Chapter 3 
METEOROLOG Y—FORECASTING 
FORECASTING TEMPERATURE AND 
MOISTURE DISTRIBUTION OVER 
MASSACHUSETTS BAY * 
[2 lage GOING into a description of the forecast 
program and results it will be profitable to describe 
the method used in coordinating the observations. 
Meteorological Observations 
Soundings were made according to two major 
plans. The first was in conjunction with the radio 
path. According to that plan, airplane soundings were 
made once or twice a day at two or three points along 
the transmission path. The boat would take either 
mast or balloon soundings along the path while meas- 
urements at Race Point Light (Provincetown) would 
continue at 2- to 4-hour intervals during most of the 
day and sometimes at night. The Race Point Station 
had the advantage of being well away from land (for 
all but easterly winds) and soundings there would 
thus represent the condition of the air over a large 
portion of the path. The soundings just described 
were made primarily for correlation with the signal 
strength measurements. 
The second plan was to obtain soundings in succes- 
sive steps in air moving off the coast as that air be- 
came more and more modified by the cool ocean sur- 
face. For this reason, days during which the air was 
westerly or nearly so were set aside for this type of 
measurement. The Duxbury soundings gave a rep- 
resentation of the structure of the air before it left 
the land. One airplane made soundings at about 2, 
%, and 25 miles offshore. The times of take-off were 
staggered to allow the first airplane to complete its 
third ascent before the second plane would begin its 
first sounding. 
The boat played a vital role in such a plan. It ran 
along the line of the air trajectory for as long as was 
practicable to take water temperature measurements 
and mast soundings, usually an 8- to 10-hour period. 
The Race Point Meteorological Station was coor- 
dinated into this general plan by having it take con- 
tinual balloon soundings at, say, 2-hour intervals 
both before and after the airplane ascents. The pur- 
pose of these soundings was both to fit in as an extra 
sounding in the general plan and also to yield some 
information as to amounts of change of the meteoro- 
logical conditions with time. Also, in general, the 
"By I. Katz, Radiation Laboratory, Lt. J. R. Gerhardt, 
Lt. W. E. Gordon, Army Air Forces, and P. W. Kenworthy, 
U.S. Weather Bureau, Boston, Mass. 
235 
times of soundings at Race Point Station were sched- 
uled about 1 hour later than those at the overland 
station to give the air sufficient time to travel from 
one to the other. 
Forecast Program 
A forecast program was begun during July and 
continued to October 10, 1944, in order to try out 
existing methods of forecasting and to help develop 
new techniques. A more natural step would have been 
to analyze the data taken during the summer and 
then to put that analysis into the form of forecast 
procedures, as was done at the end of the 1943 Boston 
Harbor transmission experiment. However, since speed 
was essential it was decided to initiate a forecast pro- 
gram simultaneous with the observations. The very 
act of forecasting tended to focus attention on the 
important weather factors, at the same time giving 
invaluable help in planning the day-to-day observa- 
tions 
The type of forecast made was different from the 
usual form. It consisted of a “space forecast”? rather 
than the usual time forecast. That is, knowing the 
conditions at one point at a given time the problem 
was one of finding the conditions at another location 
at the same time. It involves the entire problem of 
modification of an air mass by a water surface. 
The forecasts were in the form of curves of tem- 
perature and moisture, from which the modified in- 
dex curve was computed. A time and a location in 
Massachusetts Bay were selected at which it had been 
determined previously that a sounding would be made. 
Almost invariably airplane observations were chosen 
to use as verifications because those soundings were 
at sufficient altitudes so that both the modified and 
the unmodified air were sampled. The forecasts were 
made from the surface to 1,000 ft, whereas the air- 
plane soundings started from about 20 ft and con- 
tinued to 1,000 ft. For verification, the forecast and 
the sounding were plotted on the same graph. 
Army Analysis and Forecasts 
The program of the Army forecasters included the 
forecasting of transmission and radar ranges; the ap- 
proach to this problem was empirical. The basis of the 
program was again the analysis of the first 6 weeks’ 
data, this time including the signal strength meas- 
urements which have been described. Signal strengths 
were divided into four ranges qualitatively described 
