WEATHER OR NOT
Chief Physicist, Meteorological Service of Canada
DURING the next twenty years, the public may confidently look forward to increasing accuracy and precision in the regular weather forecasts for thirtysix hours ahead. For example, the Western farmer will not need to be satisfied with the general forecast, “scattered showers over the three prairie provinces.” Such a forecast will be replaced by definite official pronouncements like, “light showers in southwest Saskatchewan beginning about 4 p.m., moderate southwest winds turning about midnight to strong northwest winds in Regina district,” and similar information for each locality in the Dominion.
Progress in forecasting weather all over the world has been almost exclusively toward improvement of the forecast for a day or at most two days ahead. The success achieved in Canada has been due to the increased information available from the Far North, made possible by radio stations at Aklavik, Chesterfield Inlet and on the Hudson Straits.
Now a new attack is being made by means of airplanes in the age-long struggle with the weather. The United States Weather Bureau began on July 1, 1934, to make ascents with twenty airplanes at as many points over the United States. These flights are made every day at 5 a.m. to obtain the temperature and humidity of the air up to 15,000 feet. Just as the geologist employs the diamond drill to bring up samples of ore at various depths far down in the earth, so the aviator obtains a record of the air conditions at those high levels where clouds float. When the number of aerial scouts is extended to cover adequately the whole continent, the forecasts for tomorrow’s weather will be more reliable and precise than at present.
Official forecasters in this country have made no serious efforts to forecast for a year or even a week ahead. The outlook has appeared so hopeless that little money has been risked on investigations of weather trends for the future.
Present Limit of Forecasts
THE OFFICIAL weather forecasts made at the head office of the Meteorological Service of Canada cover as a general rule only the thirty-six hours following the time they are made, and only rarely the forecaster ventures to give his opinion of the probable weather for a further period of twelve or twenty-four hours. About eighty-five forecasts out of 100 are verified by the subsequent weather reports. The method used in making these forecasts in Canada is
to have the weather reports telegraphed to Toronto from about 225 stations scattered all over the continent from Mexico to Greenland and Alaska, and from ships at sea. Knowing how weather generally moves, a forecaster, by using a few physical principles, by many artifices and by long experience, is able to judge with a high degree of accuracy the future weather for thirty-six hours. The forecaster, looking at a new map, frequently recognizes features which closely resemble those present on some day in a previous month or year and, remembering what occurred on that former occasion, forecasts a somewhat similar development.
Unfortunately, two weather maps of North America looking almost exactly alike will gradually develop differences which at first appear insignificant but which increase rapidly in importance until by forty-eight hours the weather maps are widely diverse. It is the fear of the development of some unforeseeable change that makes the official forecaster wary of a forecast longer than forty-eight hours.
Animals as Forecasters
THERE IS no lack of weather prophets and forecasters, male and female, insect and animal, animate and inanimate, on which human frailty places confidence in their power to foretell the weather for months and seasons ahead. To list all agencies credited with the power to divine the weather would require a whole page, but a few most current in Canada are:
Squirrels laying up an unusually large quantity of nuts— severe winter; if these nuts have thick shells—very severe winter; long fur on dog—severe winter; small muskrat domes—mild winter; birds flying south earlier than usual or higher than usualsevere winter.
Now it should be clear to every observer of nature that these signs are the result of past weather and have no indication of the future. The glossiness and length of the hair of domestic animals around the farm depend largely on good feeding of the animals, and it would appear reasonable that wild animals’ pelts also depend on the abundance of the wild animals’ natural food supplies.
The muskrat builds his winter’s nest out of grasses, sedges and aquatic plants. If there has been a luxurious growth of these plants during the summer, the muskrats’ domes will be larger than after a summer when drouth or other cause has made vegetation scarce. The instinct to gather food would be a great deal more natural than one to foresee the rigors of an approaching winter. Squirrels gather a great store of nuts when there has been a plentiful yield, and their winter supply doubtless varies from one year to another, depending on the yield of nuts.
The patent medicine almanacs distributed broadcast
throughout Canada usually contain forecasts for a whole year ahead, and enjoy a widespread reputation in spite of their vagueness and unreliability. Thus in one almanac, advertising the superlative merits of a famous long-bearded doctor’s medicines, which the publishers state is sent to 1,800,000 Canadian homes, the weather for the first ten days in February, 1934, reads, “comes in cold, fair and cold, unless wind turns south or southwest.”
What would happen in this event is left to the reader’s imagination. However, in Toronto, the wind blew' from the north and the thermometer fell on February 8 to twenty-one degrees below zero, the lowest temperature recorded at Toronto for many years.
