FLOOD08489 - Laserfiche WebLink

Home Browse Search

The drainage density is defined as the ratio of the channel length per unit area: L Dd s (4) A where Dd is the total length stream channels per unit of watershed area. Ls is the total length of all channels both ephemeral and perennial. There is a practical limita- tion to the evaluation of the drainage density. The only way that all of the ephemeral drainage channels can be identified is by a detailed survey in the field or by careful analysis of aerial photographs. A great deal of the details required to determine the drainage density are lost on topographic maps of scale 1:24000 (7-1/2 minute quadrangle sheets). Even with maps of the scale 1: 24000, the task of determining the drainage density is laborious. As originally conceived and used by Horton (1945) and later by Langbein (1947). the channels were defined by the blue lines shown on the topographic sheets. This practice resulted in some inconsistency in the re- sults depending upon the season of the year and the relative wetness or dryness of the year in which the maps were prepared. During the time when flood runoff is occuring, many depressions or otherwise ephemeral channels are also part of the active channel system. Therefore, when attempting to establish re- lationships between flood hydro graph parameters and the channel network system it is perfectly valid to consider these depressions and ephemeral channels as part of the drainage system. Carlston (1963) extended the channel networks into all depressions and drainage ways suggested by upslope "V" shaped in- terruption in the contour lines. The preparation of the extended channel network in a watershed of any appreciable size is laborious. In order to reduce the task to acceptable magnitude, Balayo (1967) using the technique described by Carl- ston obtained estimates of the drainage density using the extended channels on sample blocks in the catch- ment. Standard 4-centimeter square blocks were selec- ted at random on the watershed map. All channels were extended and the length of the extended channel net- work was determined. The average drainage density was computed for five sample blocks. The data for the sixth block was entered and the new average drainage density was computed. The procedure was repeated adding one block at a time until the change in the average was less than one percent. The average drain- age density was then adopted as the drainage density for the entire watershed. Using the method of obtain- ing estimates of the extended channel networks, Carlston (1963) found this regression equation for relating the unit area mean annual flood (Q peak for Tr = 2.33 years) and the drainage density: O2.33 Drainage Area 2 1. 3 Dd Wolman and Miller (1960) found that the channel capa- city developed in a watershed tended to equal to the value of recurrence interval of 2.33 years. This value is used as a standard or normal value in geomor- phic processes. In general Carlston found that the drainage den- sity for the Appalachian watersheds he studied varied from 3.0 to 9.0 miles per square mile. The highly permeable sandy watersheds always had lower values of the drainage density. When the rainfall readily in- filtrates into the watershed, overland flow is not available for development of the drainage network. In addition, as the watershed develops a channel network, the flood peak discharge increases. In a complementary part of the investigation on the influence of channel networks on the runoff hydro- graph, Carlston (1963) found that the unit area base flow was inversely related to the drainage density: Obase Drainage Area 14D -2 d Since the base flow is supplied from the groundwater storage, it is logical for the unit area base flow to be greatest under those conditions when the surface runoff is least efficient. The drainage density is directly related to the surface runoff drainage effi- ciency (considering the watershed slope, channel slope, area and roughness to be constant). It is assumed that the channel networks and con- sequently the drainage density evolved naturally with- out man-made restrictions. In the urban environment, the channel network in existence before the landscape was urbanized is drastically altered. In some cases the major drainageways remain, but these are altered. The overbank areas are reduced, channels are straight- ened; sometimes the roughnesses are removed. The result is deeper flowing water, higher velocities and a hydraulically more efficient channel. Urbanization often obliterates entirely the secondary channel networks. These are replaced by a network of roadside ditches, or curb and gutter net- works. The curb and gutter network is relatively smooth and the alignment is straight. These rela- tively deep straight hydraulically efficient channel networks decrease the transit time of a flood wave in the channel network system. CHANNEL SLOPE The channel slope in the watershed relates the rate with which the potential energy of the streamflow is consumed in friction losses, turbulence and kinetic energy. Kirpich's (1940) relationship for the time of concentration contained a slope term in the length- slope parameter. Likewise USSR (1965) enlarging on Snyder's (1938) work with the unit hydrograph found that the channel slope appeared in the lag time rela- tionships. Dempster (1974) found that the length-slope parameter was a significant parameter in the regression model predicting the peak discharge of the T-year flood. The past researchers have found various ways of de- fining the channel slope. In the typical natural watershed, the stream channels increase in size proceeding in the downstream direction because the watershed area contributing the flood runoff increases. The increase in channel size may also be attributed to the decrease in the average stream velocity proceeding in the downstream direction. Usually the channel gradients are greatest in the headwaters region and progressively flatten in the downstream direction. 3