ABSTRACT
Ford, M.R.; Ryan, E.J.; Mules, T.; Tuck, M.E.; Sengupta, M., and Dickson, M.E., 0000. An assessment of the accuracy of shoreline indicators interpreted from aerial photos and very-high-resolution satellite imagery.
Shoreline change analysis, underpinned by the interpretation of features from remote imagery, is widely employed in coastal science, engineering, and management. Understanding and quantifying the different sources of uncertainty in the position of shoreline features is essential to provide confidence in the descriptions and calculations of the modes and rates of coastal change. A series of practical tests was developed to better understand and quantify the sources of uncertainty in the position of shoreline indicators mapped as part of a national-scale assessment of coastal change in Aotearoa/New Zealand. First, the positional accuracy of aerial photographs was calculated using tests that were independent of the georeferencing process. The accuracy of the aerial photos examined ranged between 3.61 and 5.03 m (95% confidence), with results showing that the positional accuracy is a product of photo scale and that large-scale photographs have the greatest accuracy. Second, tests were conducted to optimise the zoom level used when digitising different image types to standardise approaches to mapping within a GIS. Third, the precision and accuracy of shorelines mapped by different operators was assessed and showed that despite statistically significant differences in the position of shorelines across operators, the differences were smaller than the pixel size of the imagery. Finally, results reveal that the ability to map a shoreline indicator precisely and repeatedly was a product of the shoreline feature and the clarity of the boundary between landcovers. Results indicate that the positional uncertainty of shoreline features mapped in Aotearoa/New Zealand varied from 0.44 to 10.08 m.
