RADIOSONDE LAG CONSTANTS 39 



when the lithium chloride element measures relative humidity it must be 

 used with the saturation vapor pressure. Since the saturation vapor 

 pressure of water is a function of temperature, an error in temperature 

 produces an error in the estimated water vapor pressure. 



The lag constant of the lithium chloride (LiCl) humidity element be- 

 comes significantly larger for temperatures below °C [42]. Wexler [11] 

 has made detailed studies of the lag constants of the LiCl elements at low 

 temperatures under laboratory conditions. Bunker [43], however, has 

 found quite different lag constants in the free atmosphere. He attributes 

 this discrepancy in lag constants to the laboratory-determined values 

 which were obtained for isothermal conditions; whereas, in rising through 

 the free atmosphere, the radiosonde normally passes from warm to cooler 

 air. Bunker has raised a serious question, namely that the temperature 

 lag of the lithium chloride element is possibly as important as the iso- 

 thermal humidity lags studied by Wexler. It is quite possible that the 

 interplay of these two lags could produce a total effect either greater or 

 less than the humidity lag alone. We now have a quandary since there is 

 not currently in the literature a complete analysis of the interplay of the 

 temperature and humidity lags of the LiCl humidity element. Although 

 Bunker considers some aspects of this problem, he does not consider the 

 case of decreasing humidity and increasing temperature, a typical condi- 

 tion giving rise to the superrefraction of radio waves. The differences 

 between Wexler's and Bunker's estimates of the LiCl lag constants are so 

 great as to make one wonder at the validity of applying any correction for 

 this effect. Since neither Wexler's nor Bunker's tabulation of lag con- 

 stants is complete, the choice of lag constant appears to be arbitrary. 

 Wexler's lag constants will be adopted for the present discussion. Both 

 the temperature and humidity elements are corrected (as much as is 

 possible) for their respective time lags in this chai)ter for the purpose of 

 preparing refractive index profiles. 



In what follows the theory of sensor time lag will be examined. Data 

 from several climatically diverse locations will then be examined to 

 illustrate the relative importance of the various lag constant corrections 

 under conditions of superrefraction of radio waves. 



2.5.2. Theory of Sensor Time Lags 



Middleton and Spilhaus [44] give 



f = - -X (^' - "•) (2.1) 



as the basic differential equation of the time lag of a meteorological 

 sensor, measuring the variable 6, where t is time, X the appropriate lag 



