Laserfiche WebLink
<br />Seeding was normally conducted for ei the:r 24-hour periods or on <br /> <br />a storm duration basis. Since this analysis was based on 6-houl' <br />blocks of time, a determination of whether or not a block was <br />seeded had to be based on the seeding ge:nerator log entries from <br /> <br />each proj ect . Generator turnon and turnoff times were taken into <br /> <br />account by assuming a block was seeded i.f one or more generators <br /> <br />were operating from 1 hour before the start of the 6-hour block <br />to 1 hour before the end of the 6-hour block. In addition, <br />contamination of nonseeded blocks preceded by seeded blocks was <br /> <br />a major concern in determining the category of a given 6-hour <br /> <br />block. However, it was decided not to take carryover contam- <br /> <br /> <br />ination into account when deciding whether or not a block was <br /> <br />seeded. This could be done later by substratification of the <br /> <br /> <br />data in the analysis. <br /> <br />Knowledge of the effect of seeding as a function of position on <br />the mountain barrier was required. To select single gages to <br />represent entire areas across the barri(~r could cause a signifi.. <br />cant 1055 of data for some projects because of missing individual <br />gage data. Therefore, it was decided to divide the barrier <br />into various regions of expected effect for computing precipi- <br />tation rates. Groupings of gages to compute an average regional <br />precipi tation rate was done typically for upwind, crest, downwind, <br />and far downwind regions of the barriers analyzed. Since there <br /> <br />13 <br />