r 25 



"5; 5 



^ 4 







" ^ K " Sample Size {meosuremenf points per 1.6 Km) 



Figure A-2. Comparison of maximum difference between Xg, the mean 

 bluff recession rate of a sample of n equally spaced 

 measurements per 1.6 kilometers, and x^y, the mean 

 computed for all 57 stations. Lines are shown for 1- 

 and 4-year periods (curves are hand-fit). 



From Figure A-2, it appears that for data from 1-year periods, 20 measure- 

 ment stations at equally spaced increments per 1.6 kilometers would result in 

 a mean recession rate within ±0.3 meter per year of the value which would be 

 determined using 57 stations. This amounts to a 65-percent reduction in the 

 number of measurements needed. 



Since Figure A-2 is based on the maximum difference between 

 (instead of the mean difference), it should be conservative. 



and 



^5 7 



Fewer stations are needed for longer period data as indicated by the shift 

 of the line for the 4-year data to only 10 stations required for the same accu- 

 racy. If an accuracy of ±0.3 meter per year is inadequate. Figure A-2 can also 

 be used to select a larger san^le size. 



This rather simplified procedure for estimating measurement station fre- 

 quency is presented for general guidance in setting up a similar study. The 

 actual number of stations needed will vary depending on the quality and scale 

 of the air photos, the time period considered, the accuracy desired, the uni- 

 formity of bluff type, and the analysis procedure. It is suggested that some 

 experimenting be done with measurement density along a short reach of coast 

 before a final station frequency is selected. 



74 



