The Cincinnati Rainfall Curve which is used in the design of sewers was adopted by the Cincinnati Engineering Department in the year 1913.

16 T2

The equation of this is I= where T is the time or duration of rainfall in minutes and I the intensity or rate of rainfall in inches per hour for the period of time specified. One inch of rainfall per hour corresponds very closely to one cubic foot per second per acre.

Summarizing Cincinnati records for the last 26 years, it is found that the curve was exceeded once for the 5 minute period, 3 times for 10 minutes, 4 times for 15 minutes, 3 times for 30 minutes and 2 times for 60 minutes.

Run-off

Run-off is the result of precipitation on a watershed. From the sewerage engineer's viewpoint it is the amount of surface drainage for which provisions should be made in the design of open or closed conduits or storm sewers.

In computing run-off, the main factors generally taken into consideration are as follows:

(1) Intensity or rate of rainfall in inches per hour, taken for a period of time which will produce the maximum run-off. This period of time is called the time of concentration and fixes the rate of rainfall to be used as shown by the adopted rainfall curve.

(2)

The area, shape and slopes of the territory to be drained.

(3) The relation between rainfall and run-off, or the proportion of rainfall reaching the sewers from surface drainage, or in other words, the percentage of imperviousness of soils or surfaces from which the run-off occurs.

On a naked hillside area, the percentage of run-off is usually 10 to 20 per cent. of the rainfall, the high percentage of run-off prevailing where the hillside is quite steep with scanty soil covering underlaid rock and shale formation.

In a suburban residential district, the coefficient of imperviousness is found to be about 30 per cent.

In a compact residential district, with a house on every lot, with four large or five smaller houses to the acre, the run-off is found to be 40% of the rainfall.

In factory areas, the percentage of run-off is usually found to be about 80 per cent of the rainfall.

In mercantile districts such as the Walnut street area between Seventh street and the Ohio river, it has been found from actual sewer gagings that the run-off is 85% of the rainfall.—(Excerpts.)

Discussion: Sir Frederic Stupart: "Rainfall intensity in Toronto is less than in Cincinnati. Topography plays a very important role. It is necessary to study it separately for every city where drainage is important."

Professor H. J. Cox told of the unprecedented torrential rain of Aug.te 11, 1923, in Chicago, when 2.30 inches fell in 80 minutes in the early morning. The flooding of the sewers was probably the worst experienced in the history of the city, and the fire department was kept busy all day pumping water from basements and other low places, only to have them

flooded again in the evening, when approximately another inch of rain ar fell in about an hour. The total rainfall, midnight to midnight, was 3.70 inches.

tio

已

duit

ons

eri

nf

Mr. G. A. Loveland remarked on the need for a formula for run-off, and said that with increase in paved area more sewers are needed. Professor W. I. Milham inquired if there was a factor of safety,-i.e., were sewers built for the usual or the extreme maximum?

Mr. Sullivan replied that the sewers were designed in accordance with the curve, and that a flood from inadequate sewers was expected once in about eight years. The summarization in the excerpts printed above,

for example, shows that in 26 days, half hour falls exceeded the sewer

capacity on three occasions.

Some of the River Problems the Meteorologist Can Best Handle

By W. C. DEVEREAUX, Meteorologist, Weather Bureau

It has been recognized practically since the beginning of the Weather Bureau work that the meteorologist is in the best position to forecast the floods in the rivers and, in fact, to forecast all changes in the height of the water whether that be at a high or low stage. This is natural and as it should be; the weather and the river forecasts go hand in hand, they cannot be separated. The forecasting of the river stage is the most important phase of river work, just as the forecasting of the weather and storms is the most important work of the Weather Bureau.

Those not entirely familiar with the river work of the Weather Burti reau believes that it ends here,--when the river forecasts and flood warnings are made and fully distributed. Those holding this belief are most of the authors of text-books on Meteorology as well as many of the officials of the Weather Bureau. This is no fault of those authors or of10 ficials but is rather the result of the failure of the "river man" fully to giadvise of all of the work he is doing. There is presented here only a few od of the "other” river problems that have been handled at the Cincinnati station during the last 12 years. In solving most of these problems as well as many others we have had the splendid co-operation of the United States Engineers, and we in turn have assisted the Engineers in solving many of their problems.

