the storm centers, or any number of them, should pass to the southward of the Ohio Valley, the precipitation over that district would be much less in quantity and the temperature would be much lower, thereby preventing the melting of any considerable quantity of the snow that might be on the ground. Consequently the lower Mississippi flood would pass into the Gulf of Mexico without unusual incident unless additional heavy rains should bring the western tributaries to very high stages-a comparatively infrequent occurrence and one not of any considerable moment so far as the lower river itself is concerned—as the Ohio River is the principal factor in the question. THE FLOOD OF 1912. The annual rise of the lower Mississippi River for the year 1912 began on February 21, at which time a severe storm from the Southwest was moving northeastward over the lower Mississippi and the Ohio Valleys, attended by general and heavy rains. A second storm of similar character moved northeastward four days later, and the rise was well under way in the lower Ohio River and in the lower Mississippi between Cairo and the mouth of the Arkansas River. By the end of February the Ohio River was above the flood stage from Evansville, Ind., to Shawneetown, Ill., while the Mississippi was rising rapidly as far south as the mouth of the Red River, and had nearly reached the flood stage of 34 feet at New Madrid, Mo. The winter had been a cold one, without much snow over that portion of the Ohio watershed where it would have remained on the ground for any considerable period after it had fallen, and it so happened that on February 26 there was no accumulated snowfall over any portion of the Ohio River watershed except a negligible quantity over central and southern Indiana and southeastern Illinois. There were no heavy rainstorms probable in the near future, and consequently no immediate fears of a great flood. While rains and snows were comparatively frequent during the first decade of March, they were not heavy, and after March 4 the Ohio River at Cairo began to fall after reaching a stage of 41.8 feet, 3.2 feet below the flood stage. Shortly afterwards the Mississippi River began to fall at New Madrid and the fall continued for about 10 days, but did not extend below the mouth of the Arkansas River, as the occasional rains were sufficient to maintain the original rise from above. On March 10 an extensive barometric depression moved in from the Pacific Ocean to southern California, and by the following morning it had reached Kansas, with a secondary storm center reaching down over southeastern Texas. During the day (March 11) the general disturbance moved eastward and northeastward over the normal path of southwestern storms, and moderately heavy rains, averaging less than 1.5 inch, fell over Louisiana and Mississippi. Over the Ohio Basin the rains, while well distributed, were not heavy, and some light snow fell, the quantity being just about sufficient to maintain the average depth of from 1 to 3 inches that had covered that portion of the Ohio watershed north of Tennessee since the earliest days of March. While this storm was moving across the country another appeared on the north Pacific coast. It moved southeastward over the districts west of the Rocky Mountains, and by the time (March 14) that the preceding storm had passed off the Newfoundland coast, the second one had moved to Kansas, accompanied, like its immediate predecessor, by a secondary depression extending southward over southeastern Texas. During the succeeding 24 hours rains again fell over Louisiana and Mississippi and extended into the lower Ohio Valley, with lighter rains above. The heaviest rains fell over the Yazoo Valley, but the quantity was not excessive. The rains did not continue for more than 24 hours, as a rule, but they were sufficient to take out the little snow that covered the Ohio watershed and to augment a rise in the upper Ohio River that had set in after the rains from the preceding storm, and also to start another rise in the lower Ohio and in the Mississippi from Cairo to New Madrid. At the same time the Missouri River east of Kansas City and the Mississippi from Alton, Ill., to Cairo began to rise. This second storm passed into the north Atlantic Ocean during the night of March 15-16, and during the next four days there was no precipitation of consequence over the Ohio and lower Mississippi watersheds. On the morning of March 19 a disturbance was central over Utah, and at this time some of the river stages were as follows: At Helena the second rise of the month was just about to begin. The Utah depression, the third storm of the month, moved over the Ohio Valley by way of Oklahoma and Arkansas, somewhat to the northward of its immediate predecessors, with moderate rains and snows above Cairo, but with little or none below. The precipitation from this storm apparently had no effect other than to check the decline in the upper Ohio River, and the storm center passed into the north Atlantic Ocean during the night of March 21-22. However, another storm had appeared over Arizona, and an offshoot from it traversed the usual path