APPLIED CLIMATOLOGY 
The use of climatological information for forecasting 
is really nothing new. The recognition of certain per- 
sistencies in series of observations led to some early 
forecasting rules. Other autocorrelations in time-series 
of meteorological observations at one point have been 
discussed repeatedly as aids to forecasting [36, 54]. 
The modern approach substitutes an areal net of 
observations for local time series. From series of 
synoptic charts, stratified or multiple correlations of 
meteorological parameters at distant stations are used 
as predictors for the locality for which a forecast is 
desired. Frequency distributions of antecedent and sub- 
sequent events are obtained and stochastical relations 
are worked out so that the probability of a given event 
(in dependence upon previous events) can be obtained. 
Machine tabulations have facilitated this approach. 
Subjective forecasting experience, which in many cases 
is nothing but a subconscious system of statistics, is 
replaced by an objective summarization of the past 
performance of weather. While this is a potent tool, it 
should not be overlooked that it is essentially a static 
extrapolation technique and disregards dynamic and 
energetic factors which can influence the weather in the 
interval for which the time-lag relations are worked out. 
The procedures have been adequately presented in 
several recent papers to which reference only is made 
here [91]. Much useful work in this direction remains to 
be done. 
Multiple Point or Areal Problem—Complex Time Re- 
lation—Single Climatic Element 
Problems of this class are most frequently encoun- 
tered among the large number of questions concerned 
with the probable maximum runoff of flood waters as 
determined by the areal patterns of intensity-durations 
of rainfall. Such information is essential in the evalu- 
ation of possible peak floods or in the design of storage 
or flood-control dams, levees, culverts, storm sewers, 
and the lke. The hydrometeorological literature 
abounds with examples of the employment of the tech- 
nique which, in its best form, makes use of the detailed 
precipitation intensity-duration data accumulated from 
the records of a large number of precipitation gages 
(preferably recording gages) distributed over a given 
draimage area. Where such detailed data are not avail- 
able, the use of theory is required to supplement the 
inadequacy of the observational record. The technique, 
itself, is so commonplace that there appears to be little 
requirement that it be further elaborated upon in this 
section. For further details the reader is referred to the 
extensive and readily available hydrologic literature on 
the subject [34, 76]. 
Another example of this class is afforded by the 
necessity for evaluating the frost hazard or assessing 
frost damage to crops as the problem applies to large 
agricultural areas of diverse topography. In an earlier 
section the evaluation of the frost hazard as it applies 
to a single point in an orchard or to a small agricultural 
plot was discussed. In the present case the identical 
temperature-duration data are required, but now we 
987 
are concerned with the pattern of such conditions as 
they occur over a large area during a given time inter- 
val. For the rapid evaluation of the probable freezing 
damage from a single frost, the temperature-duration 
data from a large number of representative points 
within the agricultural area for the single frost night 
may be considered sufficient. For crop estimation pur- 
poses, however, the cumulative data for the several 
series of frost occurrences during the entire growing 
period are required. 
The variations in minimum temperatures and du- 
rations that occur over a region are due largely to dif- 
ferences in (1) exposure to winds, (2) relative elevation 
(in the local sense), and (8) nature of the soil cover 
(whether wet or dry, cultivated or fallow). Of the three 
factors, only (2) remains constant throughout the grow- 
ing season. The most exposed areas will present fewer 
frost nights and the duration of critical temperatures 
will be shorter than in more sheltered locations, due to 
the “mixing” effects of the frequent nocturnal winds. 
The temperature effect of the wind, however, will 
depend upon the degree of stability that has been 
attained in the lower layers of the atmosphere at the 
time the wind is initiated; the degree of exposure to 
winds will vary from night to night depending upon the 
direction of the pressure gradient during the frost 
period. It is doubtful if the pattern of damaging temper- 
atures recorded for an area during a given frost period 
is ever other than wnzque. 
For the purpose of rapidly evaluating the frost 
damage to the crop, the citrus industry in the Far West 
makes extensive use of temperature-duration data ob- 
tained from thermographs exposed in a large number 
(approximately 400) of representative citrus plantings. 
Through the mechanical or manual integration of the 
individual thermograms, the analysts can estimate the 
probable frost damage with surprising accuracy within 
a day or two after the frost occurrence. Similar esti- 
mates based on field sampling of the fruit would require 
weeks to accomplish. 
Multiple Point or Areal Problem—Complex Time Re- 
lation—Multiple Climatic Element 
As we expand the variables into all dimensions the 
applied climatological problems become very involved. 
At the same time well-analyzed examples become rare. 
The problem of agricultural land usage in relation to 
climate has probably had more attention than any 
other. In agriculture we are dealing with complicated 
questions that can be disentangled only by the cooper- 
ation of four sciences: soil science, plant biology, land 
management, and climatology. 
But even with proper soil, proper tillage methods, 
proper fertilization, and proper seed material, the cli- 
matic conditions determine to a large extent what 
plants will or will not grow, mature, and ripen at a 
given locality. A most intricate system of time se-~ 
quences of various climatic elements enters into the 
problem. We find that photosynthesis is dependent on 
radiation intensities in specific wave lengths. In some 
