CHAPTER XIX.

WEATHER FORECASTING.

In a letter dated at Philadelphia, July 16, 1747, Benjamin Franklin wrote to his friend Jared Eliot as follows: "We have frequently along the North American coast storms from the northeast which blow violently sometimes three or four days. Of these I have had a very singular opinion for some years, viz.: that, though the course of the wind is from northeast to southwest, yet the course of the storm is from southwest to northeast; the air is in violent motion in Virginia before it moves in Connecticut, and in Connecticut before it moves at Cape Sable, etc. My reason for this opinion (if the like have not occurred to you) I will give in my next."

In a second letter to the same correspondent, dated Philadelphia, Feb. 13, 1749-50, Franklin states his reasons as follows: "You desire to know my thoughts about the northeast storms beginning to leeward. Some years since, there was an eclipse of the moon at nine o'clock in the evening, which I intended to observe; but before night a storm blew up at northeast, and continued violent all night and all the next day; the sky thick-clouded, dark, and rainy, so that neither moon nor stars could be seen. The storm did great damage all along the coast, for we had accounts of it in the newspapers from Boston, Newport, New York, Maryland, and Virginia; but what surprised us was to find in the Boston newspapers an account of the observation of that eclipse made there; for I thought as the storm came from the northeast it must have begun sooner at Boston than with us, and consequently have prevented such an observation. I wrote to my brother about it, and he informed me that the eclipse was over there an hour before the storm began. Since which time I have made inquiries from time to time of travelers, and observed the accounts in the newspapers from New England, New York, Maryland, Virginia, and South Carolina; and I find it to be a constant fact that northeast storms begin to leeward, and are often more violent there than to windward" (Sparks's _Life of Franklin_, VI, 79, 105, 106).

The fact that our northeast storms come from the southwest, which was first noticed by Benjamin Franklin some years before he put the suggestion just quoted in writing, was one of the great contributions to meteorology made by Americans. Modern weather forecasting essentially depends upon the general eastward movement of cyclones and anticyclones, with their accompanying weather conditions.

The daily weather map shows us the actual condition of the weather all over the United States at 8 A.M., "Eastern Standard Time." The positions of cyclones and of anticyclones; of areas of clear, fair, cloudy or stormy weather, and of regions of high or low temperatures, are plainly seen at a glance. These areas of fair and foul weather, with their accompanying systems of spiralling winds, move across country in a general easterly direction. Knowing something of their direction and rate of movement, we can determine, with greater or less accuracy, their probable positions in 12, 24, 36, or 48 hours. The prediction or foretelling of the weather which may be expected to prevail at any station or in any district is _weather forecasting_.

Weather forecasts are usually made on our daily weather maps for 24 hours in advance. It is by no means an easy thing to make accurate weather forecasts. Careful study and much practice are required of the forecasters of the Weather Bureau before they are permitted to make the official forecasts which are printed on the daily maps and in the newspapers.

A simple extension and application of the principles learned through the preceding exercises make it possible for us to forecast coming weather changes in a general way. These suggestions are, however, not at all to be considered as a complete discussion of this complicated problem.

Weather forecasts include the probable changes in _temperature_, _wind direction_ and _velocity_, and _weather_. Pressure is not included. Begin your practice in weather forecasting by considering only the changes that may be expected at your own point of observation, and at first confine yourself to predicting temperature changes alone.

=Temperature.=--Provide yourself with a blank weather map. Draw an isotherm east and west across the map, through your station. Draw a few other isotherms all the way across the map, parallel with the first one, and so arranged that they will be equal distances apart, the most northerly one running through northern Maine and the Northwestern States, and the most southerly one through southern Florida and Texas. Recalling what you have already discovered concerning the eastward movement of our weather conditions, what forecast will you make as to the coming temperatures at your station? Add some additional east and west isotherms, so that there will be twice as many on your map as before. What effect will this change in the temperature distribution on your map have upon the temperature forecast you make for your station? Formulate a general rule as to temperature forecasts under the conditions of isothermal arrangement here suggested.

