Most arguments against reconnaissance reports (other than high cost) 

 are that such reports are usually inaccurate. Arguments such as these 

 are a carry over from previous years when aerial navigation was accom- 

 plished soleh^ by dead reckoning, celestial and radio means, or a com- 

 bination of all. By use of long-range navigation (LORAN) it is possible 

 to fix positions within 1 to 5 miles of the actual and with such frequency 

 that other methods are seldom used. In case of LORAN failure, air 

 plot to a fixed position and recomputation will give position to within 20 

 to 30 miles. Considering the scale of our present synoptic charts, both 

 are negligible as far as practical forecasting is concerned, as the centre or 

 " eye " of a well-developed typhoon will generally have a diameter as 

 large as the maximum navigational error. If the argument is advanced 

 that such error is too great it can be truthfully said that errors arising 

 from triangulation fixes from surface reports are much greater due to the 

 fact that the gradient wind must be ascertained in order to actually obtain 

 a fix from triangulation and also the accuracy of a triangulation fix is 

 directly proportional to the symmetry of the typhoon. 



The degree of error in wind computations seldom exceeds 15 degrees 

 and 5 knots, and when multiple drift winds are obtained using modern 

 drift equipment the error is much less. Due to rapidly changing 

 directions and velocities near typhoon centres actual winds are difficult 

 to obtain. Radar fixes are valuable in wind computations as the wind 

 is vectorially computed by use of true airspeeds and headings compared 

 with the fix. However, the degree of accuracy in such computations is 

 directly proportional to the distance between fixes up to the neighbour- 

 hood of 100 nautical miles. Some error does exist in such computations, 

 as stated previously, due to the ever-changing directions and velocities 

 found in typhoons, and the error increases with distance. The pressure 

 differential or Bellamy method is fairly accurate in giving cross-wind 

 drift components by comparision with true airspeed, and when ground 

 speed is ascertained by radar fixes will give a fairly reliable wind reading. 

 In all instances, though, when flying at angles to isobars the varying 

 directions and velocities make extremely accurate wind computations 

 difficult. Actually the observation of the ocean's surface gives the best 

 indication of wind velocity as the observer soon learns to associate 

 wind velocity with types of white caps, sea swells, patches of foam, 

 or complete obscurement of the surface by froth. With the centre 

 accurately determined by a LORAN fix the direction of wind near the 

 centre is not too important ; however, knowledge of the velocity is of 

 paramount importance in typhoon warnings, but the fact remains that 

 precautions for an 80-knot wind would be about the same as for one of 

 90 knots, and if it can be said that the velocity is over 75 or 80 knots, 

 and such observation is correct, the warning service is more than 

 adequate. In storms which have not reached typhoon intensity wind 

 velocities can be computed to the nearest 5 knots as cloud covering is 

 usually such that an increase or decrease in altitude of 100 feet will 

 allow a multiple drift computation. 



True altitude, temperature, humidity, and pressure readings are 

 exceptionally accurate as they are scale or scope readings or mathe- 

 matical derivations of both. Maximum errors are 50 ft. in altitude, 

 one degree Centigrade of temperature, 5 per cent, of relative humidity, 

 and at low levels surface pressure is computed to the nearest milHbar. 



16 



