Floods in the high Sierra Nevada, California, USA by uke86868


									                Hydrology in Mountainous Regions. H - Artificial Reservoirs; Water and Shoes
                (Proceedings of two Lausanne Symposia, August 1990). IAHS Pubi. no. 194,1990.

                Floods in the high Sierra Nevada, California, U S A

                RICHARD KATTELMANN
                Center for Remote Sensing and Environmental Optics
                Computer Systems Lab, 1140 Girvetz Hall
                University of California, Santa Barbara, CA 93106,USA

                ABSTRACT A variety of mechanisms generate peak flows in the high
                elevation parts of the Sierra Nevada. Snowmelt floods keep discharges
                high for several weeks each spring but rarely attain damaging stages much
                beyond bankfull. Rain-on-snow events generally account for the highest
                peak flows in most rivers. Although summer rainfall is rarely substantial
                in the Sierra Nevada, thunderstorms have produced the highest flows on
                record in basins where floods produced by other processes are constrained
                in magnitude. Lake outbursts generated by impoundment failure or
                avalanche displacement of lake water are localized in extent but produce
                the largest possible floods. These different types of floods vary in their
                potential for geomorphic change, but in combination, they continue to
                alter the channel system of the Sierra Nevada.

Rivers originating in the high elevation part of the Sierra Nevada both sustain and
occasionally threaten people and property at lower altitudes. Water flowing from the
Sierra Nevada is a critical resource for California's cities, agriculture, and hydroelectric
generating system. Excessive flows of this mountain water have also created difficulties
for communities and structures since the Gold Rush of .the mid-nineteenth century.
Attempts at regulating and redistributing runoff from Sierra Nevada have created a vast
network of dams and diversions along both sides of the range. Considering the extensive
development of water resources in the Sierra Nevada, the basic hydrology of this
mountain range has received surprisingly little study, particularly at higher altitudes.
    Most of our hydrologie knowledge of the high Sierra Nevada has come from interest
in water supply forecasting (Miller, 1955), water yield improvement (Anderson, 1963;
Kattelmann & Berg, 1987), or potential effects of acidic deposition (Dozier, et al., 1989).
Little work has been done concerning high-elevation floods and their geomorphic
consequences. This paper was written as part of the background review for a
comprehensive study of the hydrology and hydrochemistry of alpine catchments under
the NASA-Eos program. It discusses the primary mechanisms of flood generation in the
upper Sierra Nevada based on records from the limited set of gaged streams at high

Geographic Overview
The Sierra Nevada runs roughly northwest-southeast for more than 600 km and is about
100 km wide, on average. The subalpine forest zone is generally considered to begin
above 2100-2400 m in the northern Sierra Nevada and above 2500-2900 m in the south.
The alpine zone begins above 2800-3000 m in the north and above 3000-3300 m in the
south (Storer & Usinger, 1963). The subalpine zone is about 9000-9500 km 2 in area, and
the alpine zone covers about 2000-2500 km 2 . Snow dominates the hydrology of the high
Sierra Nevada, accumulating for six to seven months and then melting for three to four

Richard Kattelmann                                                    312

months. At higher elevations, the snow-free part of the year lasts only from mid-July
through mid-October. During four years of a recently-completed study in Sequoia
National Park at elevations of 2800-3400 m, more than 90 percent of the precipitation fell
as snow. Precipitation is minor and infrequent from May through September. Few stream
gages or climate stations exist above 2000 m, but about 100 snow courses and snow
sensors are used to monitor the winter snow cover. Most of the Sierra Nevada above
2500 m is reserved from economic use or habitation as National Parks or designated
wilderness. Consequently, floods at higher elevations are of little direct concern to
society. However, these high-elevation floods obviously contribute to low-altitude floods
as well as continue to alter the landscape of the Sierra Nevada.

Considering the overwhelming role of snow in the hydrologie cycle of the Sierra Nevada,
snowmelt floods are the most obvious source of peak flows. Snowmelt floods are an
annual event each spring of sustained high flow, long duration, and large volume.
However, they usually do not produce the highest instantaneous peaks. The Sierra
Nevada snowpack at the peak of winter accumulation represents an enormous reservoir of
potential runoff. The long-term record of snow survey measurements suggests that peak
snowpack water equivalence is about 75 to 85 cm, on average, at elevations above 2500
m (Kattelmann & Berg, 1987). In wet years, peak snowpack water equivalence can
exceed 200 cm and may persist well into the summer. Differences in snowmelt runoff
resulting from differences in snowpack water equivalence may be illustrated by
comparing hydrographs from three snowmelt seasons from a small alpine basin in
Sequoia National Park (Figure 1).

