EXAMPLE OF A MEDIUM-SIZED SYSTEM 



The Ocean Physics Group of the Applied Physics Laboratory, 

 University of Washington has operated a medium-sized computer 

 system on five major cruises since February 1966. The computer 

 is an IBM 1130 with card, paper tape, printer, plotter and 

 magnetic disc input-output. To its interface channels are 

 attached devices for reading and writing voltages, frequencies, 

 switches, magnetic tape and a complete shipboard acoustic tracking 

 system. The system has been used on-line with acoustic propaga- 

 tion experiments, sound velocity, salinity, temperature and depth 

 measurement and tracking of many types of in-water apparatus such as 

 our self-propelled underwater research vehicle. 



A complete set of programs has been written for our system to 

 allow us direct access via the FORTRAN language to our experiment 

 sensors. This has resulted in significant flexibility to alter 

 experiments in progress via simple program changes. The program- 

 ming allows us to read volts, frequencies or whatever by easily 

 understood program steps. This eliminates the necessity of 

 having each of us be a computer expert as is the case for the 

 dedicated computer. 



An example of this flexibility can be seen in an experiment 

 performed by the Ocean Physics Group to measure vertical acoustic 

 transmission fluctuations. It was decided while on a cruise that 

 it would be interesting to examine vertical acoustic transmission 

 fluctuations and correlate these with temperature data taken to 

 examine internal wave spectra. The experiment was designed, the 

 equipment was attached to the computer, two very simple programs 

 were written, the data were taken and some analysis was done - 

 all during a three-day period at sea. The computer system in 

 this case replaced a great deal of black box electronics which 

 would have been required to perform such an experiment. 



Another example of the on-line use of a computer is the Ocean 

 Physics Group's sound velocity, salinity, temperature and depth 

 system. The frequency output from the sensors is directly 

 connected to the frequency-to-digital converters attached to the 

 computer. The data can be taken under program control by time 

 increments or depth increments. Taking data by depth increments, 

 which is possible only with a computer, requires a computer 

 algorithm which will take data while the SVSTD probe is traveling 

 one direction. This procedure eliminates differential flushing 

 errors and produces a clean data file monotonic in depth. Further 

 analysis of SVSTD data is performed to produce results available 

 almost immediately for planning our other experiments. This rapid 

 tum-arovind. time is essential to us in planning acoustic, 

 microstructure or diffusion experiments. 



39 



