APPLICATION OF FORECASTING TECHNIQUES AND CLIMATOLOGY 



123 



the east coast ul' Xew Guinea when a Fohn wind tiows 

 out over a sea breeze, the latter producing simple sur- 

 face trai^iDing (Figure 18). (See Section 8.3.5.) 



Factors Affecting the Extent of Trapping. The prin- 

 cipal factors which determine the extent of trapping 

 are: 



EFFECT OF 

 FOHN WIND 



1 



Figure 18. Combination of types 4 and 5. (See Figures 

 15 and 16.) 



1. The amount by which M decreases through the 

 M inversion. 



2. The duct width, for the wider the duct the 

 more energy will be trapped. 



3. The elevation of the transmitter with respect to 

 the duct, the trajtping being most complete when the 

 transmitter is at the base of the M inversion. 



4. The angle at which the rays are propagated from 

 the transmitter ; the smaller the angle made with the 

 top of the duct, the greater the range. 



5. The frequency of the propagated waves; in 

 general, the higher the frequency, the greater the 

 extent of trapping. 



"^^ Specific Relationships Between 

 Meteorological Elements and 

 Radar Performance 



Reseaech on Foehcasting of Eadio 



AND EaDAK EaNGES 



Army Air Force Board Project. Eealizing the im- 

 portant and direct effects of temperature and humidity 

 distributions on microwave propagation, several proj- 

 ects have been undertaken in the attempt to develop 

 a systematic method of forecasting the meteorological 

 conditions leading to nonstandard propagation. Of 

 these methods, one which has met with a considerable 

 degree of success is descriljed below. The methodology, 

 developed by the Army Air Force Board working in 

 conjunction with the Radiation Laboratory at MIT,-'' 

 is designed to predict the formation of surface ducts 



over water. Its fundamental concepts are quite similar 

 to those used in other methods of radio and radar 

 forecasting.^" 



General Procedure. In essence the method consists 

 of an analysis of the modification that air undergoes 

 in the lower 1,000 ft as it moves from a large land mass 

 out over the ocean. The study was carried out in the 

 vicinity of Cape Cod, but indications are that the 

 numerical factors entering into the procedure are 

 much more generally applicable. In the modification 

 of the air moving over the sea, the following assump- 

 tions are made. 



1. The air initially (before moving off the land 

 mass) is well mixed, i.e., it exhibits conditions close 

 to neutral equilibrium (see Figure 11 ) . 



2. The stability conditions of the air as it moves out 

 over water are determined by its initial temperature 

 (over land) relative to that of the sea surface. 



3. The modified air at the sea surface acquires the 

 same temperature as the sea. 



J:. In the modified air at the sea surface the mois- 

 ture content becomes that corresponding to satura- 

 tion at the sea surface temperature, except for a cor- 

 rection owing to the salinity of the sea. 



5. The resulting M curve is determined by the 

 quantities : 



a. Temperature excess." 



b. M deficit.s 



c. Wind .speed and direction. 



d. Distance of over-sea travel (in some cases). 

 Thus the method attempts to relate duct formation 



to a limited number of easily determined meteoro- 

 logical factors. It involves a simplified consideration 

 of the ujDward diffusion of heat and moisture. It turns 

 out, however, that the simplified assumptions yield 

 results which in practical application are of sufficient 

 accuracy to be of definite use in forecasting the ex- 

 istence of nonstandard conditions. It should also be 

 mentioned that, although the method is designed 

 primarily for situations in which air over land moves 

 out over the sea, it can also be satisfactorily applied 

 to situations in which the air has a purely over-sea 

 trajectory. 



The particular steps to be taken in carrying out the 

 procedure follow. 



Method or Deteemining Duct Width 



Ohservation of Initial Conditions. The necessary 

 meteorological measurements to be taken should be 



^These terms are defined on page 124. 