      (m /day)
       [ x 10 4 ]

                           A'M'   J ' J 'A 1   '   'A'M' J ' J ' A1       ' A ' M 1 J1 J1 A
                                1985                   1986                      1987
                FIG. 1 Peak daily flow during the snowmelt season was almost twice as
                great in the deep snowpack year of 1986 as in slightly below-average
                years of 1985 and 1987 in the 1.2 km 2 headwater basin of Emerald Lake.
                The volume of snowmelt runoff in 1986 was four times larger than in 1985
                and three times larger than in 1987.

    Snowmelt runoff production from a basin depends on the spatial distribution of both
snow and energy available for melt. In the rugged terrain of the alpine Sierra Nevada,
snowpack water equivalence is highly variable over large and small spatial scales (Elder,
et al., 1989). When snow cover becomes depleted and progressively less of the basin is
contributing to runoff, streamflow declines quickly. Snowmelt in the alpine zone
generally begins and peaks several weeks later than in the forest zone. In large river
basins, snow may have disappeared from the lowest elevations before snowmelt begins at
the highest elevations. Such desynchronization of snowmelt contributions helps to
reduce peak discharge.
                313                            Floods in the high Sierra Nevada, California, USA

     Snowmelt rates also determine the daily and seasonal peak flows. In the Sierra
Nevada, net radiation is the dominant influence on snowmelt (Miller, 1955; Aguado,
 1985). Widespread cloud cover is infrequent during spring in the Sierra Nevada, and
solar radiation input and longwave re-radiation are both high under clear skies.
Therefore, the exposure of each slope to solar radiation largely determines the energy
balance of the snow cover. Consequently, melt rates may change from 0 on the north side
of a ridge to several millimeters per hour on the south side. These drastic differences also
difiuse melt water input to streams over time, particularly when snow disappears from
south-facing slopes before melt rates peak on north-facing slopes. Melt rates at high
altitude are limited by the physics of energy exchange to generally less than 1 cm per
hour and 5 cm per day. Daily melt in the alpine zone is highest when several conditions
coincide: solar radiation, humidity, and wind are high; snow albedo and nighttime heat
loss from the snowpack are low; and the snow cover is nearly continuous with some of it
thin enough to allow radiation penetration to the ground. Snowmelt can be rapid once
rocks begin to be exposed because re-radiation from the sun-warmed rocks melts snow
much faster than does shortwave solar radiation. However, snow covered area must
decline simultaneously so the enhancement of streamflow is determined by tradeoffs
between higher melt rates and smaller contributing area.
    Particularly large snowmelt floods in Sierra Nevada rivers have been documented in
1906, 1938, 1952, 1969, and 1983. In all cases, snow deposition was more than twice
average amounts and persisted into April or May. Thus, snow cover was still extensive in
late spring when energy available for melt was much greater than in early spring.
However, the relative disposition of that energy in 1983, for example, was not
particularly different from that in other years with smaller snowpacks that melted earlier
(Aguado, 1985). The high rates of snowmelt runoff in the years of late-lying snow cover
led to widespread overbank flows, which are generally limited under snowmelt
conditions. For example, in 1983 in Sagehen Creek near Lake Tahoe, daily mean
discharge exceeded the bankfull level for 45 days and exceeded 230 percent of bankfull
discharge for two weeks (Andrews & Erman, 1986). This value of discharge had been
equaled or exceeded on only 16 days in 29 years of record before 1983. Compared to
more typical snowmelt floods, this event was highly effective in transporting bedload
(Andrews & Erman, 1986). In the lowlands, damage from these floods was mostly limited
to occupied floodplains and areas drained and converted to agriculture (Stafford, 1956;
Dean, 1975).

Mid-winter rainfall on snow cover has produced all the highest flows in major Sierra
Nevada rivers during this century. Rainfall has occurred up to the highest elevations of
the Sierra Nevada during winter, but the freezing level of winter storms generally
fluctuates between about 1000 m and 2500 m. Even during the warmest storms,
snowpacks above 2500 m rarely melt much because temperatures are close to freezing.
Snow in the forest zone, particularly at elevations between 1500 m and 2000 m, can
dramatically add to floods with melt from convection and condensation processes at
temperatures of up to 15°C. The interaction of precipitation amounts, freezing level,
energy availability, and basin characteristics determines the relative response of rivers at
different elevation zones. For example, a storm in mid-February 1986 produced some of
the largest flows of record in major rivers, but it did not produce much of a contribution
from areas above the 2000-2300 m rain-snow level. Three weeks later, a storm with
much higher freezing level but less precipitation produced some of the highest flows of
record in headwater basins, but generally only moderate flows downstream.
    In the past 60 years, six large-magnitude floods have occurred in almost all rivers
draining the snow zone. Snowmelt from convection and condensation processes was
important in all but one of these events when snow was absent. None of these floods was
Richard Kattelmann                                                  314

