984 CLIMATOLOGY 
taining satisfactory climatic information for the design 
of shoes from these conventional observations needs no 
further elaboration [10]. 
Single Point—Complex Time Relation—Single Climatic 
Element 
In most existing works on problems of climatology the 
whole series of climatological observations has been 
used. All observations are accorded equal weight. The 
time factor enters only as length of the series or as a 
periodic variation, such as diurnal or annual changes. 
Many problems, however, include the duration of a 
climatic condition, or duration above or below limiting 
values, or intensities per unit time. These complex 
time relations present difficulties. Usually one difficulty 
is the lack of continuous records of the climatic element. 
A very simple example in this group is the design of a 
heating system for the melting of snow falling on a 
pavement. For this purpose, the rates of fall of snow 
(or sleet) must be known. We must construct amount- 
frequency-duration curves for the particular locality 
(generally only for temperatures below the freezing 
point). Complications are encountered in establishing 
rates and durations of snowfalls in the absence of 
heated recording precipitation gages. A first approxi- 
mation can be arrived at from cumulative 12-hr or 
24-hr precipitation measurements, temperature data, 
and hourly or three-hourly reports of the synoptic 
network. 
A rather simple example of the type of problem 
associated with this class is afforded by the evaluation 
in situ of the “frost hazard” to tender fruits and 
vegetables. For each fruit or plant (at a given stage 
of development) there exists a minimum temperature 
which constitutes an upper limit for damage through 
freezing. Freezing damage to the fruit, blossom, or 
plant will not occur at temperatures above this level. 
Nevertheless, experience and experiment have shown 
that freezing damage is a function of both the actual 
minimum temperature and the duration of temperatures 
below the critical limit for freezing. For example, a 
fifty per cent commercial damage to a particular crop 
can result from a minimum temperature of 28F with 
a duration of four hours, whereas no more than fifteen 
per cent damage might result from the situation im 
which the minimum temperature is 26F for a duration 
of thirty minutes, below 27F for forty-five minutes, 
and below 28F for one hour. For this reason, attempts 
to correlate frost damage and the frequencies of mini- 
mum temperatures have seldom been successful, al- 
though excellent results have been obtained when the 
temperature-duration data have been substituted [89]. 
In evaluating the probability of frost damage, time 
becomes a further complex to the extent that the dates 
of susceptibility of the plant or crop to freezmg may 
vary widely from year to year. For example, in the case 
of deciduous fruit crops, the most susceptible period for 
freezing damage is during the blossom (or small green 
fruit) stage of development. However, the phenological 
data from a single orchard between earliest and latest 
dates of blossoming may show an extreme range of five 
weeks. Obviously, the probability of a complete or 
partial loss of crop through frost damage is much 
greater in an early-bloom year than when the blossom- 
ing period occurs so late in the season that there exists 
a small probability that freezing temperatures will occur. 
A somewhat similar problem exists in the evaluation of 
the frost hazard during autumn to the seed corn crop at 
points in the Middle West [79-81]. In the latter case the 
probability of frost damage to the crop is increased 
during a season of late seed maturity. In neither case will 
a simple frequency distribution of the occurrences of 
critical temperature durations according to calendar 
dates give a true measure of the probability of frost 
damage to the crop. In all such agricultural problems, 
the reference base is the “crop calendar,” not the 
Gregorian calendar. 
Another intensity-duration problem of a single ele- 
ment at a single locality concerns solar radiation as 
related to the design of window surfaces, blinds, and 
awnings. Their practical purpose may be either the use 
of solar energy for supplemental heating, or the elimi- 
nation of excessive radiation. Theoretically, the problem 
is readily solved. Practical difficulties arise from the 
lack of radiation measurements, especially on exposures 
other than the horizontal surface [35, 40]. 
Another example which illustrates this category of 
problems deals with the exploitation of wind for power 
production. In contrast to other possible examples we 
are favored in this instance by the existence of an 
excellently documented case history [71]. A brief sum- 
mary of the meteorological phases of the New England 
experiment for engineering exploitation of the wind is 
given below. In it we follow Putnam’s conclusions 
closely. 
There are many basic questions concerning wind 
behavior that have to be answered in the selection of a 
site for a wind power plant. In the specific case analyzed 
here, the choice of a suitable locality for a test site was 
an important problem. We can pass over the restricting 
operational requirements, such as accessibility and 
proximity to the center of maximum electric load, 
although in any practical case these limit the choice 
between all meteorologically favorable sites. 
The basic ingredients for a preliminary climatic 
evaluation are: 
1. Free-air wind speed at mountain-top height. 
2. Effect of the geometry of a mountain on retarda- 
tion or speed-up of wind flow over the summit. 
3. Prevailing wind direction. 
4. Influence of the turbulence structure of the wind 
on design. 
5. Influence of the wind structure on estimates of 
output. 
6. Influence of atmospheric density on estimates of 
output. 
7. Influence of estimates of icing at higher elevation 
on design. 
Factor 6 can be derived from theoretical considera- 
tions with sufficient accuracy; factor 7 is of importance 
