1040 
basin only, since the production of rain over other 
basins is largely controlled by different orographic 
features. For nonorographic regions as, for instance, 
the Great Plains of the United States, the fixed controls 
are absent. Distribution of vertical velocities is, instead, 
a function of storm type, but the storm type is probably 
only remotely a function of geographical location within 
the limits of, say, a half-dozen degrees of latitude or 
longitude. In nonorographic regions, consequently, 
while fixed barrier heights are not at hand for use in 
computation, present meteorological knowledge permits 
use of data from storms that have occurred not only 
within the confines of a project drainage basin but also 
within a sizable area surrounding the basin. 
The first estimates of maximum possible precipitation 
over a drainage basin utilized the highest depths of 
record within that basin; if no large storms had hap- 
pened to occur withim the period of meteorological 
history of the basin, use of factors of safety was often 
adopted. A degree of refinement was then added when 
the rainfall records of adjoming basins were employed. 
With the introduction of the process of “transposition” 
of storm-rainfall values from one region to another came 
the need for meteorological studies of the major storms 
of record. Decisions—largely subjective, and based upon 
the weather-map-analysis experience of trained meteor- 
ologists—had to be made as to the area of trans- 
posability of each of the great historical storms. In 
conjunction with other procedures, storm transposition 
is still in wide use. It will undoubtedly continue to be so 
until accurate and reliable theoretical rainfall equations 
can be developed for general usage in nonorographic 
regions. Use of this procedure obviously requires DDA 
processing of large numbers of recorded storms. 
Moisture Adjustment. The equation of continuity of 
moisture indicates that rainfall depths increase with 
water-vapor content of the air, other factors remaining 
equal. Thus, it can be reasoned that a given storm of 
record would have produced greater rainfall values had 
its water-vapor charge been greater, or, in terms of the 
end product, were a “moisture adjustment” to be ap- 
plied to the DDA values. Wide usage is made of such 
a procedure, in combination with that of storm trans- 
position, for regions in which theoretical rainfall equa- 
tions have not been developed. 
Since nearly all of the great historical storms have 
occurred without available samplings of upper-air hu- 
midities, the use of surface indices has had to be 
adopted. For reasons already pointed out, the surface 
dew pomt—reduced to 1000 mb—is usually chosen as 
the index of the moisture charge of the air in which the 
rain is formed. In practice, for each storm of record, 
the dew point is selected at a location, nearest the rain- 
fall center, where tropical air is to be found at the 
surface. The method of selecting these values is de- 
seribed in detail in a report [30] which also lists the 
representative storm dew points of about 500 major 
storms in the United States. 
The so-called maximum possible dew points, to which 
the representative storm dew points are adjusted, are 
derived statistically from the records of first-order 
HYDROMETEOROLOGY 
Weather Bureau Stations. A close envelopment of the 
highest values—reduced to the 1000-mb surface—re- 
corded at these stations yields, for each month, a map 
of maximum possible dew-point isotherms. A detailed 
description of the process is given in a recent hydro- 
meteorological report [28]. Both the maximum possible 
and the representative storm dew points are 12-hr 
values; actually, they are the highest minimum values 
occurring through any 12-hr duration of the period of 
record of a station or during the period of a storm, 
respectively. 
Through storm transposition the DDA values of all 
transposable storms are applied to a project basin and 
each storm dew point is adjusted to the maximum 
possible for the season and new location. Envelopment 
of the transposed and adjusted values of all the storms 
yields a preliminary estimate of the maximum possible 
precipitation for the project basin. 
Various methods of performing the moisture adjust- 
ment have been developed. Of these, only the three 
which seem to be closest to physical reality will be 
outlined here. The first, called the cyclonic-adjustment 
method, has frequently been applied to large-area 
storms. Here, the adjustment is simply multiplication 
of the rainfall values by the ratio of W at maximum 
possible dew point to W at the representative-storm 
dew point, under the assumption of a saturated, pseudo- 
adiabatic atmosphere. Precipitable-water depths for a 
layer of air between 1000 and 300 mb are used in the 
calculations. 
The second, the thunderstorm-adjustment method, 
differs from the cyclonic-adjustment method in that 
the top of the layer for which precipitable-water values 
are used varies from 100 to 300 mb when the 1000-mb 
dew point varies from 78F to 50F, respectively. Choice 
of the range was based upon available data relating to 
the variation of cumulonimbus cloud tops with surface 
temperatures and dew points. This method, discussed 
at length in a hydrometeorological report [28] on 
generalized estimates of maximum possible precipita- 
tion, is intended for use in adjusting storms m which 
rainfall was mostly from cells of convective action. 
The third, or g-adjustment method, is intended to 
apply to the moisture adjustment of storms m which 
only the magnitude of the moisture charge is to be 
varied, and all other factors, such as moisture stratifica- 
tion, spatial dimensions of the storm cells, and magni- 
tude and distribution of wind, are to remain unchanged. _ 
As described by Fletcher [3], it is possible to factor a 
mean value of specific humidity from the equation of 
moisture continuity, and to introduce a factor of propor- 
tionality between the mean value and the value of 
specific humidity corresponding to the surface dew 
point representative of the storm’s moisture charge. 
For purposes of adjusting storms for moisture only, 
with all other characteristics to remain unchanged, use 
of ratios of surface specific humidity is more rigorously 
correct than use of precipitable-water ratios. However, 
it has not been established whether the concept that 
moisture charge can vary without any appreciable 
change in the rest of a storm’s characteristics 1s physi- 
