|
<br />3 COLORADO FLOOD STUDY
<br />
<br />The aultidtacipllnary study of streems in the foothills and moun.
<br />talns of Colorado has concentrated on: (1) An analysis of available
<br />streamflow and precipitation data, (2) the use of paleohydrologic
<br />eechniques in flood-hydrology studies, and (3) flood information~
<br />transfer techniques (Jarrett, 1987; Jarrett and Costa, 1988). This
<br />study answers questions about the flood hydrometeorology that
<br />studies limited to a single discipline cannot provide because of
<br />limited dsta.
<br />
<br />3.1 Streamflow and precipitation
<br />
<br />In tbe foothills and mountains of Colorado, annual peak flows are
<br />derived from snowmelt at higher elevations, from rainfall at lower
<br />elevations, and/or from a combination of rain falling on snow (mixed
<br />population hydrology). Snowmelt.runoff peaks were distinguished
<br />from rainfall-runoff peaks on the basis of daily and seasonal occur~
<br />renee, hydrograph shape, and local weather conditions for 69 unregu~
<br />lated streams in Colorado in the South Platte. Arkansas, and
<br />Colorado River basins (Jarrett, 1987). Snowmelt. and rainfall-
<br />senerated peaks were used to develop snowmelt, rainfall, and coma
<br />posite flood-frequency curves that improved flood. frequency es.
<br />timates (Jarrett, 1987).
<br />
<br />Evaluation of rainfall and snowmelt floodafrequency curves can in.
<br />dieate which meteorological cause predominates for a stream. Com.
<br />parisons of flood-frequency curves for two streamflow.gaging stsa
<br />tions demonstrate that the change from snowmelt- to rainfall a
<br />dominated peak flows occurs abruptly over a 700aDleter range of
<br />elevation within .bout 25 kilometers in the Big Thompson River basin
<br />(this distance would vary by basin). The flood. frequency curves for
<br />the higher elevation (2,290 m) ststion (~ig Thompson River at Estes
<br />Park, A drAinage area of 355 .quare kilometers). indicates a
<br />snowmelt. dominated stream (Figure 1). In contrast, the flooda
<br />frequency curves for the lower elevation (1,615 m) station (~ig
<br />Thompson River at the mouth of the canyon near Drake, a drainage
<br />area of 790 km2), indicates a rainfall-dominated stream (Figure 2).
<br />Similar analyses were made for all 69 stAtions; these analyses indi.
<br />caCe that snowmelt runoff dominates above an elevation of about
<br />2,300 m (the elevation is lower for some river basins). Above this
<br />elevation. rainfall did not significantly contribute to peak flows.
<br />Below 2,300 m, rainfall produces laTge floods. These analyses also
<br />indicated that rain-on-snowmelt peak flows generally were small and
<br />very infrequent and tbat there is a relation between peak flow and
<br />elevation.
<br />
<br />The .1xed-populatIon analysis was done for 69 representative sta.
<br />tions in Color.do; however, a question remained: could rainfall
<br />floods have occurred elsewhere above 2,300 m? Therefore, data from
<br />all other U,S. Geological Survey streamflow-gaging stations (935
<br />stations) and miscellaneous flood measurement sites (706 flood
<br />sites) in Colorado were analyzed to determine if large rainfall-
<br />generated floods have occurred at higher elevations in Colorado
<br />
<br />".
<br />
<br />
<br />o
<br />z
<br />ZO
<br />-u
<br />~~50
<br />.""
<br />"W
<br />","
<br />uv>
<br />en 0:
<br />c~
<br />"w
<br />,,~
<br />~~
<br />..
<br />:>
<br />u
<br />
<br />r;~t.
<br />""~o
<br />,;,eO
<br />~"
<br />~t.\.~
<br />~O
<br />
<br />"
<br />....
<br />
<br />0.20 0.10 0.05 0,01
<br />EXCEEDANCE PROBAS1UTY
<br />
<br />Figure 1. Floodafrequency curves
<br />for the Big Thompson River at
<br />Estes Park, Colorado;
<br />elevation is 2,290 meters.
<br />
<br />0.001
<br />
<br /> -
<br /> ,-
<br />~
<br />~
<br />0:
<br />~
<br />en
<br />0:
<br />w
<br />t;
<br />~
<br />~ 1..
<br /> <--
<br />U ~ ~:}..;
<br />; ~t, S'" fI}9~
<br />~ ot::;; ~
<br /> 11-.' q.:..........-
<br /> 1...-
<br />0: -I
<br />"
<br />'"
<br />lil
<br />0
<br />"
<br />~
<br /> 10010 '50 0.10 0.1 ... 0.01
<br /> EXCEEDANCE pROUJMUTY
<br />
<br />
<br />Figure 2. Flood-frequency
<br />curves for the ~ig.Thompso
<br />River at mouth of eanyon.
<br />near Drake, Colorado;
<br />ele~ation is 1.615 meters.
<br />
<br />(Jarrett, 1987), The unit dischsrge (discharge divided by drainage
<br />area) for each peak discharge value was computed and ranked for eac
<br />county. For each major river basin, maximum unit discharge in each
<br />county was plotted &r.ainst elevation. The data and an envelope
<br />curve are shown in Figure 3 for the South Platte River basin, which
<br />includes the Big Thompson River basin. The magnitude of rainfall.
<br />generated floods decreases dramatically as elevation increases.
<br />Below 2,100 m, unit discharge has exceeded 40 mS/s/km2. Above
<br />2,300 m, unit discharge has not exceeded 1.1 m'/s/km2. Data for
<br />the other major river basins above 2,300 m in Colorado show that
<br />unit discharge also has not exceeded 1.1 mS/sfkm2.
<br />
<br />2.000 2.500
<br />ELEVATION. IN METERS
<br />
<br />Relation between maximum unit discharge and elevation.
<br />including envelope curve. for the South Platte River
<br />Basin in Colorado.
<br />
<br />",0: so
<br />",w
<br />wI;;
<br />t;::; ..
<br />::;9
<br />S:!i2
<br />"'w
<br />a~30
<br />~5
<br />wIll
<br /><0",20
<br />O:W
<br />"o.
<br />Xc
<br />~~10
<br />cu .
<br />k::~
<br />Zo: .
<br />:>~ 1,000
<br />
<br />......-
<br />/'
<br />/
<br />/
<br />/
<br />/
<br />,,/
<br />.
<br />
<br />.
<br />.
<br />
<br />1,500
<br />
<br />Figure 3.
<br />
<br />.
<br />
<br />. MAXIMUM UNrr
<br />DISCHARGE IN EACH
<br />COUNTY
<br />
<br />3,000
<br />
<br />3,500
<br />
|