On a second blank weather map draw an isotherm through your station inclined from northwest to southeast. Draw a few other isotherms parallel to the first, and each one representing a temperature 10 higher than that indicated by the adjacent isotherm on the east. Make a general forecast of the temperature conditions that may be expected at your station, as to _kind of change_, if any; _amount of change_, and _rapidity of change_. Of the isotherms just drawn, erase every second one; still, however, letting those that are left represent differences of temperature of 10. What forecast will you now make as to temperature? How does this forecast compare with that just made?

Now draw twice as many isotherms on your map as you had in the first place, still letting these lines represent differences of temperature of 10 in each case. Make a forecast of the kind, amount, and rapidity of temperature change at your station under the conditions represented on this map. How does this forecast compare with the two just made? Formulate a general rule governing temperature forecasts in cases of isothermal arrangement such as those here considered.

Take another blank map. Draw through your station an isotherm inclined from northeast to southwest. Draw other isotherms parallel to this, west of your station, letting each successive isotherm represent a temperature 10 lower than that indicated by the adjacent isotherm on the east. Make a temperature forecast for your station under these conditions. Diminish and increase the number of isotherms on your map, as suggested in the preceding example, making temperature forecasts in each case, and comparing the three sets of forecasts. Formulate a general rule for temperature forecasts made under these systems of isotherms.

Make temperature forecasts from the daily weather maps for your own station, using the knowledge that you have already gained as to the progression of cyclones and anticyclones (Chapter XVII), and as to the temperature distribution in these areas (Chapter XIV), to help you in this work. Study each day's map carefully before you decide on what you will say. Then write out your own forecast, and afterwards compare your forecast with that made by the Weather Bureau. Note also, by reference to your own instrumental observations, whether the succeeding temperature conditions are such as you predicted.

=Wind Direction.=--The weather maps already studied taught us that our winds habitually move in spirals. The composite picture of the wind circulation around cyclones and anticyclones (Chapter XII) further emphasized this important fact. Evidently this law of the systematic circulation of the winds around centers of low and high pressure may be utilized in making forecasts of wind direction.

Applying the knowledge already gained concerning cyclonic and anticyclonic wind circulations, ask yourself what winds a station should have which is within the range of the cyclonic wind system, and is in the following positions with reference to the center: _northeast_, _north_, _northwest_, _east_, _at the center_, _west_, _southeast_, _south_, _southwest_. Ask yourself precisely the same questions with reference to a station within an anticyclonic wind system. Write out a general rule for the kinds of wind changes which may be expected to take place under these different conditions.

When a station is south of the track of a pa.s.sing cyclone its winds are said to _veer_, and the change in the direction of its winds is called _veering_. A station north of the track of a pa.s.sing cyclone has a change of direction in its winds which is known as _backing_, the winds themselves being said to _back_.

=Wind Velocity.=--What general relation between wind velocities and areas of low and high pressure did you discover in your study of the weather maps? What was the result of your work on the correlation of the velocity of the wind and the barometric gradient in Chapter X? And what general statement as to the relation between the velocity of the wind and its distance from a cyclonic or anticyclonic center may be made as the result of your work in Chapter XII, on the correlation of cyclones and anticyclones with their wind circulation? These results must be borne in mind in making predictions of coming changes in wind velocities. Forecasts of wind velocities are made in general terms only,--_light_, _moderate_, _fresh_, _brisk_, _high_, _gale_, _hurricane_,--and are not given in miles per hour.

Make forecasts of wind direction and velocity from the daily weather maps for your own station. Continue these for a week or two, keeping record of the verification or non-verification of each of your forecasts. Then make daily forecasts of temperature and of wind direction and velocity together. Write out your own forecast for each day before you compare it with the official forecast, and if the two differ, keep note of which one seemed to you to be the most accurate.

=Weather.=--What general relation between kind of weather and cyclones and anticyclones was ill.u.s.trated on the six maps of our series? What is the average distribution of the different kinds of weather around cyclones and anticyclones, as shown by your composites? (Chapter XVI.) What changes in weather will ordinarily be experienced at a station as a cyclone approaches, pa.s.ses over, and moves off? What conditions will prevail in an anticyclone?