dramatically larger than the others in a majority of rivers, but the relative magnitude
depended on storm characteristics and the area-elevation distribution of each basin.
These floods at low elevation, which were 4 to 10 times the magnitudes of the mean
annual floods, have recurrence intervals of about 10 to 20 years based on the period of
record. Large-magnitude warm storms do not seem to occur during spring in the Sierra
Nevada. There are only a few moderate rain-on-snow events superimposed on spring
snowmelt floods in the streamflow record. The one large flood of this type occurred in
April 1982 and ranks in the top tenfloodsin the annual series in many headwater streams.
Storms in April and May generally do not incorporate warm air masses from low latitudes
that lead to the warm storms that occasionally occur in the winter months.
    In basins that are largely above 2000 m, the highest peaks also tend to be caused by
rain-on-snow events, even though almost all the other floods in the annual series are of
snowmelt origin. For example, in the Merced River in Yosemite National Park, the four
highest floods were caused by rain-on-snow and were 1.5 to 1.8 times greater in discharge
than the maximum snowmelt peak of record in May 1983. A hydrograph from a smaller
basin in Yosemite, Tenaya Creek, illustrates the contrast between a large snowmelt flood
of long duration and a high, but short, rain-on-snow peak (Figure 2). The impact of mid-
winter rain-on-snowfloodson channel processes may be enhanced by the confining effect
of snow along the banks. When present, snow may effectively block overbank spreading
of flood water and, thereby increase flow depth and bed shear stress for discharges above
bankfull (Erman, et al., 1988).

       (m 3 /s)

                         'O'N'D'J'F'M'A'M'J'J'A^'O'N'DIJ^ÏM'A'M'J 1 J1 A1 S
                                  WY1951                 WY1952
                  FIG. 2 Peak discharge from a rain-on-snow event in November 1950 was
                  about twice the maximum discharge of one of the largest snowmelt floods
                  on record in 1952. Mean elevation of this 120 km2 Tenaya Creek basin is
                  about 2000 m.

Although the summer and early autumn seasons in the Sierra Nevada tend to be dry, a few
minor storms or brief showers occur in most years (Hannaford and Williams, 1967). For
example, during the summers of 1985 to 1987, precipitation was measured on 29 days at
Emerald Lake at 2800 m (Dozier, et al., 1989). Daily precipitation exceeded 10 mm on
only 8 of these days and was in the form of snow on 6 of these 8 days. Intense rainfall,
possibly exceeding 20 mm per hour, was noted in the same basin on one day in each of
1984 and 1988 when gages were not present. In general, summer rainfall is much less of
a flooding concern in the Sierra Nevada than in the Rocky Mountains (e.g., Jarrett &
Costa, 1982).
               315                           Floods in the high Sierra Nevada, California, USA

    However, subtropical storms occasionally move into the southern Sierra Nevada in
late summer. Intense thundershowers occurring over a period of three or four days can
generate localflooding,cause extensive surface erosion, and destabilize hillslopes. Such
storms were a common event in the late 1970's and early 1980's when several hikers were
killed and bridges were washed out far downstream. These storms may generate the
greatest floods in some alpine basins that are sufficiently high to avoid mid-winter rain-
on-snow events and are oriented so that snowmelt rates are kept low because of northerly
exposure over much of the basin. For example, the four highest floods of Bear Creek
(gage near Lake Thomas A. Edison) were generated by summer rainfall. The peak
discharge of two of thesefloodswas more than twice that of the largest snowmeltfloodin
this basin of 136 km2 with a mean elevation of about 2850 m. Three other summer floods
appear in the annual flood series of this stream in 67 years. Widespread summer rainfall
has also generated one large flood of approximately 20-year return period in the 640 km2
upper San Joaquin River. In August 1989, a flood and debris flow generated by a
thunderstorm in the 2000 to 3000 m headwaters of Olancha Creek in the southeast part of
the Sierra Neavada damaged the Los Angeles Aqueduct several kilometers downstream at
1200 m (Sahagun & Warren, 1989).

