FLORIDA GEOLOGICAL SURVEY
requires both a temporal analytical magnitude that we must keep the number of component and a spatial analytical computational steps to a minimum in order component. to minimize the propagation of error in computing (bear in mind that in addition to Of the numerical methods available to the tempral analytical component a spatial analyze such data, many can actually component remains, which further increases magnify the uncertainty and/or error analytical computation). associated with the final results of an involved computational approach. Caution The "bottom line" is that we need to with respect to this aspect of analysis use the most appropriate and cannot be over emphasized. In fact, the computationally simple analytical topic is so important that a series of methodology available. The most standard equations for assessing the appropriate statistical analytical tool is propagation of error in computing have been undoubtedly end analysis which already provided in the Appendix. includes measures of determining the associated error or variability. In addition, The nature of historical shoreline what we might learn and quantify about location data is such that there is associated nature's own systematic variability can be error and variability. Surveying error used to our advantage both in terms of includes inherent closure errors, error due to assessing the acceptability of data, and as older technologies, and non-adjustment error an analytical tool. Such is the usefulness of for more recent vertical and horizontal epoch horizontal seasonal shoreline change. readjustments. Survey nets established for county surveys may not precisely relate to An example of temporal analysis is adjacent county nets as they would in a illustrated in Figure 12 for a locality about state-wide net. Long-term sea level 2.7 kilometers south of a major inlet on the changes, though slight, affect long-term east coast of Florida. Equation (1) was shoreline changes. These sources of error evaluated using the appropriate wave data of may be called map-source errors
after Demirpolat and others
(1989), for which a magnitude of 400 -a: -6.25 m/yr c: -0.4S m/yr  9 to 15 m may be appropriate b: +110 m/yr d: +1.64 m/yr (Demirpolat and others, 1989). Artificial Nourishment Interpretive plotting of errors of 3
shoreline location (depending on 3
data concentration) on original Jetty ConstructionBegansurvey maps must be assumed, v J especially for older maps. 200 Inlet Artificially Cut Present digitizing technology ( b \ results in an error of  3 to 4 m (Demirpolat and others, 1989). 100 ..--Except for recent -'technologies, magnitudes of errors for examples suggested 1900 19s50 2000 above are not known with Year certainty in the majority of cases. Figure 12. Example of long-term shoreline change rate Even so, it can be envisioned that (solid lines) temporal analysis using seasonal shoreline they are of sufficiently large shift data (dashed lines); see text for explanation.
18