Laserfiche WebLink
<br />process of successive approximations. computing flood runoff from <br />various trial sequences, using equation 1. <br />Selection of critical sequences of melt factors in equations 2 <br />and 3 is much more complex than the selection of temperature sequences <br />alone, since variations of the different parameters must be consistent. <br />The process is largely a matter of judgment and of successive approxi- <br />mation of resulting runoff severity. <br />Complete snowpack and snowmelt computation techniques described <br />above are contained in the computer programs described in Appendix 1. <br /> <br />Section 2.08. Ground conditions <br /> <br />Different wetness of the ground and different vegetal cover can cause <br />large differences in runoff from storm rainfall and snONmelt. In very <br />porous soils particularly, loss rates can vary greatly with ground wet- <br />ness. In cold regions, freezing of the ground can reduce substantial <br />losses to practically zero. Since a .ajor portion of the rainfall in <br />most storms is commonly lost to interception and infiltration into the <br />ground, it is apparent that selection of ground conditions is a major <br />consideration in deriving hypothetical floods. <br />For various reasons, loss rates in natural drainage basins average <br />much less than rates observed in controlled laboratory experiments with <br />similar soils--sometimes an order of magnitude smaller. Consequently, <br />loss rates used in the computation of hypothetical floods should be <br />derived from data on a watershed similar to the one under study. Rates <br />in nature range from less than 1 to more than 10 millimeters per hour <br />when expressed as average for an entire basin. <br />Loss rates can be related to ground conditions where data on rain- <br />fall and runoff are available for a large number of floods. USually an <br />index of ground conditions such as total antecedent precipitation is <br /> <br />2-14 <br />