The sudden release of water from storage generates the most extreme floods but occurs
under a limited set of conditions in a small fraction of the Sierra Nevada. Although this
type of flooding is localized, it may produce flood peaks that are at least several times
greater than those caused by any other process and is likely to produce debris flows.
Sierra Nevada lakes tend to be stable with little risk of failure of their impoundments of
bedrock or broad moraines. Only one failure of a small artificial dam is known to have
occurred at high altitude in the Sierra Nevada: North Lake at 2850 m during a summer
storm in 1986, although several dam failures have occurred at lower altitudes. The
failure of landslide and snow-avalanche dams that temporarily impound streams
undoubtedly occurs at a variety of scales in the Sierra Nevada, but large events of this
type are not known to have been documented. There is also the possibility of minor
outbursts of water stored within the small glaciers of the Sierra Nevada although on a
smaller scale than occurs in the Cascades (Richardson, 1968). The greatest potential for
this type of flood would be on Big Pine Creek below the Palisade Glacier, but the author
is unaware of any reported glacier releases there.
    Displacement of lake water by snow avalanches is yet another flood generation
process in high elevation streams of the Sierra Nevada. The impact of an avalanche on to
the ice cover of a lake can force large volumes of water into the outlet channel.
Avalanches are known to generate large flood waves when impacting ice-covered lakes
(Schytt, 1965; J0rstad, 1968), but the geomorphic significance of lake water displacement
in the absence of dam failure has not been described (e.g., Luckman, 1977). Following
the major avalanche cycle of April 1982, the author observed avalanche-induced
displacement of lake water in the Virginia Lakes basin where lake ice-cover was piled to
several meters depth on the shore near the lake outlet in a manner similar to that
described by J0rstad (1968) in Norway. During the Emerald Lake study mentioned
above, avalanches occasionally ran on to the lake and displaced water into the gaged
outlet stream. In February 1986, a massive avalanche struck the3lake and displaced up to
70 percent of the unfrozen water in the lake (about 90,000 m ) into the outlet stream
(Williams & Clow, 1990). The resulting flood removed all snow from the outlet channel
for 700 m downstream to the next lake and severely scoured the streambed. Williams and
Clow (1990) also observed lake displacement damage in Bishop Creek and Rock Creek
on the east side of the crest in 1986. We heard other personal accounts of avalanche-
induced outbursts at Corbett Lake in Sequoia National Park and Reds Lake near
Mammoth Mountain resulting from the same avalanche cycle. These events may be
Richard Kattelmann                                                  316

relatively common and are the only means of generating high flow immediately
downstream of lakes, which otherwise tend to attenuate floods. Avalanche-induced
outburst floods may also be caused by water supply development where small reservoirs
or storage tanks are located in avalanche paths (Oaks & Dexter, 1987). Such a flood
almost occurred in Incline Village near Lake Tahoe during the February 1986 avalanche
cycle when avalanche debris reached but did not destroy a large water tank.

These various flood-generation mechanisms modify stream channels to various extents.
Although debate continues about the relative effectiveness of common events (e.g.,
annual snowmelt floods) versus catastrophic events (e.g., rain-on-snow events) in shaping
the landscape (e.g., Wolrnan & Gerson, 1978; Beven, 1981), large floods would seem to
be particularly important in mountain streams because of the high proportion of material
transported as bedload. In mountain rivers, rare high-magnitude floods are generally
required to significantly alter the channel because material comprising the bed and banks
tends to be large and resistant to entrainment (Lisle, 1987). However, the sequence of
events of different magnitudes also determines the geomorphic effectiveness of particular
floods (Beven, 1981). Large floods that destabilize a channel can lead to enhanced
sediment transport from low-magnitude events over several decades (Lisle, 1987). Such
effects have been documented in the Lake Tahoe basin following extreme rain-on-snow or
thunderstorm events (Nolan & Hill, 1987; Glancy, 1988). Similarly, two large rain-on-
snow events in 1982 may have created channel conditions favorable for the high bedload
transport measured in the snowmelt flood of 1983 (Andrews & Erman, 1986). These
interactions of different flood processes may be a critical influence on channel form and
sediment transport in the Sierra Nevada.

ACKNOWLEDGEMENTS Studies contributing to this review of flood processes were
supported by the NASA-Eos program, California Air Resources Board, and Pacific Gas
and Electric Company.