Make a series of daily forecasts for your own station of probable weather changes, omitting temperature and winds at first. Include in your weather forecasts the state of the sky (_clear_, _fair_, _cloudy_); the changes in the state of the sky (increasing or decreasing cloudiness); the kind of precipitation (_rain_ or _snow_) and the amount of precipitation (_light_ or _heavy_). Write out your forecasts; compare them with the official forecasts, and notice how fully they are verified. Then add temperature and winds to your forecasts so that you will make a complete prediction of probable changes in temperature (kind, amount, and rapidity of change), wind (direction and velocity), and weather. Practice making these complete forecasts for several weeks, if time allows. Use all the knowledge that you have gained in the preceding work to aid you in this. Study each weather map very carefully. Do not write down your forecast until you are sure that you have done the best you can.

[Ill.u.s.tration: FIG. 53.]

Vary this exercise by extending your forecasts so as to embrace the whole section of country in which your station is situated (as, _e.g._, New England, the Gulf States, the Lake region). Pay special attention to making forecasts of cold waves, of heavy rain or snowstorms, of high winds over the lakes or along the Atlantic coast, etc. When possible, obtain from the daily newspapers any particulars as to damage done by frost or gales, or concerning snow blockades, floods due to heavy rains, etc.

Fig. 53 summarizes what has thus far been learned as to the distribution of the various weather elements around a well-developed center of low pressure. The curved broken lines represent the _isotherms_ (Chapter XIV).

The solid concentric oval lines are the _isobars_ (Chapter XI). The arrows represent the _winds_, the lengths of the arrows being roughly proportionate to the wind velocities (Chapter XII). The whole shaded area represents the region over which the sky is covered by heavy lower clouds.

The smaller shaded area, within the larger, encloses the district over which rain or snow is falling (Chapter XVI). The lines running out in front of the cloudy area represent the light upper clouds (_cirrus_ and _cirro-stratus_) which usually precede an area of low pressure.

Imagine this whole disturbance moving across the United States in a northeasterly direction, and imagine yourself at a station (1) directly in the path of the cyclone; (2) south of the track; and (3) north of the track. In the first case, as the disturbance moved on in its path, you would successively occupy the positions marked _A_, _B_, and _C_ on the line _AC_, pa.s.sing through the center of the cyclone. In the second case you would be first at _D_, then at _E_, and then at _F_. In the third case you would be at _G_, _H_, and _J_ in succession. What changes of weather would you experience in each of these positions as the cyclone pa.s.sed by you? Imagine yourself at some station halfway between the lines _AC_ and _DF_. What weather changes would you have in that position with reference to the storm track? In what respects would these weather changes differ from those experienced along the line _DF_? Imagine your station halfway between the lines _AC_ and _GJ_. What weather changes would you have there? How would these changes differ from those experienced along the line _GJ_?

It must be remembered that Fig. 53 is an ideal diagram. It represents conditions which are not to be expected in every cyclone which appears on our weather maps. If all cyclones were exactly alike in the weather conditions around them, weather forecasting would be a very easy task. But cyclones are not all alike--far from it. Some are well developed, with strong gradients, high winds, extended cloud areas, heavy precipitation, and decided temperature contrasts. Others are but poorly developed, with weak gradients, light winds, small temperature differences, and it may be without any precipitation whatever. Some cover immense districts of country; others are small and affect only a limited area. It therefore becomes necessary to examine the characteristics of each approaching cyclone, as shown on the daily weather map, very carefully. Notice whether it is accompanied by heavy rain or snow; whether its winds are violent; how far ahead of the center the cloudy area extends; how far behind the outer cloud limit the rain area begins; what is the position of the cloud and rain area with reference to the center, and other points of equal importance, and govern yourself, in making your forecast, according to the special features of each individual cyclone. Well-developed cyclones will be accompanied by marked weather changes. Weak cyclones will have their weather changes but faintly marked.