over Texas, reaching the Ohio Valley on the morning of March 24, when the Missouri River east of Kansas City, the Mississippi from Alton to New Orleans, and the Ohio, Cumberland, Tennessee, and Wabash Rivers were rising steadily. The rainfall resulting from this fourth storm was the heaviest of the month, beginning on March 22 over Louisiana and southern Mississippi and extending northward and eastward during the two following days with an average fall of 2.5 inches over the lower Mississippi Valley and somewhat less over the Ohio Valley. At the same time a heavy blanket of moist snow was deposited over Missouri and Kansas, with an average water equivalent of over 1 inch. There was also some snow, probably sufficient to make one-half inch of water, over the northern portion of the Ohio River watershed. As the soil was deeply frozen after the long and cold winter, this large snowfall was equivalent to another heavy rain, and it must necessarily run into the rivers as soon as the temperatures rose to normal conditions or another rainstorm came. Should the high temperatures and the rain come coincidently, as they do during southwestern storms, conditions would become still more threatening. There were no longer any doubts that a severe flood would occur. The only question was as to the final outcome which was now entirely dependent upon the contingency of additional heavy rains in the near future. If there should be more the flood heights must certainly prove to be the greatest of record from Cairo to the Passes should the levees remain intact. The suspense was not prolonged, for on March 26, one day after the fourth storm moved into the Atlantic Ocean, the fifth one appeared over Nevada. The river stages at various places and the changes in one week from Helena northward, due to the third and fourth storms, were as follows: TABLE III.-River stages, Mar. 26, 1912. Cincinnati, Ohio.. Evansville, Ind. New Orleans, La.. Owing to stagnant pressure conditions in the extreme West, the fifth storm did not reach Texas until the morning of March 28, when rains were falling over the great central valleys. Approximately 1 inch of rain fell, with the greatest fall below Cairo; and the storm also melted and brought out the heavy, moist snow that had fallen during the preceding storm, so that the run-off was practically doubled. It was now certain that, should the levees hold, all previous high-water records from Cairo to Memphis would be exceeded, and equally certain that another heavy rain would result in similar conditions from Memphis south to the mouth of the river. Again the period of suspense was short, for on the evening of March 29 another disturbance, the sixth and last of the remarkable series, was central over Utah. It moved more slowly than its predecessors and did not reach eastern Texas until the morning of April 1, moving then to the northeastward over the Ohio Valley. It happened, unfortunately, that the rains from this storm were heaviest over those sections where they were least desired. Over Louisiana and southern Mississippi, where another inch or so of rain would not have changed conditions materially, the rainfall was light, while to the northward almost as far as St. Louis and to the eastward over Tennessee and Kentucky it was heavy, especially over eastern Arkansas and the Yazoo Valley. As a result the rate of rise in the lower Ohio and the lower Mississippi was maintained or increased, the Cumberland passed the flood stage, the Tennessee rose more rapidly, while the Missouri east of Kansas City, the Mississippi from Hannibal to Cairo, the lower Arkansas, the White and the Black Rivers also passed the flood stage. There could be no further doubt, and warnings were issued that the coming flood heights from Cairo to the Gulf of Mexico would be the greatest in history. The stages at this time and the changes in one week were as follows: The rains continued during April and May with more or less frequency, and at times they were heavy, but they do not appear to have done more in the main stream than to prolong the high-water period, as in most instances the loss of water through the crevasses more than offset the effects of the rains. An exception was noted at New Orleans, where during the evening of May 10 nearly 7 inches of rain fell, forcing the river up to the remarkably high stage of 22 feet, or 4 feet above the flood stage, a few hours later. High southerly winds, however, were an important factor in causing this stage, and a decline to the actual high-water level followed shortly afterwards. Among the appendices to this report will be found copies of the United States Weather Bureau charts (Nos. 2 to 7, inclusive) showing the storms directly contributory to the flood, and also copies of the Weekly Snow Charts (Nos. 8 to 11, inclusive) showing the depth of snow on the ground at different times during the flood. DURATION OF THE FLOOD. As the Ohio River at Evansville, Ind., reached the flood stage of 35 feet on March 18, the flood may be considered to have begun on that day. The Mississippi River at New Orleans, La., fell below the flood stage of 18 feet on June 8, and this day