A story is told of how one of these patent-medicine almanacs made an enormous reputation. The office clerk was busy preparing the predictions for the coming year, putting down with much effort weather conditions—“mild spell followed by severe cold, continues stormy,” etc.—but inspiration failed him and he repeatedly asked his chief for suggestions. The latter became more and more exasperated at the interruptions, so that when the clerk enquired regarding the weather for the middle of June, the chief shouted, “Put down snow, heavy snow.” Heavy snow went down, and. by one of those vagaries of fate, heavy snow came and the almanac’s reputation wras made.
Most of the predictions are expressed in general terms, and the region for which they are valid is unstated. Thus for the first two weeks of July an almanac could safely write “F'air and warm all this week. Fair again, but weather is changeable.”
In ordinary circumstances, for these almanac forecasts to be really useful and not one hundred per cent humbug, the publisher should state the region for which the forecast is to be considered.
A Sixty-Year Record
NOW THE Meteorological Sendee of Canada publishes a monthly record of weather observations taken at 900 stations scattered over the Dominion from coast to coast, from Pelee Island in the latitude of Rome to Craig Harbor, about 800 miles from the North Pole. These records are available for Toronto for over a century, for Eastern Canada from the fifties, and for Western Canada from the seventies and eighties. The enormous amount of statistics about past Canadian weather covering tens of thousands of quarto pages overpowers a single investigator, and in order to arrive at any conclusions it is necessary to select a very small fraction of the information available.
We begin by taking three stations with long records from the prairies—Prince Albert with records kept from 1885,
Continued on page 34
Continued from page 11
Qu’Appelle, near Regina, with records from 1883, and Edmonton from 1883. Toronto and Montreal weather records from 1875 represent fairly well weather in Central Canada, and the records from Saint John, N.B., and Halifax indicate the temperature variations in the Maritimes. The records since 1890 from Victoria, B.C., have been studied to find the trend of weather on the Pacific Coast.
Although winter sometimes sets in earlier than Decembei T and continues after February 28, for the sake of ease in comparison, the average temperature for the three months, December, January and February, is considered the winter temperature.
January is the coldest month of the year in Central Canada, followed in order by February and December. In a few Northwestern stations like Dawson, December is colder than January, while in the Maritimes February is frequently the coldest month in the year.
The weather records for the past fifty years show that the mildest winter in British Columbia occurred in 1925-26, on the prairies in 1930-31, in Central Canada (Toronto and Montreal! in 1931-32, and in the Maritimes in 1932-33. The most severe winter in each of these regions since 1880 has occurred as follows: On the prairies in 1885-86, in Central Canada in 1917-18, and at Saint John and Halifax in 1933-34.
Generally a very mild winter on the prairies brings somewhat above the normal temperature eastward to the Atlantic, but British Columbia often differs widely from the rest of Canada. Thus the Maritime provinces experienced the most severe winter in 1933-34 in half a century, while British Columbia enjoyed the second warmest winter on record, exceeded only by 1925-26.
Attempts to Forecast Winters
SCIENTISTS have attempted by two quite different methods to forecast the weather in coming seasons or years— better known as cycles — not the businessman’s irregularly occurring periods of prosperity and depression, but regularly repeated years of high rainfall or temperature separated by a very definite number of years.
The most famous cycle in nature is the regular change in the number of spots on the sun’s face, which even,' 11.1 years reaches a high value, then slowly decreases for the following years. The last high value occurred in December, 1929, when 108 spots daily flecked the sun, and this decreased to December, 1933, when only one spot every three days v'as observed. The most recent figures show that the sun is once more growing spotty.
Sun spots certainly affect the behavior of the magnetic needle all over the earth, producing magnetic storms during which for several days the needle will wander back and forth over its normal position. At these times “static" is more often bad than when the needle is quiet, and aurora or Northern lights flare up in the sky. Since this elevenyear cycle of sun spots has such pronounced effects on the magnetism of the earth, one might reasonably expect it would affect the weather.
If weather records from the tropics are examined over a long period of years, it is found that temfwratures are slightly higher when sun sjxrts are few than when they are numerous. Koppen, a great German investigator, believed he had proved this for stations all over the earth, but the most recent investigations by Dr. Clayton, an American, show that it is true only for certain low-lying regions near the sea coast.
Now the heat from the sun reaching the earth is not constant from year to year or even day to day, but has varied between one and two per cent since it was first accurately measured by Abbott and Langley for the Smithsonian Institution of Washington, D.C. The greatest value of solar heat for a whole year was measured in 1921, the smallest in 1922, so that it obviously does not fluctuate with the sunspots. If the sun’s heat should increase slightly, the first effect over Canada would be to lower the temperature. This surprising result arises from the increased cloudiness caused by evaporation from the vast oceanic region inside the tropics.