Aguado, E. (1985) Radiation balances of melting snow covers at an open site in the
   central Sierra Nevada, California. Water Resources Research 21 (11), 1649-1654.
Anderson, H. W. (1963) Managing California snow zone lands for water. Research Paper
   PSW-6, USDA-Forest Service, Pacific Southwest Forest and Range Experiment
   Station, Berkeley.
Andrews, E. D. & D. C. Erman (1986) Persistence in the size distribution of surficial bed
   material during an extreme snowmelt flood. Water Resources Research 22 (2), 191-
Beven, K. (1981) The effect of ordering on the geomorphic effectiveness of hydrologie
   events. In: Erosion and Sediment Transport in Pacific Rim Steeplands, International
   Association of Hydrological Sciences, Publication 132, 510-525.
Dean, W. W. (1975) Snowmelt floods of April-July 1969 in the Buena Vista Lake, Tulare
   Lake, and San Joaquin river basins In: Summary of Floods in the United States
   During 1969, Water-Supply Paper 2030, U. S. Geological Survey, Washington, D.C.,
Dozier, J., J. M. Melack, K. Elder, D. Marks, S. Peterson, & M. Williams (1989) Snow,
   snowmelt, rain, runoff, and chemistry in a Sierra Nevada watershed. Final report,
   contract A6-147-32, California Air Resources Board, Sacramento.
               317                           Floods in the high Sierra Nevada, California, USA

Elder, K., J. Dozier, & J. Michaelsen (1989) Spatial and temporal variation of net snow
     accumulation in a small alpine watershed, Emerald Lake basin, Sierra Nevada,
     California, U.S.A. Annals of Glaciology 13,56-63.
Erman, D. C, E. D. Andrews & M. Yoder-Williams (1988) Effects of winter floods on
     fishes in the Sierra Nevada. Canadian Journal of Fisheries and Aquatic Sciences 45,
Glancy, P. A. (1988) Streamflow, sediment transport and nutrient transport at Incline
     Village, Lake Tahoe, Nevada, 1970-73. Water Supply Paper 2313, U. S. Geological
     Survey, Washington, D.C.
Hannaford, J. F. & M. C. Williams (1967) Summer hydrology of the high Sierra.
    Proceedings of the Western Snow Conference 35, 73-84.
Jarrett, R. D. & J. E. Costa (1982) Multidisciplinary approach to the flood hydrology of
     foothill streams in Colorado. In: International Symposium on Hydrometeorology,
     American Water Resources Association, Bethesda, 565-569.
J0rstad, F. A. (1968) Waves generated by landslides in Norwegian fjords and lakes.
    Publication 79, Norwegian Geotechnical Institute, Oslo, 13-32.
Kattelmann, R. & N. Berg (1987) Water yields from high elevation basins in California.
     In: Proceedings, California Watershed Management Conference (R. Z. Callaham and
    J. J. DeVries, eds.), Report No. 11, University of California Wildland Resources
     Center, Berkeley, 79-85.
Lisle, T. E. (1987) Overview: channel morphology and sediment transport in steepland
    streams. In: Erosion and Sedimentation in the Pacific Rim, International Association
    of Hydrological Sciences, Publication no. 165, 287-297.
Luckman, B. H. (1977) The geomorphic activity of snow avalanches. Geografiska
    Annaler59A (1-2), 31-48.
Miller, D. (1955) Snow cover and climate in the Sierra Nevada, California. University of
    California Press, Berkeley.
Nolan, K. M. & B. R. Hill (1987) Sediment budget and storm effects in a drainage basin
    tributary to Lake Tahoe. Eos, Transactions American Geophysical Union 68 (16),
Oaks S. D. & L. Dexter (1987) Avalanche hazard zoning in Vail, Colorado: the use of
    scientific information in the implementation of hazard reduction strategies. Mountain
    Research and Development 7 (2), 157-168.
Richardson, D. (1968) Glacier outburst floods in the Pacific Northwest. Professional
    Paper 600-D, U.S. Geological Survey, Washington, D.C, D79-D86.
Sahagun, S. & J. Warren (1989) Storms go on; new aqueduct damage found. Los Angeles
    Times 109, August 12, 1989, 1 & 22-23.
Schytt, V, (1965) Notes on glaciological activities in Kebnekaise, Sweden during 1964.
    Geografiska Annaler47A (1), 65-71.
Stafford, H. M. (1956) Snowmelt flood of 1952 in Kern River, Tulare Lake, and San
    Joaquin River basins. In: Floods of 1952 in California, Water Supply Paper 1260-D,
    U.S. Geological Survey, Washington, D.C, 562-573.
Storer, T. L. & R. L. Usinger (1963) Sierra Nevada Natural History. University of
    California Press, Berkeley.
Williams, M. W. & D. W. Clow (1990-in press) Hydrologie and biologic consequences of
    an avalanche striking an ice-covered lake. Proceedings of the Western Snow
    Conference 58.
Wolman, M. G. & R. Gerson (1978) Relative scales of time and effectiveness of climate
    in watershed geomorphology. Earth Surface Processes 3,189-208.

To top