The distance of your station from the center of the cyclone is of great importance in determining what the weather conditions and changes shall be, as may easily be seen by examining Fig. 53. If the storm pa.s.ses far to the north or far to the south of your station, you may notice none of its accompanying weather conditions, except, perhaps, a bank of clouds on your horizon. You may for a few hours be under the cloudy sky of some pa.s.sing storm, and yet not be reached by its rainy area. The shifts in the wind may be marked and the wind velocities high, or the expected veering or backing may hardly be noticeable, owing to the weakness or the distance of the controlling cyclone.

Again, the rapidity with which weather conditions will change depends upon the rate of movement of the cyclone itself. The better developed the cyclone, the higher its velocity of progression, and the nearer its track lies to the station, the more emphatic and the more rapid are the weather changes it causes. On the other hand, the weaker the cyclone, the slower its rate of progression, and the further away its track, the less marked and the slower the weather changes. The probable track of a coming storm, and its probable rate of movement, therefore, need careful study if our forecasts are to be reliable.

There are many other obstacles in the way which combine to render weather forecasting extremely difficult. Some of these difficulties you will learn to overcome more or less successfully by the experience you will gain from a careful and persevering study of the daily weather maps; others, the best forecast officials of our Weather Bureau have not yet entirely overcome. The tracks followed by our cyclones vary more or less from month to month, and even if the average tracks for each month are known, individual cyclones may occur which absolutely disregard these tracks.

While the average hourly velocity of cyclones is accurately known for the year and for each month, the movements of individual storms are often very capricious. They may move with a fairly uniform velocity throughout the time of their duration; they may suddenly and unexpectedly increase their rate of movement, or they may as suddenly come nearly to a standstill. The characteristics of cyclones vary in different portions of the country and at different times. Cyclones which have been accompanied by little precipitation on most of their journey are apt to give increased rain or snowfall as they near the Atlantic Ocean and Gulf of St. Lawrence.

Cyclones which over one portion of the country were rainy, may give little or no precipitation in another portion. Cyclones and anticyclones are found to have considerable influence on one another, r.e.t.a.r.ding or accelerating one another's advance, or changing one another's normal path of progression. While this mutual interaction is clearly seen, and may be successfully predicted in many cases, many other cases arise in which, under apparently similar conditions, the result is very different from the antic.i.p.ation. Such are some of the difficulties with which weather forecasters have to contend, and which prevent the attainment of greater accuracy in weather prediction.

PART V.--PROBLEMS IN OBSERVATIONAL METEOROLOGY.

CHAPTER XX.

TEMPERATURE.

The chief interest and value of the instrumental work in meteorology are to be found not only in the taking of the daily observations at stated hours, but in the working out of numerous simple problems, such as may readily be undertaken with the help of the instruments already described.

Thus, the temperature of the air (obtained by the sling thermometer, supplemented by maximum and minimum thermometers, and by the thermograph if available) can be determined under a variety of conditions, _e.g._, close to the ground, and at different heights above the ground; at different hours, by day and night; in different seasons; in sunshine and in shade; during wind and calms; in clear and cloudy weather; in woods and in the open; over bare ground, gra.s.s, snow, or ice; on hills and in valleys. Observations may also be made of the temperature of the ground and of a snow cover, at the surface and at slight depths beneath the surface, in different seasons and under different weather conditions.

Among the problems which may be worked out by means of such observations as these are the following:--

_A._ =The Diurnal Range of Temperature under Different Conditions and at Different Heights above the Ground.=--Under the influence of the sun the regular normal variation of temperature during 24 hours is as follows: A gradual increase, with the increasing alt.i.tude of the sun, from sunrise until shortly after noon, and a gradual decrease, with decreasing alt.i.tude of the sun, from the maximum, shortly after noon, until the minimum, about sunrise. This variation is known as the _diurnal variation of temperature_. Curve _a_ in Fig. 12 ill.u.s.trates well the normal diurnal variation of temperature, as recorded by the thermograph during a period of clear, warm spring weather (April 27-30, 1889, Nashua, N. H.). The _diurnal range of temperature_ is the difference between the maximum and minimum of the diurnal oscillation. The regular normal diurnal variation in temperature is often much interfered with by other controlling causes than the sun, _e.g._, cyclonic winds, clouds, etc.

