METEOROLOGY — FORECASTING 241 
the balloons, which were released from Army crash 
boats at distances of 20 and 60 miles from the coast. 
Possibly because of lack of radar efficiency, only the 
balloons released at 20 miles were picked up by the 
coastal radar. Although no nonstandard conditions 
were observed during the releases, the method seems 
suitable for making vertical coverage measurements. 
Radar and weather data for the period January 1 
to March 15 were tabulated and analyzed during the 
month of March. The primary data consisted of S- 
band radar reports from Winter Garden and Lees- 
burg and low-level soundings from Leesburg, sup- 
plemented by the synoptic charts and radiosonde ob- 
servations supplied by the 26th Weather Region. 
The analysis resulted not in a system of forecasting 
such as that developed for over-water use but rather 
in a series of clues to be used as an aid to over-land 
forecasting in Florida. Since the clues are closely 
related to the topography and peninsular situation 
of Florida and to the season of the year, they are not 
directly applicable to other locations in their present 
form. However, they suggest that investigation of 
these points at other locations would quickly yield 
useful correlations. Some examples of these relations 
follow. 
1. Of the 600 ft low-level sounding standard curves, 
90 per cent gave standard ranges. 
2. During early morning hours: 
a. Surface winds of 10 mph or more produced 
standard propagation always; winds of 5 to 9 
mph produced superrefraction 20 per cent of 
the time; 2 to + mph showed superrefraction 
60 per cent of the time; and calm winds pro- 
duced superrefraction almost always. 
b. Similarly, the 1,000-ft winds of 30 mph or 
more produced standard, while 1,000-ft winds 
of 10 mph or less almost always produced 
ssuperrefraction. 
e. Superrefraction occurred with clear skies ex- 
cept on two occasions, one with broken high 
clouds, the other with broken middle clouds, 
never with low clouds. 
d. High-pressure centers within 700 miles and with 
gradients of 1 mb per 100 miles or less pro- 
duced superrefraction. 
e. Ground fog patches were observed during pe- 
riods of class 4 propagation, with two excep- 
tions. 
3. Simple surface ducts of 70 ft and AM,, (refer- 
ence level 50 ft) of 4 or more and elevated S curves 
with ducts above 200 ft and AM,, of 6 or more pro- 
duced class 3 or 4 propagation with possibly one ex- 
ception. 
4, Large AM’s observed by radiosonde between 
1,000 and 3,000 ft showed no correlation with S-band 
propagation but did show fair correlation with super- 
refraction on P band. Superrefraction on both S and 
0030 EWT 
0130 EWT 
0300 EWT 
0400 EWT 
Figure 5. SCR-588, Leesburg, night of March 5 to 6, 
1945. 
P band showed good correlation with large AM’s ob- 
served below 1,000 ft. 
5. The height of the temperature inversion in- 
creased with increasing 1,000 ft wind speeds up to 
10 or 12 mph, then decreased slowly with further in- 
crease in wind. 
6. Substandard propagation conditions were never 
observed over land, either on the radar or the sound- 
ings. 
A series of low-level soundings taken at hourly in 
tervals throughout the night were related to corre. 
sponding radar ranges. The soundings were made at 
Leesburg; the radar data were taken at Tomato Hill 
(2 miles west of the sounding site) and at Winter 
Garden (25 miles southeast of the sounding site). 
The general weather situation for the night of 
March 5 to 6 shows maritime tropical air pouring up 
over Florida around the western end of the Bermuda 
high, giving clear skies and southerly winds of 10 mph 
at 1,000 ft and 2,000 ft at 2000 EWT,° increasing to 
20 and 25 mph respectively by midnight, and to 23 
and 35 mph by 0400. Figure 5 shows the PPI scopes 
of the P-band SCR-588. The arrows point north; the 
small grid squares are 5 miles on a side. Before mid- 
night, propagation was standard, as illustrated by the 
0400 frame. The ranges built up rapidly, reaching 65 
miles and decreased slowly between midnight and 
0400. (Radar shadows of surrounding topographical 
features account for the uneven distribution of range 
increase. ) 
Figure 6 shows the progression on the S-band SCR- 
584. The bold line points north, the range markers are 
at 10,000-yd intervals. (The sounding site is roughly 
315° at 45,000 yd.) From 1900 to 2200 propagation 
All times to follow are Eastern War Time. 
