Although the flow of Colorado River has been controlled at Boulder Dam since February 1935, flood danger still exists in the basin below the dam. This report on the first general floods to occur below Boulder lam since the dam was closed presents facts that should prove helpful in planning protection and reservoir operation to minimize the ill effects of future floods.

The floods of September 1939 were caused by a series of tropical disturbances that moved northwestward along the west coast of Mexico and culminated in unprecedently heavy rains in northwestern Mexico and southwestern United States. Three separate storms, occurring September 3-7, 8-13, and 23-26, moved across the lower Colorado River Basin. At many points in that area the mean annual precipitation was exceeded by the precipitation for September.

Because little rainfall preceded the storms, runoff from them was less than would have occurred under more unfavorable conditions. At streamflow measuring stations where past records are available no new records were set. On Gila River no flood occurred. On Colorado River the flood was not so great as those that had occurred almost every year prior to the closing of Boulder Dam. On Williams River, however, the peak discharge was of the magnitude of a major flood, and in many of the smaller drainage basins peaks occurred which probably have not been exceeded in the previous 50 to 100 years.

The relatively low flood peaks on Colorado River do not mean that there was no flood danger. The regulation of the river since Boulder Dam was closed has prevented the scouring of the channel by floods, has permitted the encroachment of vegetation in the channel, and has allowed the accumulation of sediment. As a result, flood stages today are about as high as they formerly were for discharges twice as large.

Flood peaks in Colorado River were greatly reduced by storage. Between September 5 and 20»Lake Mead stored 330,000 acre-feet of water. Between September 4 and 7 Havasu Lake stored 135,000 acre-feet, and the channel storage between Topock and Yuma, exclusive of Havasu Lake, accounted for about 110,000 acre-feet at the peak.

Storage in Havasu Lake is limited in relation to the flood flows that may enter it. When the September storms began, Havasu Lake was at normal minimum level. Had it not been low, the lake probably would have filled by the morning of September 6, and that day's peak inflow, which may have exceeded 75,000 second-feet, would necessarily have passed through with little reduction.