during the project, the horizontal velocity matched within 2 percent of the car- 

 riage velocity. The time delay has little effect on the validity of the calibra- 

 tion because there are no individual parts or elements in acoustic Doppler 

 systems that have significant "drift." 



Sampling and time reference 



Two types of data files were collected with the 2.4-MHz BB-ADCP. The 

 first is based on single ping velocity estimates, recorded at approximately 

 1.8 Hz. The other file contains data that are averaged over 20 or 50 pings, 

 recording one average profile every 3 to 8 sec. 



The 600-kHz system collected three different types of data files. The first 

 contains raw data from individual pings recorded at a rate of approximately 

 6 Hz. These pings alternate between short pulses (two carrier cycles) and 

 longer coded pulses (34 carrier cycles). Due to the large amount of data 

 collected in this mode (typically 12 KB/sec), these files were generally only 

 collected over 2-min intervals. In the second configuration for the 600-kHz 

 system, time-averaged root-mean-square values of the signal were recorded. 

 The signal was averaged over a 40-sec interval, using a short pulse, and each 

 rms file covers approximately 8 min. In the third mode, various pulse combi- 

 nations were tried for later experimentation with Doppler calculations. 

 Appendix H contains a complete list of file naming conventions, for both 

 systems, as well as a list of files recorded with each wave set. 



The computer operating the 600-kHz unit was hooked up to a UTC satellite 

 receiver. This receiver continuously updates the computer clock from infor- 

 mation contained in the satellite transmissions. The computer clock is subse- 

 quently used to tag the data files. The time was also transmitted over the 

 serial line at regular intervals to update the clock in the computer operating 

 the 2.4-MHz BB-ADCP. The clock in the computer operating the 2.4-MHz 

 system was noted to drift 1 to 5 sec over a 1-hr run. The time was thus not 

 synchronized as well as hoped for. 



Pre-processing and filtering 



The BB-ADCP measures the velocity by measuring the phase shift of the 

 echo from two or more transmit bursts. In "short-range mode" (or pulse-to- 

 pulse coherent mode), the second pulse is transmitted after the first pulse has 

 died down, leading to a velocity estimate with low statistical uncertainty. 

 Through a series of algebraic manipulations, along-beam velocities can then be 

 used to measure the vertical profile of turbulent kinetic energy and of 

 Reynolds stress (Lohrmann, Hackett, and Roed 1990). 



Due to the inherent problem of measuring phase changes larger than ±tt 

 ("wrap-around" or "ambiguity problems"), the measurement scheme required 

 to operate in short-range mode is complicated and requires pre-processing to 



Chapter 8 ADCP Measurements at SUPERTANK 



147 



