192 BELL SYSTEM TECHNICAL JOURNAL 



atmosphere. The instruments used to measure temperature and humidity 

 require a few seconds to reach equilibrium — a length of time comparable at 

 times with the period of fading. To measure the variation of dielectric 

 constant with height, the measuring instruments are usually carried aloft 

 by means of captive balloons. A half hour may be required to measure 

 to heights of six or seven hundred feet with the result that the final curve 

 represents an unknown combination of variations of dielectric constant with 

 height and with time. It is also extremely doubtful that the atmosphere is 

 uniform in the horizontal plane — an assumption which is usually made in 

 the theoretical treatment of microwave propagation. It seems likely that 

 the lower atmosphere is far from being a homogeneous fluid but rather may 

 contain small air masses or ''boulders" with properties which differ con- 

 siderably from these of the surrounding air. Reflections from these boulders 

 may be the cause of radar echoes received from the lower atmosphere.^ 

 Scintillation fading of microwaves is another evidence of these inhomo- 

 geneties in the atmosphere. Scintillation fading, a rapid fluctuation in 

 signal level about a more or less steady average value, increases as the wave- 

 length becomes less and as the path length is increased. 



In order to evaluate, on a statistical basis, the effect of atmospheric 

 changes on a typical microwave circuit, extensive measurements of trans- 

 mission were made over a 40-mile overland path between New York City 

 and Neshanic, New Jersey. The tests covered a period of about two years. 

 Most of the data were obtained at wavelengths of 10, 6.5, and 3.2 centimeters 

 although some data were taken at wavelengths of 42 centimeters and 1.25 

 centimeters. The results are described in a recent paper.® In many 

 respects, observations were in agreement with those made earher on the 

 39-mile Beer's Hill-Lebanon path at wavelengths of 4 and 2 meters and on 

 the 38-mile non-optical path between Deal, N. J. and Lawrenceville, N. J. 

 at a wavelength of 2 meters.'^ The same seasonal and diurnal trends in 

 fading were found; transmission was generally steady during the midday 

 hours and during periods of windy or rainy weather; fading was the same 

 on vertical and horizontal polarizations. However, the character of the 

 fading was different ; the fading at microwaves was much faster and deeper 

 than that observed on the ultra-short-wave path. The average daily fading 

 range for July on the New Yoik-Neshanic path was 20 db at 6.5 centimeters 

 compared with a median daily fading range of 8.5 db for 2.0 meters observed 

 in July on the Lebanon-Beer's Hill path. 



» H. T. Friis, "Radar Reflections from the Lower Atmosphere", Proc. I. R. E., vol. 

 35, pp. 494-495; May 1947 (Correspondence Section). 



" A. L. Durkee, "Results of Microwave Propagation Tests on a 40-mile Overland Path", 

 Proc. I. R. £., vol. 36, No. 2, pp. 197-205, Feb. 1948. 



i« C. R. Burrows, A. Decino and L. E. Hunt, "Stability of Two-Meter Waves", Proc, 

 I. R. £., vol. 26, pp. 516-528; May 1938. 



