Now bear in mind the fact that charts identical with this one, except that they contained much more detailed information, were issued in all the more important cities and towns of the United States on the morning of January 25, 1905, about a couple hours after the taking of the morning observations, at 8 a. m., Eastern time. The same morning, weather forecasts, cold-wave warnings and storm warnings, deduced from this map, were issued to some hundreds of thousands of addresses by telegraph, telephone, mail, and messenger. The map itself conveys information comparable in interest to the news of public events published on the first page of the newspapers the same day. The forecasts and especially the warnings are, in such a case, worth millions of dollars to the people of the United States. The following account of the cold wave appears in the Annual Report of the Chief of the Weather Bureau for 1905:
“A severe cold wave appeared over the Dakotas, Minnesota, Nebraska, and Iowa on January 24, 1905, and on the 25th covered the central and upper Mississippi valleys and extended over the northern portions of the east Gulf States, the line of zero temperature reaching into northern Tennessee. On the 26th the cold wave covered Florida, and temperatures below freezing were reported as far south as Tampa and Jupiter. At the latter place, the minimum temperature, 24 degrees, equaled the lowest ever recorded since the establishment of the Weather Bureau station at that point. Considerable damage was done to orange trees where groves could not be fired or protected. Ample warnings had been given of the expected low temperatures.”
The subject of making forecasts from a weather map is one concerning which some big books have been written, and it cannot be dealt with very satisfactorily in the brief space at our disposal. There are two cardinal rules—viz., (1) the weather has a characteristic distribution in relation to the distribution of barometric pressure, and (2) pressure systems, with their attendant winds and weather, move, in a general way, from west to east—but these rules require various qualifications and are subject to various exceptions. Thus highs and lows generally take rather circuitous routes in getting, eventually, to the eastward, and sometimes they break up or fade out. The high shown on the annexed map actually moved much more south than east during the following twenty-four hours, while the low moved up the coast; i. e., more north than east. These were, however, the movements expected by the experienced forecaster.
Well-developed highs are nearly always regions of clear weather and, in winter, of cold weather. Lows are attended by clouds and precipitation; rising temperature usually precedes them and falling temperature follows them. The professional forecaster recognizes several types of pressure distribution other than ordinary highs and lows, and they, also, have their characteristic winds and weather. The seven typical forms of isobars, as classified many years ago by R. Abercromby, are: Cyclone; anticyclone; secondary; V-shaped depression, or trough; wedge of high pressure; col, or saddle, between two anticyclones; and straight isobars. These may be combined in a variety of ways on the weather map. The forecaster learns to classify these combinations consciously or subconsciously, and grows familiar with their habits and mannerisms. Forecasters spend a good deal of time in studying the files of weather maps for past years; but the results of such studies are not easy to reduce to definite statements. Scientific forecasting is, in its present stage, almost wholly empirical. The dependence of weather changes upon the phenomena of atmospheric circulation is generally easy to make out, but the vagaries of winds and pressure are still in the main mysterious; notwithstanding such interesting developments as (1) the much-discussed rules of M. Gabriel Guilbert for predicting the movements of barometric depressions from abnormalities in the force and direction of the winds; (2) the systematic charting of “isallobars,” or lines of equal pressure change, associated especially with the name of Dr. Nils Ekholm; and finally (3) the hypothesis of a sharp line of demarcation between masses of equatorial and polar air along the so-called “polar front,” forming the basis of a system of forecasting that originated at the Geophysical Institute of Bergen, Norway, and has had a marked influence upon the methods of forecasters in other parts of the world.
As to the practical results of this empirical art, one point of the utmost importance is commonly overlooked by the public when it complains about the mistakes of the forecaster. He is required to make forecasts of weather day after day, regardless of the kind of map that is laid before him. Sometimes the map is so featureless (or, as the forecasters say, “flat”) that there is little in it on which to build a forecast. At other times there is an abundance of features, but they are in process of rapid and disconcerting change. In either case the ordinary day-to-day weather forecast is likely to go astray. The brighter side of this picture is that the atmospheric phenomena that count heavily in terms of dollars and imperiled human lives are not found on “flat” maps, and, when they appear on the map, generally behave in a simple, straightforward way. In other words, such events as great storms and cold waves are far easier to forecast than everyday weather, and it is the successful prediction of these events that furnishes the principal raison d’être of an expensive telegraphic forecasting service.
