— 
METEOROLOGY — FORECASTING 237 
the M inversion in both very close to 150 ft. This was 
one of the better forecasts. It can be seen that actual 
values of temperature and vapor pressure between 
forecast and verification might vary by several de- 
grees, but as long as they have the same slopes at the 
same elevation they will produce M curves reason- 
ably close to one another. 
Conclusions. It is felt that the method of forecast- 
ing used was a marked improvement over the tech- 
nique employed previously. It is potentially capable 
of dealing with the low-level modification problem 
in the case where the intial air is stratified as well as 
the one in which the initial air is homogeneous before 
it passes out to sea. 
RADAR PROPAGATION FORECASTING» 
This is a report of results obtained by an AAF 
board project investigating radar propagation fore- 
casting, which was started as two distinct programs in 
September 1944. The first part of the project was 
carried out at the Radiation Laboratory with facili- 
ties used by Group 42 during over-water transmission 
Measurements in the summer of that year. During 
that time, with the invaluable assistance of Group 42, 
a forecasting system was developed for the over-water 
case, the results of which are presented in previous 
text® These reports gave preliminary results of the 
MIT program and the recommended forecasting pro- 
cedures. The second part of the propagation forecast- 
ing program was set up at Orlando, Florida, to study 
particularly the over-land forecasting phase and to 
investigate some of the operational uses of such fore- 
casts. 
With this in mind, AAF Board Project H3767, 
“The Determination of the Practicability of Forecast- 
ing Meteorological Effects on‘ Radar Propagation,” 
was initiated late in 1945 with the following specific 
objectives : 
1. To determine the practicability of forecasting 
those low-level meteorological conditions which affect 
radar propagation. 
2. To determine the accuracy with which radar 
propagation forecasts can be made from the corre- 
sponding meteorological conditions. 
3. To determine the operational uses of such fore- 
casts. 
4. To determine the optimum meteorological ob- 
servation site with relation to the site of the radar 
employing the forecasts. 
5. To determine the suitability of available low- 
level sounding equipment. 
It was originally planned to study the over-land 
and over-water problems simultaneously, but because 
of the lack of a coastal radar site until the last month 
of the program the project was divided into two 
>By Lt. J. R. Gerhardt and Lt. W. E. Gordon, AAF Board. 
°Elaborated in references 1 to 3. 
phases: (1) the general study of the over-land prop- 
agation variations in an attempt to devise a suitable 
forecasting procedure and (2) an evaluation of the 
results obtained from both the over-water and over-land 
methods, with possible tactical applications under field 
conditions at the site of a powerful coastal radar. 
Figure 2 is a map of central Florida giving in de- 
tail specific facilities used throughout the project. 
Headquarters was established at the Weather Central, 
Orlando, where complete weather information, fore- 
casts, and analyses were available. The meteorological 
data used throughout the project consisted of surface 
and upper air observations for the general central 
Florida area. 
Detailed synoptic maps of Florida were drawn cov- 
ering periods of 6 hours each to determine wind 
patterns and representative land temperatures and 
dew points ; piballs* for Orlando, Sebring, and Tampa 
were plotted up to 2,000 ft to determine trajectories and 
wind speeds above surface levels, while the Orlando, 
Tampa, Tallahassee, and Jacksonville raobs were 
studied to correlate subsidence and radiation effects 
with radar propagation variations. : 
During the first phase of the project, sounding sta- 
tions were established at Leesburg and New Port 
Richey using both the Washington State College 
wired sonde and the MIT psychrograph. Radar data 
were taken from the S-band V beam and the P-band 
SCR-588 at Tomato Hill, only a few miles from the 
Leesburg sounding site, the SCR-584 at Winter Gar- 
den, and the P-band SCR-271 at Crystal River dur- 
ing their operating hours. The Tarpon Springs pro- 
gram employed a medium early warning and an SCR- 
615 radar, both on S band, located on the Gulf coast, 
and the Crystal River SCR-271 and Winter Garden 
SCR-584. The sounding station was located within 
a half mile of the Tarpon Springs radar site. During 
the entire program sea surface temperatures were 
measured several times weekly at either the Cedar 
Keys or Anclote Crash Boat bases out to a distance 
of about 20 miles at 2- to 4-mile intervals. 
Low-level soundings were made during the entire 
project primarily as an aid in interpreting radar per- 
formance and in determining representative air values 
and secondarily in an attempt to evaluate the opera- 
tional suitability of the available sounding equip- 
ment. Results of the latter portion of the work will 
be presented later in this report. Ground-based sound- 
ings were made by means of various combinations of 
350- and 700-g balloons, 7-ft Seyfang kites, and a 
small barrage balloon. 
The sounding stations were originally located so 
as to be as representative as possible of interior and 
coastal areas, although it was found later that, with- 
out the additional mobility of airborne measurements, 
4A small balloon with standard rate of rise released for 
tracking by a theodolite for estimation of upper-air winds. 
