used to play these tapes into the Modcomp so that the data record could be 

 digitized and tapes compatible with the telephone-line system generated. From 

 August 1978 through September 1979, the Modcomp was not operational; conse- 

 quently, only the FM recorder was used for data collection. 



Regardless of which system was used, the voltage signal from the sensors 

 required certain conditioning. For the Modcomp system, the signal was first 

 amplified and biased to ensure a 0- to 5-volt range, then converted to a 

 frequency-modulated signal by exciting a voltage controlled oscillator (VCO). 

 That signal was then transmitted to Fort Belvoir via telephone line where a 

 discriminator was used to convert back to a voltage signal. This signal was 

 fed into a demultiplexer to convert to a serial data stream which was then 

 sampled by the Modcomp. For the FM recording system, the 0- to 5-volt signal 

 was fed directly to the FM recorder which operated on a maximum output of 3 

 volts; thus it linearly scaled the 0- to 5-volt signal by a factor of 3/5. 



b. Data Collection. The signals from the BEM wave sensors were sampled 

 four times per second for 20 minutes every 6 hours, beginning as near as pos- 

 sible to 0100, 0700, 1300, and 1900 hours eastern standard time (e.s.t.). 

 These hours correspond to the times used for the NWS daily sjmoptic weather 

 maps. Since the Modcomp system was automated, data were recorded during 

 nonduty hours and on weekends-holidays. The FM recorders were run manually 

 and for most dates only two observations, one in the morning and one in the 

 afternoon, were obtained. In general, the FM recorder was not run on the 

 weekends and holidays unless there was a particular event in progress, such as 

 a storm or experiment. If raw data records are to be obtained from CERC, it 

 is important to determine which recording system was used since the digital 

 data from the FM system require a 5/3 amplification to exactly duplicate the 

 scale of the data recorded via the telephone-line system. 



The Baylor wave staff gages required little maintenance except to keep the 

 biological growth cleaned off the cables and to replace defective parts, e.g., 

 transducers after a lightning strike. The staff gages including the trans- 

 ducer elements and associated electronics were calibrated at least once a year 

 and each Baylor transducer element was calibrated prior to installation (see 

 App. A for gage histories and the dates new transducers were installed) . 

 Recent tests indicate that gage calibrations performed this often ensure 

 accurate wave height information but that the amplification electronics should 

 be checked and adjusted at least monthly to obtain accurate water level infor- 

 mation from the staff gages. 



At least every 9 months the two Waverider buoy gages were rotated with two 

 that had been cleaned, repainted, and new batteries installed. CERC did not 

 calibrate the buoys during 1978 and 1979; however, recent semiannual calibra- 

 tions of the same buoys show an 8-percent or less amplitude response error in 

 the 10-second period range when compared to the manufacturer's calibration 

 curve. The Datawell Company believes this has been a gradual deterioration 

 of the accelerometer system that started when the buoy was initially put into 

 service; thus, there is reason to believe that during 1978 and 1979 the error 

 was considerably less than 8 percent. 



c. Data Tapes . The wave data from January 1977 through December 1979 

 were recorded in digital form in the following basic tape format: two records 

 of header information which included the station identification number, the 



15 



