100 

 90- 



80 



o 70 



H 

 < 

 > 60 



LU 



§50 



U- 



°40 



UJ 



^30 



LU 

 Q. 



20 

 10 







CENTRAL 70 PER CENT OF DATA 

 ICHANGEI LESS THAN 25 FEET 

 ISLOPEI LESS THAN 0°28' 



2.7 FEET 



1.4 FEET 



CENTRAL 50 PER CENT OF DATA 

 ICHANGEI LESS THAN 1.35 FEET 

 SLOPE! LESS THAN 0° 15' 



-20 



-10 +10 



DEPTH CHANGE (FEET) 



+ 20 



+ 30 



Figure 4. Example of cumulaiive percentage distribution of differences in depth 

 between haif-minute or 304-foot spaced readings of 20°C isotherm on section H. 

 The 25 and 75 percentile delineating the central 50% of Data and the 75 and 85 

 percentile delineating the central 70% of Data. 



To summarize the variability of depth changes (Appendix A), 

 points were scaled from each of the cumulative percentage distri- 

 butions of changes in depth per 304 feet, which corresponded to the 

 25th and 75th percentile of depth change and the absolute values 

 averaged. Similarly the 15th and 85th percentiles were also de- 

 termined and averaged. The latter average is somewhat analogous 

 to the "significant wave" method whereby the upper 30 per cent is 

 "average. " However, here the value represents a depth change 

 greater in absolute value than 70 per cent of the observations. 



As an alternate treatment of the data, a cumulative percent- 

 age distribution curve for the absolu-te values of depth differ- 

 ences could be plotted. The new 70th percentile change would 

 agree almost exactly with the average of the absolute values of the 

 15th and 85th percentiles discussed above. Hence, that average 

 wUl be designated 70th percentile of absolute value of depth change 

 (70th percentile - depth change). Likewise the 50th percentile in the 

 alternate treatment would agree with the average of the absolute 

 values of the 25th and 75th percentiles discussed in the previous 

 paragraph, and that average will be designated 50th percentile - 

 depth change. 



14 