Great ice gorges do not form frequently in the Ohio river and that makes the forecasting of both the forming and breaking of the same the more difficult due to lack of experience or knowledge. The great ice gorge below Cincinnati in 1918 destroyed more property in this region than any of the great floods. The first gorge that year was of the more usual type. The river was frozen nearly its entire length and the break in the ice proceeded down the river as the flood wave advanced, until the ice jammed in the great bends at Sugar Creek and formed a dam 61 feet high which backed the water up the Ohio river for more than 100 miles.

This dam held for 12 days during which time the water above fell 25 feet, but it broke with the next flood wave when the water rose to 62 feet, the break not being due to the great pressure behind it but to the higher temperature, the 50-mile gale blowing against the face of the gorge, and the bright sunshine.

Forecasting the breaking up of these gorges was not so difficult a problem as was the question of whether or not to issue warnings of an im

pending break in a tributary three days before the final break in the Ohio. The Licking river enters the Ohio directly opposite the business center of Cincinnati. A great flood 55 feet high, higher than ever before known, was moving slowly down the river breaking the ice and forming great gorges. It was 40 miles from Cincinnati; the next report showed it to be 30 miles away. Would it come through and throw a wall of water 10 to 20 feet high across a frozen Ohio and a flooded Cincinnati? The city could be warned in a short time by special editions of the newspapers, the telephone, and the police force, but night was approaching and action, if any, must be taken soon. The bends in the river and the cold weather held the gorge, and only a few knew of the impending danger.

The great government dams in the Ohio river are wonderful works of engineering. The weather determines the rate at which they can be built, and after being built it determines largely when and how they shall be operated. The running of "artificial waves" the last few years, which bring down large shipments of coal from the Kanawha, depends on the weather.

Other work on the river, possibly the outlets for a city sewer system, must be done when the river is low. What is the period, on the average, most favorable for the work? The Weather Bureau has the record available.

A business firm is considering locating an important plant a short distance up a tributary and wishes to use the river for transportation. The questions in this case are not only how many days each year the river may be above a certain stage but also how many days will navigation be interrupted by high water and by ice.

We are frequently requested to determine how many times a certain piece of property has been flooded during the last 50 years. The property may be a considerable distance from any river gage and we must consider the slope of the flood plane as well as the height of the river. Such information may be requested in connection with the sale of the property or in condemnation proceedings. The river records like the weather records are frequently in court.

A few years ago the river and weather records were taken to court in Ohio. The case involved timber lands in Arkansas and had been on trial for nearly a week. Witnesses had stated that the land was flooded during all of February and one witness stated that he had rowed over the property on or about Washington's Birthday. The river records showed that the river flowing through the property was not half full at any time during the month. The case was settled that same evening out of court. These problems and many similar ones are on the edge of the science of Meteorology, they involve Hydrology and Engineering, but who can solve them as well as the Meteorologist? We speak of this service as the "river work" but that is a poor and inappropriate expression. But what word describes this phase of science? The only one I can suggest is Hydro-Meteorology.—(Author's summary.)

Some of the River Problems the Meteorologist Can Best Handle By MONTROSE W. HAYES, Meteorologist, Weather Bureau

A sharp line of demarcation between closely related activities usually does not exist, but in hydraulics it seems to be generally conceded that

eOk

enter

know

it to

r 1

acti

precipitation run-off and river gage relations, in so far as they apply to river forecasting, properly belong to the meteorologist. River stages are so intimately associated with precipitation, temperature, and at times wind velocity and direction, the fluctuations of the stages can be forecast in a satisfactory manner only by one having at his command extensive and fresh meteorological information.

hec The interests that can be served by the river stage forecaster are in a pape great measure peculiar to the locality in which they lie, but the work being done in Cincinnati, which has been so well described by Mr. Devereaux, is representative of the tasks confronting all the forecasters on large streams, and it is unnecessary to consider the matter in greater detail.

eati

Ork

can It might be of interest, however, for a brief mention to be made of forecasts of the first run of ice at St. Louis. Contrary to the general belief, the winter's first run of ice does not come from far upstream. Alon most invariably the first floating ice to appear at St. Louis forms at St. Louis, or a few miles upstream. This ice is thin; river men call it "window glass" ice, but it is dangerous to light river craft and some enengineering equipment. In the last five winters the Weather Bureau has been able to forecast the first floating ice 24 to 48 hours in advance, and the warnings given have been timely enough to enable craft and equiptiment to be moved to safe places.

