SCIENCE 



NEW YORK, APRIL 1, 1893. 



STORAGE OF STORM- WATERS ON THE GREAT 

 PLAINS. 



Somewhat exaggerated expectations have been aroused by 

 the speculations of certain theorists in regard to the possi- 

 bilities of water storage on the high, wind-swept, treeless 

 plains lying between the 98th meridian and the Rocky Moun- 

 tains. These visionaries have virtually promised every 

 farmer a reservoir on his land if he would only make the 

 effort to secure it. 



The need of storage, if it can be made a success, is indis- 

 putable. Rivers are few, and, as a rule, inadequate to the 

 irrigation of more than the lands of their own valley. Arte- 

 sian wells are limited to certain sharply defined basins. 

 Other wells are generally too deep for profitable irrigation 

 by pumping, except for small plats of fruit and garden vege- 

 tables. If the mesas are to be extensively irrigated it must 

 be by storage of storm-waters. Can it be done ? If there is 

 any doubt about it we would better know the truth than to 

 encourage delusive hopes. Let us seek some quantitative 

 numerical expression for the possibilities and limitations of 

 storage. 



The great robber of moisture on the plains is evaporation. 

 The activity of the winds is so great and constant that more 

 vapor is raised from exposed water surfaces than in many 

 regions of greater heat. The annual evaporation is seldom, 

 if ever, less than four feet, and may rise to eight feet. We 

 may safely put the average as high as five feet. 



The rain fall varies from one to two feet. Its seasonal 

 distribution is favorable, the late spring months and the 

 summer months receiving the greatest amount. So far as 

 the quantity and seasonal distribution of the rainfall are 

 concerned the chances of impounding some of it look en- 

 coui'aging. But it is not so much the aggregate of precipita- 

 tion as the percentage of -it which flows off on the surface, 

 which determines the feasibility of storage. In a treeless 

 region of great evaporation and porous soil atid subsoil, the 

 total run-off is always low, and much of that is subterranean. 

 Humphreys and Abbott, in their report on the Mississippi 

 River, estimate the total run-off of the Missouri valley at 15 

 per cent of the rain fall. This includes the springs which 

 feed the nvers, as well as the superficial run-off. This sub- 

 terranean factor is unusually large on the plains, because 

 there are large areas on the mesas and among the sand hills, 

 which have no surface streams. All moisture reaching the 

 rivers from these areas percolates beneath the surface, and 

 the superficial run-off is by so much diminished. 



Again, if the average for the whole Missouri valley is 15 

 per cent of the rain fall, it is less than that on the plains, 

 because the whole basin includes wooded areas and steep 

 mountain slopes, from both of which the run-off is more 

 than the average. If we reckon 7.5 per cent as the superfi- 

 cial run-off of the plains, that will certainly be as favorable 

 as the considerations just presented will possibly admit. It 

 is more likely to be too high than too low, for fully half, if 



not more, of the run-off is subterranean, and the total is less 

 than 15 per cent, while we have allowed half of 15 per cent 

 for surface flow which may be impounded. 



The third important consideration is seepage. A reservoir 

 may be made absolutely water-tight, but it is not likely to 

 be. Rather is it absolutely certain that for small storage on 

 the farm, executed without the aid of professional engineer- 

 ing, and under rigid conditions of economy, so that cement- 

 ing, or puddling with clay, is out of the question on account 

 of the expense, the loss by seepage will always be consider- 

 able. The possible variations of such loss are so great that 

 we can do no better than to make a somewhat arbitrary as- 

 sumption of its amount, say two feet annually. If the site 

 is so badly selected, and the dam so poorly built, that the 

 water will.be lowered more than two feet annually by per- 

 colation, success is improbable; on the other hand, less than 

 two feet would be too small a margin to allow for seepage 

 under the circumstances. More would be fatal, and less is 

 improbable. 



The fourth consideration is the ratio of catchment basin to 

 reservoir surface. This factor is more under human control 

 than the others. At first blush it might be thought to be 

 wholly a matter of choice. And so it is if the reservoir is 

 artificially excavated. It may be dug deep and narrow to 

 prevent evaporation. Its surface may be made only one- 

 millionth of the catchment basin, if that is desirable. But 

 the economy of water storage for irrigation will not admit 

 of more excavation than that required to procure earth for 

 the dam. Aside from the cost of digging it, a deep pit 

 would require a pump to raise the water. Natural depres- 

 sions must be utilized. But these are always broad and 

 shallow on the uplands. Deep caiions and valleys are ex- 

 cluded because they are below the lands to be irrigated. 

 They may answer for the valley lands below them, but not 

 for the table- lands which we ai'e considering. In the wide 

 shallow basins of the uplands, if the waters have any con- 

 siderable depth, they will spread abroad, cover much good 

 land, and lose much by evaporation. But they must have 

 considerable average depth throughout the year for two 

 reasons. The maximum depth will occur after storms, the 

 minimum during periods of drought. Unless the average is 

 high it may readily happen that little or no water is avail- 

 able just when the crops need irrigation. Furthermore, the 

 depth should be considerable, or else the reservoir will flood • 

 nearly as much land as can be irrigated from it. E. S. Net- 

 tleton, chief engineer of the Irrigation survey, U. S. Dept. 

 of Agriculture, estimates that an annual average of nine 

 inches of water over the whole surface of the field will be 

 required for successful irrigation on the plains. One acre of 

 reservoir with an annual average depth of four and one-half 

 feet will therefore irrigate six acres of land. The alue of 

 the flooded land will absorb the profits of the operation if 

 the ratio is greater than that, that is, if the depth of water is 

 less. 



It is evident that when water is impounded in natural de- 

 pressions on the table-lands the reservoir will necessarily 

 cover a considerable fraction of its catchment basin. Take 