The one known cause which produces widespread cold is a great volcanic eruption throwing up an enormous amount of ash high into the air, which effectively screens the earth’s surface from the sun’s rays. In America the famous “year without a summer” followed the titanic eruption of Tomboro, Dutch East Indies, in 1815, when this volcano threw so much dust into the air that "for three days there was darkness at a distance of 300 miles.” The eruption of Krakatoa on August 27, 1883—the greatest volcanic activity for the past century — lowered the earth’s temperature about one degree Fahrenheit for a year. In Canada the following winter was cold but not nearly as severe as 1885-86, which was one of the coldest on record. The more genial world temperatures of the past ten years have been ascribed to the entire absence of great volcanic eruptions.
The Weather Balance
'“PHE OTHER and more successful method of forecasting the weather for months ahead is by connecting it with the previous weather at some other place on the earth. The rainfall in the United States has varied directly with the rainfall in South America for the past thirty years, and opposite to the rainfall in Australia. The three driest years in the past quarter of a century in the United States—1910, 1917 and 1930—were the three wettest years in Australia in the same period. It is evident there is something in common in the weather in widely separated parts of the earth, and its changes are in some way related to one another over the entire world.
One of the most extraordinary relations of this kind is the fairly close connection between the average temperature and pressure in Central India for the earlier months of the year and the temperature at Winnipeg during the following winter. When the temperature and more especially the atmospheric pressure in Central India is high the following winter in Winnipeg is mild, and vice versa. Out of forty-six years from 1875 to 1920, this relation was found to be correct for thirty-seven years, wrong in seven and indecisive twice when weather conditions were exactly normal in India. There appeared so little reason or logic for connecting summer weather on India’s coral strand with the following winter in Winnipeg that computations have not been made for recent years. It is an interesting by-product of an investigation by a German engineer named Grossmaier.
Why Wintens Vary
YUr E ALL KNOW that the polar regions W within the Arctic Circle remain during winter at a temperature from twenty-five degrees to forty-five degrees below zero, while the tropics bask in temperatures from seventy degrees to eighty degrees. Cold air creeps out in great uniform masses from the Arctic Basin, advancing 300 to 400 miles per day along a wide front stretching eastward from the south of the Mackenzie. These air masses frequently halt temporarily crossing the Northwest Territories, becoming intensely cold, and then, once more moving forward rapidly as a “cold wave,” travel onward to the Southern United States with very little warming up. The regions in the Arctic that would have been left empty by the cold air are occupied by warm air moving
up from the tropics heavily laden with moisture.
The paths for these great currents of warm and cold air are not confined to definite limits like a river within its banks, but shift their position from time to time. Thus, last winter cold air streamed southward during February on the Eastern side of this continent, while over the Western provinces and states there was a persistent flow from the South.
The precise cause which starts the cold air southward along a particular route may possibly be very small, just as a few grains of dust on a smooth floor may deflect an immeasurably larger mass of spilled water one way or another. Once the spilled water has established a track over the floor it has a tendency to flow in it, and a similar tendency is observable in air moving over the earth. Mountain ranges do exert an effect in guiding these enormous currents, but mountain barriers rarely exceed a height of two miles, while these great currents of air which decide whether our winter will be cold or warm extend upward from six to ten miles.
Today, meteorologists of America are busystudying the temperature and moisture of these vast air streams, but the precise factors that control their path, making the outbreak of Arctic air flow for one winter on the Eastern side and for another winter on the Western side of Canada, remain undiscovered.
In other parts of the world, where conditions are simpler than in Canada, considerable success has already been achieved in weather forecasting for two or three months ahead. The Indian Meteorological Service has successfully predicted for years the time of the beginning of the rainy season, on the basis of the great ebb and flow of air from Siberia to the Indian Ocean and back again. Nearer home, Professor McEwen of the Scripps Institution, has forecast with considerable accuracy the rainfall on the Pacific Coast states from the movement of the air over the Pacific.
When the temperaiure of the ocean is warmer or colder than usual, then the air is more humid or drier than the normal. Since the ocean temperature is affected by the sun’s heat during previous months, this factor carries considerable weight in making the forecast. For the coming year 1934-35, rainfall in the Pacific Coast states has been forecast to be about twenty-five per cent below the average.
These few successes lead meteorologists to expect that, just as the height and time of the oceans’ tides may be computed for years in advance for almost any seaport, so too it will eventually be possible to predict the movements of air which constitute weather.