Meteorology: The Science of the Atmosphere

by Charles Fitzhugh Talman

The climatic data for any place are a sort of digest of the meteorological observations that have been made there, special emphasis being given to those features of the meteorological record that bear important relations to the life and activities of mankind. Temperature and rainfall are the leading elements of climate; others are wind, humidity, evaporation, cloudiness, etc. We have not space to enumerate here all the kinds of data found in elaborate climatic tables; but in order to illustrate how the records of a meteorological station are utilized in compiling climatic statistics and to show what complications may arise in this process, we shall consider the question of temperature alone.

The instruments used in measuring temperature have been described in another chapter. From these instruments are obtained the current temperature of the air, the wet bulb temperature (used to compute the humidity), and the maximum and minimum temperatures of the day. Readings are made at fixed hours, known as “term hours.” At regular stations of the United States Weather Bureau the term hours for the observation of all the meteorological elements are 8 a. m. and 8 p. m., Eastern Time, and an observation of temperature, humidity, and clouds is made at noon. In most other countries tri-daily readings have been the rule, though in Europe four or more observations a day are now taken at many stations in order to supply the frequent weather bulletins required by aeronauts. Important stations are generally equipped with thermographs, which make a continuous record of temperature.

Theoretically, the mean temperature of any day is the average of 24 hourly observations, from midnight to midnight. In practice, the mean is generally computed from the observations at the term hours, or from the maximum and minimum. Having obtained the mean daily temperature for each day of a month, the average of these values gives us the mean monthly temperature. The average of the mean temperatures for the twelve months of the year is the mean annual temperature.

These data for each day and month, and for the year—sometimes also for other intervals, such as five-day periods, or “pentads,”—are computed year after year, and eventually the values for all the years of the record are averaged to form what are called “normals.” We thus obtain, for example, for a given station, the normal temperature for January 21, the normal temperature for the month of March, the normal annual temperature, etc.

All this is a mere beginning toward the complete discussion of a body of temperature observations for the purposes of climatology. We have still to obtain from the readings of the maximum and minimum thermometers the normal maximum and minimum temperatures and range of temperature for each day, each month, and the year; also the “absolute” maxima, minima, and ranges (i. e., the extreme values that have occurred during the entire record) for corresponding intervals of time. These data furnish answers to such questions as: What was the lowest temperature ever recorded on January 21? What is the lowest on an average January 21? What is the average range of temperature in March? What was the highest temperature ever recorded, on any day, at the station?

Having thus disposed of the extremes and ranges, we may compute what is called the “variability” of temperature, i. e., the average difference between the means of two successive days in a given month, and the corresponding average for the entire year. These data are of considerable importance in medical climatology. We may also compute the frequency of occurrence of various values among the temperature data above enumerated. The most frequent value is often quite different from the average value. Many climatologists compute the number of days, in an average year, on which the temperature rises to 77 degrees (Fahr.) or above (“summer-days”), and the number of days on which the temperature does not rise above the freezing point (“winter-days”). Especially valuable in agricultural regions are data of the average and extreme dates of the last frost in spring and of the first frost in autumn. These define the length of the “growing season.” Statistics of the temperature of the ground at the surface and at various depths below the surface are also of agricultural interest.

From the foregoing outline it will be seen that a bewildering variety of climatic statistics may be computed merely from observations of temperature, and the same is true of the other elements. Moreover, the list set forth above is by no means exhaustive even for temperature. In fact, there is almost no limit to the number of ways in which the raw material of climatic data—i. e., the original records of observation—may be grouped, averaged, or otherwise treated in order to bring out certain features of the climate that may conceivably serve some useful purpose. The reader will now be able to understand why a treatise on the climate of a single locality often fills a substantial volume.

The numerical data contained in such a work are generally supplemented by text descriptions and by various graphic devices, such as curves showing the normal fluctuation or “march” of a weather element during a day, year, or other interval. Works which deal with the climates of larger areas, such as whole countries, are usually accompanied by climatic charts. These charts furnish a quick and easy way of getting a general idea of the climate of a region. Among the more important climatic charts are the following:

1. Temperature (isothermal) charts. These include charts showing the distribution of normal temperatures for months, seasons, and the year; normal range of temperature for similar periods; highest and lowest temperatures ever recorded at the different stations; etc. Lines known as isotherms are so drawn as to pass through places having identical values of the element in question (mean temperature, highest temperature, etc.).

2. Rainfall (isohyetal) charts. These show the distribution of rainfall (including snowfall, expressed in equivalent depth of water); especially for each month and for the year. Other charts may show the average number of rainy days; average snowfall (actual depth, not water equivalent); seasonal distribution of rainfall; etc.