(2) Special Surveys . These surveys were taken along profiles at 

 ranges 0.5 foot apart. The ranges in the plots and tables are numbered 

 from 05 to 60 in the 6-foot tank and 05 to 99 in the 10-foot tank. Eleva- 

 tions were determined at stations 0.5 foot apart; however, some of the 

 surveys did not cover the full length of the profile. 



b. Elevation Measurements . Elevations were measured using 3-foot- 

 long (0.91 meter) gages mounted on the instrument carriage (Fig. 3). The 

 point on the gage was replaced with a 1-inch-diameter (2.54 centimeters) 

 gimbaled foot. The pivoting ability of this foot allowed it to attain the 

 slope of the local sand surface and the large diameter prevented the foot 

 from penetrating the profile surface, excep* in rare circumstances where 

 the readings were obviously wrong. The gages, graduated in thousandths 



of a foot, were read to the nearest hundredth. 



c. Survey Datum . To ensure that all elevations were measured to a 

 fixed datum, a carriage elevation survey was taken annually, measuring the 

 distance of the carriage above the water surface at different ranges and 

 stations along the tank. A special gage was used to survey the distance 

 from the carriage to the water surface. The point on a gage was modified 

 by adding a small transistor oscillator-amplifier. V/hen the point of the 

 gage came in contact with the water surface, the circuit was completed, 

 activating the oscillator and generating a tone which drove a speaker. A 

 precise elevation of the carriage at each station could be determined with 

 this device. The carriage elevation at station 23.5 (position of the 

 original toe of the slope) in the center of each tank was used as the ref- 

 erence datum. Fluctuations in carriage elevation of more than 0.005 foot 

 (0.15 centimeter) were corrected in the data reduction procedure. The 

 gages were adjusted so that elevations were measured with the bottom of 

 the tank at station 23.5 as the datum, and 2.33 feet was later subtracted 

 from each gage reading during the automatic data reduction to convert to 

 the SWL datum. 



d. Survey Accuracy . Ripple formations on the sand surface can affect 

 the accuracy of the local average sand surface elevation by half the ripple 

 height. A special survey was made in 1970 to give an indication of the 

 effect of ripples on survey accuracy. Using a point gage, distances and 

 elevations were measured at the crest and trough of all ripples along range 

 9 in the 10-foot tank after 42 hours. The surveyor visually determined 

 that the point was at the ripple crest or trough. Table 4 presents data 



on the frequency of occurrence of ripple heights and indicates the poten- 

 tial effect on survey accuracy. The data indicate that if all elevations 

 were measured at the worst points, i.e., always at a crest or trough, 82 

 percent of the measurements would be within 0.02 foot (0.61 centimeter) of 

 the actual local average elevation and 99 percent would be within 0.03 

 foot (0.91 centimeter). 



To achieve a more realistic estimate of the effect of ripples on 

 regular-spaced surveys, the actual error caused by ripples was determined 

 by comparing the elevations from the regular survey with those from the 

 special ripple survey. These data (Table 5) indicate that 64 percent of 



39 