I. Study and compare the diurnal ranges of temperature as indicated by the maximum and minimum thermometers, or the thermograph, in the instrument shelter, in clear, fair, cloudy, and stormy weather, during winds and in calms, in different months. Summarize your results by grouping them according to the general weather conditions, and according to the months or seasons in which the observations were made. For example, group together and average the ranges observed on clear, calm days in winter; on similar days in early summer or autumn; on clear days with brisk northwest winds in winter; on similar days in early summer or autumn; on calm days with overcast sky in the different seasons; on stormy days with strong winds, etc. Study carefully the weather maps for the days on which your observations are made. Pay special attention to the relation between the diurnal ranges and the control exercised over these ranges by cyclones and anticyclones through their winds and general weather conditions.

II. Observations of diurnal ranges of temperature at different heights above the ground may be made by means of maximum and minimum thermometers fastened (temporarily) outside of the windows of different stories of the school or of some other building. These observations should be made out of windows facing north, and care should be taken to check, so far as possible, any draft from within the building out through the window during the taking of the observation. If a fire escape is provided on the building, the instruments may often be conveniently fastened to that.

Study the ranges under different conditions of wind and weather at various heights above the ground, and compare these results with those obtained under I. Notice the relations of all your results to the cyclonic and anticyclonic areas of the weather maps.

The diurnal range of temperature in the air over the open ocean from the equator to lat.i.tude 40 has been found to average only 2 to 3. In southeastern California and the adjacent portion of Arizona the average diurnal temperature range in summer is 40 or 45. Over other arid regions, such as the Sahara, Arabia, and the interior of Australia, the range also often amounts to 40. Observations of temperature above the earth's surface, in the free air, made on mountains, in balloons, and by means of instruments elevated by kites, indicate very clearly that the diurnal range of temperature decreases with increasing elevation above sea level. The results obtained at Blue Hill Observatory, Ma.s.sachusetts, by means of kites, show that the diurnal range of temperature almost disappears, on the average, at 3300 feet (1000 meters).

_B._ =Changes of Temperature in the Lower Air, and their Control by the Condition of the Ground, the Movement of the Air, and Other Factors.=--Determine the changes of the temperature in the lower air by making frequent readings of the ordinary thermometer in the instrument shelter, of the sling thermometer, or by an examination of the thermograph record. Group these changes, as in Problem _A_, so far as possible according to the weather conditions under which they occurred, and try to cla.s.sify the kinds of change roughly into types. Study the control of these various types by the wind and other weather conditions accompanying them, as ill.u.s.trated on the daily weather maps. The control exercised by different conditions of the earth's surface may be studied by means of observations made with the sling thermometer over different surfaces, such as gra.s.s, bare ground, snow, etc.

Examples of temperature changes in the lower air, under different conditions of weather, recorded on the thermograph, are given in Fig. 12, and are briefly referred to their causes in the text accompanying that figure.

_C._ =Vertical Distribution of Temperature in the Atmosphere.=--The vertical distribution of temperature in the lower air may be studied by having ordinary thermometers or thermographs exposed at different heights above the ground, _e.g._, close to the surface; in an instrument shelter; out of windows on successive stories of some high building; and on the roof of the building. They may also, in cases where there is a hill in the neighborhood, be exposed in a valley at the bottom of the hill and at successive elevations up the side of the hill. It is, however, usually much simpler, as well as more practicable, to take these temperature readings by means of the sling thermometer. In the case of observations made out of the windows of a building, one observer can take the readings at different elevations in succession. When the observations are made at different alt.i.tudes on the side of a hill, it is best to have the cooperation of several observers, who shall all read their thermometers at the same moment of time. The results obtained in the previous problems (_A_ and _B_) may, of course, also be utilized in studying the vertical distribution of temperature in the atmosphere.