Since weather predictions serve a variety of purposes, many different kinds of forecasts and warnings have been developed by meteorological services. In this country there are, first of all, district forecasts and local forecasts; the former, covering whole States and groups of States, being issued at a few main forecasting centers, while the latter, applying to a single town and its vicinity, are issued at a large number of the ordinary Weather Bureau stations. There is a long list of forecasts and warnings intended for special classes of the community; indeed the specialization of forecasts is carried so far that an individual citizen or a single business firm can generally obtain, by asking for it, a forecast of any specified predictable feature of the weather for a particular place. The established types of special prediction issued regularly, or when conditions warrant, include wind and weather forecasts and storm warnings for mariners; shippers’ forecasts, relating to temperatures injurious to perishable goods; aviation forecasts; “fire-weather” warnings, issued when the weather is conducive to fires in the western forests; avalanche warnings; and several different kinds of advices for the benefit of agriculture and horticulture. The United States Weather Bureau, although it is not the only branch of the Government that carries on work in hydrology, is the one charged with the duty of issuing river-stage predictions and flood warnings. An elaborate organization is maintained for this purpose, and the results are extremely successful.
The period of time covered by an official weather forecast is generally from one to two days. In this country the morning forecast is ordinarily for 36 hours from 8 a. m., and the evening forecast for 48 hours from 8 p. m., but occasionally the period is extended for an additional day. For some years the Weather Bureau has issued every Saturday a forecast in quite general terms relating to the whole of the following week. These long-range forecasts are made for extensive areas of the country, such as the North and Middle Atlantic States, the Ohio Valley and Tennessee, and the Great Lakes region. “From the weekly forecast,” says an official publication, “a farmer may know whether it is safe to cut his hay at the beginning of the week or whether it would be better to wait till the last of the week; and a produce dealer may know whether it is safe, at a particular time in the early spring, to start a carload of strawberries to a northern market.” The British Meteorological Office follows a more cautious plan. Its regular forecasts are for twenty-four hours, but occasionally, when conditions are fairly settled, announcements are made of what is termed the “further outlook.” The same office sends notices to the agricultural districts when a spell of fine weather, favorable for haying or the like, appears to be on the programme.
There are a few official meteorological establishments that have embarked on much more ambitious undertakings in long-range forecasting. The classic example is furnished by the Indian Meteorological Department, which has issued seasonal forecasts of rainfall ever since 1882. These were originally based upon reports of the snowfall in the Himalaya, abnormalities of which, as noted in the spring, appear to be related to the intensity of the subsequent monsoon rainfall. Eventually the Indian meteorologists began to seek in more remote regions for clues to the character of the Indian seasons, and they believe they have found them; the barometric pressure in South America and at Mauritius, the rainfall at Zanzibar and Seychelles, the Nile flood, and summer rains in Australia all seems to bear some relation to meteorological conditions in India.
The study of world-wide interrelations of weather, although it has not generally, as in the case just mentioned, furnished the basis of official forecasts, has engaged the attention of a great many able meteorologists. We have spoken on another page of the “centers of action” that seem to be such important indexes to changes in the circulation of the atmosphere, with concomitant fluctuations in weather. Telegraphic reports from some of these centers, including the Iceland and Aleutian lows and the Siberian and Azores highs, have helped to guide the Weather Bureau in making its weekly forecasts. Many attempts have been made to predict the weather months in advance from variations in the temperatures of the water in different parts of the ocean or from the distribution of sea ice in high latitudes. Another line of attack upon the problem of long-range forecasting is through observations of solar activities, as indicated by fluctuations in solar radiation, the prevalence of sun spots, etc. Lastly, an immense amount of energy has been expended in efforts to detect definite cycles or periodicities in the weather itself, without regard to their causes. The thirty-five year period of rainfall and temperature variations, announced in 1890 by Prof. E. Brückner, has found a place in all the current textbooks on meteorology, and several other alleged weather periods have been the subject of serious discussion.
From all of which it appears that the professional meteorologist is not at all inclined to discountenance attempts at long-range weather prediction, provided they are made both honestly and intelligently. Unfortunately the vast majority of people who, in all ages, have indulged in this sort of vaticination—and their name is legion—have been either dishonest or ignorant, or both. The world is still well supplied with them, and they are, undeniably, a thorn in the flesh of the scientific forecaster, who sometimes sees his predictions confounded with theirs by the public, and who commonly incurs the charge of jealousy and narrow-mindedness because he declines to acknowledge brotherhood with the cranks and impostors who hang about the outskirts of his profession.