conditions depend on du/dh, it is to be expected that 
a better correlation will exist between the radio signal 
and the measured value of (@)(de/dh) than between 
the radio signal strength and the measured value of 
(—b) (dT /dh) 
to be the case; the correlation between the radio signal 
strength and the measured value of (a) (de/dh) is 
poor, while the correlation with the measured value of 
(—b) (dT/dh) is good. 
2. A situation similar to that mentioned in (1) has 
been encountered in attempting to forecast 12 hours 
ahead for operational radar sets. Table 1 summarizes 
the results obtained. 
The continuity method of forecasting consists in pre- 
dicting that tomorrow’s conditions will be the same as 
those observed today. Forecasting from the (d7’/dh) 
. The reverse has, however, been found 
APPENDIX 
term was empirical, it being assumed that a tempera- 
ture inversion of 1.5 © per 100 ft would give very good 
propagation conditions, an inversion of 0.75 C per 
100 ft would give good conditions, and a temperature 
lapse would correspond to standard or natural condi- 
tions. 
The value of du/dh calculated from equation (1) 
using the measured values of de/dh and d7/dhk are too 
small to explain the large signals’ frequently observed 
on a wavelength of about 80 me. 
These experimental results enable the following 
conclusions to be drawn so far as the meteorological 
conditions around the British Isles are concerned. 
1. The radio signal strengths computed from the 
observed M curve do not agree with observation. The 
method of observing the temperature and water vapor 
493 
gradients to obtain an M curve is therefore unsatisfac- 
tory. 
A reasonably satisfactory forecasting system can 
be obtained on the basis of temperature gradient alone. 
This system is empirical and does not depend on com- 
plicated mathematical computations. Accordingly the 
system may be suitable under operational conditions.® 
* Since November 1944 the Group at San Diego has 
found a good correlation. between the radio data and a simple 
meteorological parameter based on temperature excess. In the 
British Isles the temperature excess has to be taken between 
a height of about 200 ft and sea level, whereas at San Diego 
the temperature excess has to be taken between a height of 
about 5,000 ft and sea level. It has also been reported that in 
the Pacific Area temperature excess is by far the best index 
for predicting radio propagation conditions. There is some 
hope, therefore, that it may be possible to work out a method 
of fairly universal application on the basis of temperature 
excess, 
TABLES AND ILLUSTRATIONS IN APPENDIX 
MASSACHUSETTS BAY 
ED ete 
a [ 
- | 
") 
FLGURE, J. Map of transmission paths. 