There is another phase of the subject of river stage forecasting that seems to have a place here, although it does not come directly under the title of the paper. It is the wonderful opportunity we, as meteorologists, have to develop mathematical working formulas for forecasting fluctuaer tions in river stages. There are no texts on the subject. There is nothing in text books that have any direct bearing on the work. Hydraulic engineers have their run-off formulae, their Bazin, their Chezy, their Francis, their Humphreys and Abbott, their Kutter, and their various of weir formulae for discharge work, but none of these formulae is directly applicable to the forecasting of river stages. In forecasting stages we must consider the rate at which rain has fallen, as well as the amount; the varying tributary increment, which is highly important; the constant t change that is taking place in the thalweg, a change that is very marked in silt-bearing streams; the variation of the slope from the normal, if er there is such a thing as a normal slope, and numerous other factors that may be peculiar to the stream under consideration.

t It is not thought possible to find a general law governing river fluctuations, for the vast amount of work done on river discharge has developed nothing more than formulae that are frankly empirical, and have the disadvantage of carrying a coefficient that can not be determined easily and accurately, and it is obvious that estimating gage heights is more difficult than determining the discharge through a given crosssection. However, it does seem that we, also, should be able to develop empirical formulae to guide us in our work. No criticism of the stage forecasting that is done now, and has been done for many years, is implied or intended, for this work has been of a high order. It has saved many human lives, and the millions of dollars in property it has saved can not be estimated. It has the confidence of the entire country, and the public may not think it in need of improvement. Those of us

who are familiar with the inside of this work are gratified by the esteem in which the river and flood work of the Weather Bureau is held, but we are well aware of the regrettable fact that the men making the river stage and flood forecasts are able to do the work satisfactorily only because they have devoted years to the study of the streams in their districts. We need something to guide a new forecaster; something to give him confidence when he takes charge of a district, without the necessity of waiting until he has gained the confidence through the experience of several years.

The River and Flood Division of the Weather Bureau, in Washington, is doing valuable work in this direction, but the field is too big, and it will take too long to develop formulae for the entire country; and because the field is so big the opportunity is presented for Weather Bureau officials of a mathematical bent to do something really original and of high value, by aiding in placing this work, that Mr. Devereaux calls "hydro-meteorology," on a sound mathematical basis.—(Author's summary.)

Discussion: Sir Frederic Stupart: "River forecasting is very important. In Canada there is flood forecasting on the Fraser and Saskatchewan rivers. The river forecasting work is very much more important here than in the Canadian West." Sir Frederic agreed that a mathematical formula should be developed so that new men could use it.

Mr. J. H. Armington doubted if a practicable mathematical formula could be developed, for, in the upper reaches particularly, a forecast of the amount of rainfall comes in.

Dr. G. W. Littlehales said that at the International Geophysical Union, Rome, 1922, a section on scientific hydrology was proposed for developing mathematical foundations. This came from the section of Continental Hydrography. The American Geophysical Union deferred the creation of such a section, since Americans think these studies should be developed in connection with meteorology.

Mr. W. H. Alexander remarked that the lowest as well as the flood stages must be indicated right along.

Mr. J. L. Kendall emphasized the immediate importance of rainfall in filling smaller streams, and therefore, in affecting river stages in a very short time. The prediction of such stages shortly after rains occur is difficult and much work is needed on this problem.

Professor H. E. Simpson remarked that on the Red river of the north there was an international problem concerning flood forecasts. Such forecasts, however, in the last decade had been needed only once, in 1916. The forecasts on that occasion were very accurate. It would be possible on a similar future occasion to give warnings for the lower portions of Winnipeg.

Mr. M. W. Hayes said if we can estimate river stages, as at present, to feet or tenths, we can develop a formula. Near headwaters, however, we cannot use the formula. Wherever a forecast is made, a formula could be developed.

Sir Frederic Stupart said he was glad that Dr. Littlehales reported this question should be left in the hands of meteorologists.

Mr. W. C. Devereaux said the U. S. Weather Bureau made a mistake a few years ago in letting a fair portion of the river observations get transferred to another agency. With respect to the possible accuracy of