WSP08775 - Laserfiche WebLink

Home Browse Search

6 J. HYDROLOGIC AND HUMAN ASPECTS OF THE 1976-77 DROUGHT distribution of precipitation nor that some areas can withstand large variations in pre- cipitation whereas other areas cannot. Other hydrologists have found that the criterion of a 15 percent annual deficiency is meaningless if applied to desert areas where the mean annual precipitation does not exceed 5 inches because precipitation in many years is less than 85 percent of the mean and a drought implies a large departure from the norm. In arid or semiarid areas, water deficien- cies are chronic. Irrigation is necessary for most agriculture, and metropolitan areas-Los Angeles, San Diego, and San Francisco for example-also depend upon imported water for domestic, municipal, commercial, and indus- trial supplies. Therefore, drought conditions can be experienced where and when the demand exceeds the supply. Hydrologists use analyses of low-flow fre- quencies to "define" a hydrological drought on an annual basis. If the actual low flow of a natural stream for a selected number of days has a small probability of occurrence then one can conclude that a drought, in a hydrologic sense, is in progress. Both the number of days and the level of probability considered to be an uncommon event are arbitrary. Low-flow frequency curves have been determined from long-term records for six streams representing different areas affected by the drought in 1976-77. Frequency curves for selected numbers of consecutive days have been plotted in figure 2. Also, the discharges for 1976 and 1977 and the minimums of record are plotted. In figure 2a, the probability that the mini- mum average discharge for 30 consecut~ve days in a given year will be less than 26ft /s is 0.10 (Point X). Because the recurrence interval, which is the average number of years between events, is the reciprocal of the prob- ability, the flow will be less than 26 ft 3/s at intervals averaging 10 years. In 1977, the 30- day minimum discharge was 13 ft :rs which has a probability of 0.005 or a recurrence interval of 200 years. Because frequency curves are most reliable in the vicinity of the mean, a probability of 0.5, probabilities determined near the extremes should be considered only as reasonable estimates, not exact probabilities. The low-flow frequency curves and the flows in 1976 and 1977 at the six locations are not directly comparable. Differences may be caused by chance, by different basin charac- teristics, by different temperatures and pre- cipitation at various times, or by combinations of these factors. However, several interesting conclusions can be made. L The North Fork American River at North Fork Dam, 31 mi northeast of Sacramento, Calif., and the Beaver River near Beaver, 48 mi northeast of Cedar City, Utah had record minimum flows in 1977 at all num- bers of consecutive days selected. Prob- abilities were roughly 0.01 to 0.005, or the recurrence intervals are in the 100 to 200 year range. 2. Flows of the Sturgeon River near Sidnaw, 90 mi northwest of Escanaba, Mich., for periods up to 183 days had probabilities equivalent to recurrence intervals greater than 200 years. The probabilities are from long extrapOlations of the frequency curves and should be considered only as estimates. That the very low flows of the Sturgeon River are extremely rare events can be substantiated by the combination of two natural events. The drought in 1976 re- duced the sum mer and fall flows to a new record low just slightly more than half the previous low flow of record. Then the extremely cold winter struck the area, and flows remained low until the latter half of March 1977. Thus the lowest 183-day period started on July 23, 1976 and ended on January 21, 1977. 3. Flows of the other three streams in other years were significantly less than in 1976 or 1977 though none of them had prObabili- ties less than 0.012, about an 80-year recurrence interval. 4; Record low flow for the entire year was the damaging feature related to the drought. Records for low annual flows were set at four of the six streams, and the recurrence intervals are in the 100- to 170- year range. Whipple (1966) derived relations between the duration of a drought, in years, versus the probability of occurrence for five streams from Illinois to New Mexico and for three streams in Massachusetts. The relation curves are nearly alike for durations of 4 years or longer. On the basis of this small sample, the probability of a drought lasting 4 years is