862-present) have occurred on the mainstem and major tributaries over the last few thousand years. Accordingly, flood-frequency analyses that incorporate paleoflood events consistently result in larger flood magnitudes at a given frequency (e.g., 1-in-100 years) than analyses based on the gaged record alone.

The largest historical floods on the mainstem and major tributaries (Table 1) show seasonal and geographic clustering that reflects multiple meteorological and hydrologic mechanisms:

• Most of the large floods on the Upper Basin tributaries and on the mainstem have occurred in May, June, and July as a result of the rapid melt of unusually large snowpacks in the headwaters, in some cases exacerbated by heavy rains.
• In the Lower Basin tributaries and the southern Upper Basin tributaries (e.g., San Juan, Virgin, Escalante) and the downstream mainstem, large floods in September and October have been associated with low-pressure troughs and excess moisture from landfalling Pacific tropical storms.
• Several of the largest floods on Lower Basin tributaries, and on the mainstem from Grand Canyon downstream, have occurred in winter (December-February), likely as the result of one or a series of Pacific extratropical cyclones accompanied by atmospheric rivers.

The flood events driven by snowmelt manifest not just in high peak discharges but also in very large monthly, seasonal, and water-year streamflow volumes. Rainfall-dominated flood events typically have much sharper peaks and shorter durations, and produce less anomalous streamflow volumes on monthly and longer timescales, given the same peak discharge.

While Colorado River basin floods have occurred during every phase of ENSO (El Nino, neutral, La Nina), there is an overall propensity for major floods during El Nino events. The floods in 1884, 1891, 1911, 1920, 1941, 1952, 1957, and 1983 coincided with El Nino events, which last for 1-2 years. This is consistent with a NOAA analysis that over the period 1895-2015, El Nino events have been associated with a much greater likelihood of very wet conditions (precipitation >80th percentile) in the Lower Basin in fall, winter, and spring (Oct-May), and also in the Upper Basin in late winter and spring (Feb-May).

The ongoing warming of the Colorado River basin is likely to increase the risk of extreme precipitation events in the basin, due mainly to warmer air having the capacity to hold more water vapor, all else being equal. Consequently, climate change may elevate the flood risk across the basin–including increased magnitudes of the largest floods–even as the warming is also causing annual average streamflows to decline.

Table 1. Historical large floods on the mainstem and large tributaries

Date and Year	Gage or Location	Gaged or estimated (*) discharge, cfs	Reference or source	Comment
January 1862	Colorado R. at Topock, AZ	400,000*	Dickinson 1944; O’Connor et al. 1994	Winter rains; likely atmospheric rivers; same systems caused floods in CA
July 4, 1884	Colorado R. at Fruita, CO	125,000*	Follansbee and Sawyer 1948	Snowmelt
July 1884	Colorado R. at Lees Ferry, AZ	210,000*	Topping et al. 2003	Snowmelt; same event as above
Early 1888 (?)	Little Colorado R. near Cameron, AZ	177,000- 194,000*	Unema et al. 2021	Assumes the high-water mark emplaced 1884-1891 was caused by the known flood in 1888
February 1891	Gila below Gillespie Dam, AZ	250,000*	Smith and Heckler 1955	Main source of flood on mainstem at Yuma
February 26, 1891	Colorado R. at Yuma, AZ	140,000- 200,000*	USGS stage-height record; discharge estimated from height	Winter rainstorm (atmospheric river(s)?); flood mainly from Gila
June 24, 1909	Colorado R. at Yuma, AZ	150,000	USGS gage record	Snowmelt
October 6, 1911	San Juan near Bluff, UT	150,000*	Webb et al. 2001	Tropical storm?
January 22, 1916	Gila R. near Dome, AZ	200,000 (230,000* at mouth)	USGS gage record; Smith and Heckler 1955	Winter rainstorm (atmospheric river(s)?); main source of flood same day on mainstem
January 22, 1916	Colorado R. at Yuma, AZ	250,000	USGS gage record	Flood mainly from Gila
June 17, 1917	Green at Green River, UT	68,100	USGS gage record	Snowmelt
June 8, 1920	Colorado R. at Yuma, AZ	190,000	USGS gage record	Snowmelt
June 16, 1921	Colorado R. at Fruita, CO	81,100*	Follansbee and Sawyer 1948	Snowmelt
June 17, 1921	Green at Green River, UT	65,500	USGS gage record	Snowmelt; contributed to the peak at Lees Ferry the following day
June 18, 1921	Colorado R. at Lees Ferry, AZ	170,000*	Topping et al. 2003	Snowmelt; same event as above
June 1921	Colorado R. at Yuma, AZ	167,000-188,000*	Topping et al. 2003	Snowmelt; same event as above
September 19, 1923	Little Colorado R. near Cameron, AZ	120,000	USGS gage record	Rainstorm
July 1, 1927	Colorado R. at Lees Ferry, AZ 1|27,000	USGS gage record	Snowmelt; heavy rains in SW CO
September 10, 1927	San Juan R. near Bluff, UT	70,000	USGS gage record	Rainstorm
May 17, 1941	Colorado R. at Lees Ferry, AZ	120,000	USGS gage record	Snowmelt
June 12, 1952	Colorado R. at Lees Ferry, AZ	123,000	USGS gage record	Snowmelt
June 12, 1957	Colorado R. at Lees Ferry, AZ	126,000	USGS gage record	Snowmelt
June 1983	Inflows to Lake Powell (Colorado R., San Juan R.)	116,000	USGS gage record	Snowmelt plus rain-on-snow; major damage to Glen Canyon Dam outlet tubes
June 1983	Colorado R. at Lees Ferry, AZ	97,000	USGS gage record	Same event as above; dam operations attenuated the peak by ~20,000 cfs
July 1984	Inflows to Lake Powell (Colorado R., San Juan R.)	125,600	USGS gage record	Snowmelt
March 29, 1996	Colorado R. at Lees Ferry, AZ	45,900	USGS gage record	Highest flow of the high-flow experiments (HFE; 1996-present)

Additional resources

State of the Science Report

Chapter # of the State of the Science report, describes...