Chapter 11 

 STORM DETECTION 



"1 STORM DETECTION BY RADAR'' 



THIS SUMMER (1944) a study was made of the 

 meteorological echoes observed on a Canadian 

 microwave S-band early warning radar set at Ottawa. 

 These were correlated with check observations on the 

 weather made by a large niimber of local observers dis- 

 tributed over the area covered liy tlie set. It was found 

 that observers inside the source area of the echo always 

 reported rain; just outside the echo (1 to 5 miles) 

 there was a half chance of light rain. Atmospheric 

 electrical disturbance was present in less than half 

 the cases checked. Echoes became less frequent at in- 

 creasing distances from the set but in some cases were 

 seen at 160 miles. 



The display system that we have used is a plan posi- 

 tion indicator [PPI] tube. This tube provides a map 

 of a circular area centered on the location of the set 

 and extending out at choice to 40 or 80 or 160 miles. 

 The last-mentioned setting was the one used most of 

 the time. 



^^"'■* Procedure 



During the hours of operation a 16-mm motion- 

 picture camera was kept running, taking pictures of 

 tlie PPI display and a clock alongside, exposing one 

 frame of the film for tlie duration of each revolution 

 of tlie array. Thus about four photographs were ob- 

 tained per minute. At the same time we watched the 

 progress of moving echoes across the screen and made 

 telephone calls to any observers that were available, in 

 the neighborhood of any echo. From the observer the 

 existing state of the weather was determined ; his re- 

 marks and the exact time were carefully noted. 



We checked the echoes and their movement as re- 

 corded on the film against the information obtained 

 about the weather from the observers. We also made 

 charts of the echoes, based on tlie film, at 30-minute 

 intervals. 



"•1-2 Weather Information 



Facilities of which we availed ourselves for obtain- 

 ing information were, among others : 



» By Col. J. T. Wilson, Director of Operational Research, 

 NDHQ, Canada. 



1. Ottawa meteorological stations. These stations 

 provided us with as many as three forecasts a day and 

 with weather information generally. 



2. Distant meteorological stations. Apart from the 

 Ottawa stations, the nearest weather station, 57 miles 

 away, is at Canton, N. Y. Its reports are included in 

 tlie teletype sequences that come to us. 



3. Unofficial observers, consulted by telephone. Since 

 the official weather stations did not provide the close 

 network that we required, we compiled a list of per- 

 sons whose location could be closely marked on our 

 map and whom we could consult about existing 

 weather conditions to the extent that an untrained 

 observer would be competent. 



^^■'■^ Correlations 



Our aim was to correlate observed echoes with 

 weather conditions. At first clouds were thought to be 

 possible sources of echoes, and it was thought that 

 fronts might produce some sort of echo quite inde- 

 pendently of cloud or precipitation along the front. 

 The earlier part of our work showed that on after- 

 noons with heavy cumulus clouds but with no pre- 

 cipitation there were no weather echoes. On days with 

 scattered showers, however, echoes were observed. 

 Also, the passage of a front did not seem to produce 

 any peculiar sort of echo or any echo that could not 

 be attributed to precipitation along the front. 



In analyzing our correlations, we ha\e found it 

 convenient to consider them in two groups. First, there 

 are correlations icith echo, that is, correlations when 

 the observer was in the vicinity of an echo, although 

 not necessarily right inside the echo. All correlations 

 involving telephone calls to local observers were of 

 this type, for we didn't make such a telephone call 

 unless there was an echo in the vicinity. Second, there 

 are correlations iviih weather, or correlations witli 

 weather stations, when we first select an occasion wlicn 

 precipitation is reported (by an official station) and 

 then go looking in our records for an echo to match. 

 Nearly all the precipitation recorded at the weather 

 station during our hours of operation was light, and 

 too light, as it proved, for us to detect it at the distance 

 we were away. Thus there is only a very small number 

 of echoes associated with correlation with weather, 



187 