may be considered to have been the last day. This would make the total duration of the flood 83 days. In the Atchafalaya River the flood lasted until June 19, on which day the river at Melville, La., fell below the flood stage of 37 feet. Flood stages in this river, however, were not recorded until April 9, when the stage of 41 feet was first reached at Simmesport, La., making the flood duration 72 days. The Yazoo River was in flood for 75 days, beginning on April 1 and ending June 14, while the lower Ouachita River was in flood for 62 days, beginning on April 4 and ending on June 4 at Monroe, La. The first rise began on February 21 with the advent of a typical southwestern storm, but early in March it was followed by a decline that continued for a week or 10 days as far south as Helena, Ark. Below the mouth of the Arkansas River the initial rise was practically unin terrupted. The duration of the flood was materially increased by some abnormally heavy rains that fell over the entire watershed east of Kansas City during the last few days of April. These rains sent a second flood wave down the river, with resulting secondary crests only 0.5 foot below the original one at St. Louis, but 5 to 6 feet lower from Cairo to the mouth of the St. Francis River. From Helena southward the effects consisted only of a considerable prolongation of the original flood wave. The hydrographs (Diagram III, appendix) show the actual conditions that prevailed from March 1 to June 10, inclusive. Hydrographs of the floods of 1882, 1897, and 1903 are also shown. (Diagrams IV, V, and VI.) From these hydrographs it will be seen that the Ohio River at Cincinnati developed two distinct major crests, one on March 27, when the stage was 53.4 feet, and another on April 5 and 6, when the stage was 51.7 feet. There was also a minor crest of 36.7 feet on April 30 following a decline to 21.9 feet on April 26. In all, the Ohio River at Cincinnati was above the flood stage of 50 feet on 10 days. There were two crests in the Mississippi River at St. Louis, one of 30.8 feet on April 5, and another of 30.3 feet on April 30, the latter resulting from the heavy rains of April 25 and 26. The river was above the flood stage of 30 feet on 10 days. At Cairo the Ohio River first reached the flood stage of 45 feet on March 22, and from that time until the end of the flood there were two crests, one of 54 feet on April 6 and 7, and another of 49.3 feet on May 4 and 5. The stage of 54 feet on April 6 and 7 was 1.8 feet higher than the previous high-water record of February 27, 1883, and only the failure of the levees in the immediate vicinity and below prevented a crest stage of at least 55 and possibly 56 feet, the latter stage representing the maximum capacity of the Cairo City Levee. Flood stages prevailed for 45 days in all. At Memphis there were also two crests, corresponding with the conditions at Cairo. The river first reached the flood stage of 33 feet on March 24, and did not fall below it for 60 days. The great crest occurred on April 6, when the stage was 45.3 feet, 5 feet above the previous high-water mark of February 3, 1907; and here again the failure of the levees prevented a stage of at least 47 feet, or nearly 7 feet higher than the stage of 1907. The second crest of 38.9 feet occurred on May 10. At Helena the flood stage of 42 feet was passed on March 26, and a stage below 42 feet was not again reached until May 26, making a total of 62 days that the river was above the flood stage. The highest stage occurred on April 21, and was 54.4 feet, 2.6 feet above the previous high-water record of April 4, 1897. There was but one crest after the great rise set in. At Vicksburg the flood stage of 45 feet was reached on March 31, and the river was at or above the flood stage until May 31, a period of 62 days. There were two crests; one a principal one of 52.1 feet on April 12, and a secondary one of 48.4 feet on May 6 and 7. Heavy rains over the Yazoo watershed were responsible for the second crest. The high stage of 52.1 feet was 0.4 foot lower than the record stage of April 16, 1897, the Panther Forest and Salem crevasses being responsible for the deficiency in 1912. Had these levees remained intact, the crest stage at Vicksburg would have been 53.5 feet or 54 feet, and, if only one of the two levees had held, it is very probable that the high-water mark of 1897 would have been exceeded somewhat. At New Orleans the river was above the flood stage of 18 feet from April 10 to June 8, inclusive, a period of 60 days. There was a single crest of 22 feet in the early morning of May 11. This was 1.6 feet above the previous high-water record of April 6 and 7, 1903, and was partially due to an unfortunate combination of high southerly winds and torrential local rains. Conditions above New Orleans had indicated a maximum stage of 21.5 feet, which was the stage actually reached after the effects of the high winds and local rains had disappeared a few hours later. Records were also exceeded in the Atchafalaya River, the excess ranging from 2.2 to 2.8 feet. The Ouachita River failed by nearly 3 feet to equal the record of 1874. 14442-13--2 |