point, the accuracy tends to decrease with distance, and large elevation 

 errors are possible (Czerniak, 1973). Each survey began near the bluff crest, 

 so errors should be greater for points on the beach than for those on the 

 bluff line. 



Surveying after September 1973 was done by transit and tape. Elevations 

 were recorded to the nearest tenth of a foot, and distances were measured to 

 the nearest foot. Because of the height of the bluffs on profile lines 2, 13, 

 and 14, stadia readings were used to determine both distance and elevation. 

 Vertical datum was changed from that used by Davis (1976) to low water datum 

 (LWD) equal to 175.81 meters IGLD. Each profile line was surveyed from the 

 bench mark, or from a point landward of the bench mark, to wading depth or to 

 the edge of the ice cover. 



Because of the severe erosion at some of the profile lines during the 

 study, bench marks were occasionally lost and had to be reset. Usually, this 

 required simply placing an auxiliary bench mark landward of the original one 

 before the loss actually occurred. However, in some instances, the original 

 monument was lost before the auxiliary monument was installed. When this 

 occurred, both horizontal and vertical control had to be reestablished. 



Vertical control was established from the lake level the day the auxiliary 

 bench mark was placed. Once the lake level was determined from a nearby gage, 

 the data were corrected to the same datum as the original bench mark. Hori- 

 zontal control was more difficult to establish. This was accomplished by 

 estimating the distance between the location of the original monument and the 

 auxiliary one. 



The accuracy of this procedure was questionable, particularly during the 

 Davis surveys when monuments were placed on the bluff face and the Emery 

 method was used to reestablish control. The problem was most acute between 

 the end of the Davis surveys in July 1973 and the beginning of the U.S. Army 

 Engineer District, Detroit, surveys in October 1973. Although original bench 

 marks were used where possible, the vertical datum was changed and new verti- 

 cal control was established. Similar problems also existed at some profile 

 lines between the two periods of the Davis surveys (August 1970 to July 1972 

 and August 1972 to July 1973). 



The importance of this surveying problem is that, because of the inaccura- 

 cies and the lack of a stable bench mark at some profile lines, there is not a 

 reliable continuous record of profile line changes. Also, the data, which are 

 available at CERQ, had to be stored in three separate files, one for each 

 study period. Consequently, changes to some of the profile features, such as 

 the shoreline, cannot be examined over the entire study period. Only monthly 

 amounts of bluff recession, which are less sensitive to vertical control 

 errors, are examined in detail over the entire period (August 1970 to December 

 1974). Annual volume changes are also discussed. Monthly volume and shore- 

 line changes, which can only be examined for October 1973 to December 1974, 

 are discussed in Appendix A. 



2. Bluff Recession and Volume Change Measurements . 



Monthly amounts of bluff recession for each profile line are tabulated (in 

 feet) and summarized in Table 3. Measurements from August 1970 to July 1973 

 are from Davis (1976). English units have been used because the data were 

 originally collected to an accuracy of the nearest foot. Figure 9 is a 

 histogram of the data in Table 3. 



21 



