Sounding the Ocean of Air / Being Six Lectures Delivered Before the Lowell Institute of Boston, in December 1898

by Abbott Lawrence Rotch

Plate IV.

The identity of cloud-forms all over the world has been established, and as a result of the measurements at Blue Hill, the heights and speed of all clouds observed there are known. The averages have been plotted in the five levels into which we separate the clouds in [Plate IV]., Heights and Velocities of Clouds, where ordinates represent heights and abscissæ velocities, and, consequently, the distances of the various forms of clouds above the horizontal base indicate their heights, and the distances from the left-hand vertical line their velocities. For comparison, the velocity of the wind on Blue Hill, a few hundred feet above the general level of the country, is represented. The mean height of the cirrus is about 29,000 feet, but this cloud sometimes reaches 49,000 feet. The mean height of the cumulus is about a mile, but the tops of the cumulo-nimbus, or thunder-shower cloud, sometimes penetrate into the cirrus level. Generally the base of the nimbus, or rain cloud, is only 2300 feet above the ground, and it frequently sinks below the top of Blue Hill, which is only 630 feet above the sea. The poetic saying, that "Earth wraps her garment closer about her in winter," has a scientific basis, for the average height of all the clouds is greatest in summer and least in winter. But the reverse is true of their velocity, for the entire atmosphere moves twice as fast in winter as it does in summer, and at the lower levels the seasonal change is even greater. The average velocity of cirriform clouds is ninety miles an hour in winter, and sixty miles in summer, but occasionally in winter cirrus have been found to have the enormous velocity of two hundred and thirty miles an hour. In the average, the velocity of the currents increases, from the lowest to the highest clouds, at the rate of about three miles an hour for each 1000 feet of height, but near the ground the increase with height is faster. It has been found that the velocity of the lower clouds is less than the velocity of the wind on a mountain of the same height, which may, perhaps, be explained on the supposition that the mountain acts like a dam to accelerate the flow of air over it. The measurements in Sweden showed that the middle and upper levels of clouds are higher than in America, but that they move less rapidly. This may be because the surfaces of equal temperature in the air are higher in the United States than in Sweden, on account of the direction of the upper currents, while the greater velocity of our high clouds corresponds with the more rapid movement of areas of low and high barometric pressure over the United States.

These results are suggestive. For instance, the energy of the upper half of the mass of the atmosphere, or that portion which lies above 18,000 feet, has been calculated to possess six times the energy of the lower half in which we live, and as yet, none of this enormous store of energy is applied to the use of man. While it appears certain that no navigable balloon or flying machine will ever be able to stem the enormous velocity of the upper atmosphere, rarified though it is, perhaps in the future aërial machines will take advantage of the prevailing currents of the high atmosphere, as our sailing ships do of the trade winds. The observations of cirrus clouds in various parts of the world show that they always move from a general westerly direction, while below this primary drift toward the east occur the relatively permanent or transient differences of pressure which cause the deviations from the normal circulation of the atmosphere, and give rise to the local circulation in storms. In the familiar daily weather map it will be noticed that there is usually some portion marked "low," and another portion marked "high." The former is an area of low barometric pressure, into which the winds at the ground blow spirally inward in the [opposite direction] that the clock hands turn; the latter is an area of high barometer, out of which the winds at the ground blow in the contrary way. The former when well developed are called "cyclones," and are usually accompanied by stormy weather, and the latter, called "anti-cyclones," bring fair weather. From the observations of the directions from which the clouds move in cyclones and anti-cyclones, we have found that above the cumulus level (at the height of about a mile) the inward inclination of the wind in a cyclone, and the outward inclination in an anti-cyclone, both disappear, and the general drift from the west prevails. The results of the observations are shown in [Plate V]., Atmospheric Circulation in Cyclones and Anti-cyclones, representing sections of the atmosphere, concentric to the earth's surface, in the five cloud-levels seen from above. The arrows fly with the wind and are proportional in length to its velocity, the dotted arrows indicating the probable flow of the air through the cyclones and anti-cyclones that are indicated by the circles, their axes being assumed to be nearly perpendicular to the earth's surface. Above the cumulus it will be observed that the wind in the cyclone tends to come from the south-west in front and from the north-west in the rear, while in the anti-cyclone the contrary is the case, indicating a deflection of the westerly upper current to the right in cyclones and to the left in anti-cyclones. This sustains the theory that the cyclonic circulation is struggling against a general atmospheric drift from the west which increases with altitude, and above the height of a mile becomes greater than the cyclonic influence. Higher than this, the atmospheric circulation is controlled primarily by the permanent temperature gradient between equator and pole, by the seasonal temperature gradient between ocean and continent, and, in the United States, by the passage of "warm and cold waves." Mr. Clayton's investigations indicate that the motion of the upper clouds is nearly parallel to the lines of equal temperature at the earth's surface. A high temperature, by expanding the air upward, causes in the upper air a high pressure; and a low temperature, by contracting the air towards the ground, causes in the upper air a low pressure, so that the lines of equal pressure in the upper air are parallel to the lower lines of equal temperature, and since there is little friction in the upper air the motion of the wind is nearly parallel to the lines of equal pressure. Below the cumulus level the winds follow the normal cyclonic and anti-cyclonic circulation. There are two theories of the origin of these areas of high and low pressure, the "driven theory" which supposes that they derive their energy and drift from the general atmospheric movement from west to east, and the "convectional theory" which attributes their formation and progression to the difference of temperature between them and the adjacent air. While the observations on mountains have favoured the driven theory, yet the inward spiral motion of the cirrus clouds above the anti-cyclone, indicating a lower pressure than in the surrounding air, contradicts the hypothesis, and the recent observations with kites at Blue Hill strongly support the convectional theory of cyclones.

Plate V.

The relation of the clouds to weather forecasting has been investigated at Blue Hill. For instance, it is found, in this region at least, and contrary to the general opinion, that cirrus clouds do not indicate rain, but do foretell a change of temperature that is proportional to the rapidity of motion of the clouds. Alto-cumulus is followed by rain within twenty-four hours three times in four. Rain follows the appearance of all high and intermediate clouds most frequently when the cloud banks are densest toward some westerly point and when they come from the west. Mr. Sweetland, an assistant, has studied special forms of cloud in their relation to the succeeding weather. He concludes that cirrus plumes precede fair weather, while dense clots of cirro-cumulus are followed by rain. Rounded pendants, or mammillated clouds, in the lower levels indicate rain, but in the upper levels fair weather. Of all the forms, the dark sheet of stratus, and clouds of lenticular shape, are most frequently followed by rain. Of clouds presaging changes in temperature, the turreted cumulus, which is connected with thunder-storms, precedes the greatest fall in temperature, and next in order come lenticular clouds, flaky cirrus, and alto-cumulus. In general, flat and flaky clouds, clouds forming and disappearing rapidly, and clouds changing to forms at a higher level precede dry and cooler weather.

It will be seen that this modern study of clouds as prognostics simply adds to the weather proverbs that have come down to us from the time of Theophrastus. It does not appear, however, that cloud forms alone can usually serve to predict rain for more than twenty-four hours, but for a few hours in advance the appearance of certain cloud forms frequently furnishes the observer more trustworthy signs of coming rain than does the synoptic weather map. To a forecaster in possession of telegraphic data, the prevalence of rapidly-moving cirrus over a wide area indicates a rapid storm movement, with sudden and marked changes of weather and of temperature, while slowly-moving cirrus indicate slight changes of temperature and dry weather. The direction of the cirrus movements in front and around a storm centre will usually point out the future movement of the storm, which tends to advance in the same general direction.