reverberation loss calculations are provided for Fleet use (Fig. 2^4-). 

 Convergence zone and bottom bounce ranges and recommended dip depths 

 and tilt angles are provided so that ships can select the optimum 

 mode of operation. Ranges are presented for in-layer and below-layer 

 targets and various ship speeds, based on a 50^ detection probability 

 of a random aspect target by an unalerted operator. And the range cal- 

 ciilations are based on noise limited performance figures corresponding 

 to sonar type and ship speed. 



What now of the future? Are there new products of a revolution- 

 ary nature just over the hill which might eq.ual the breakthroughs 

 we have had in recent years? The consensus among Naval Weather 

 Service people is that the period just ahead of us will not be one 

 of exciting discoveries, with an exception to be discussed, but one 

 of smoothing the rough edges off existing models to get improved 

 forecast accuracy. 



As I have pointed out, the poor data base with which we work 

 limits the efficiency of the models. There are a few possibilities 

 for increasing our data base. Satellites would be the most likely 

 contributor in the near future. The newest weather satellite appears 

 to be able to define sea surface temperature and vertical tempera- 

 ture structure of the atmosphere. If this is so, and reasonably ac- 

 curate soundings can be obtained over most of the earth's surface, 

 our n;;imerical atmospheric analyses will be greatly improved, with an 

 assTored by-product of improved oceanographic forecasts. Other data 

 sources which could help, if they ever come into being, are buoys 

 and an expanded observational program in merchant ships. 



There is work being done at Monterey on a new product which 

 could have tremendous impact upon both meteorological and oceano- 

 graphic forecasting. This is a new analog system which can provide 

 forecasts for periods of from one hour up to ten days. An analog 

 is selected by computer- search through hemispheric historical data 

 to find that day when the state of the atmosphere matched or nearly 

 matched that existing today. Once found, it is assumed that the 

 atmosphere will behave in the future just as it did following the 

 historical day we selected for our analog. The idea is not new at 

 all, but the complicated method of selecting the analog by computer 

 is. The analog system is being tested right now, and has shown skill. 

 With improvement and refinement, we may soon be able to provide 

 prognostic charts out to ten days which have enough accioracy to be 

 used for forecasting any atmospheric and many oceanographic para- 

 meters. 



23 



