Resolution and scale controls on the accuracy of atoll island shorelines interpreted from satellite imagery

Abstract

Understanding the positional uncertainty of vector objects within a GIS has been widely discussed and a range of approaches have been applied to quantify the uncertainty of such objects. Recently, a number of studies have adopted remote sensing approaches to examine the erosion of low-lying atoll islands. These studies are underpinned by a comparison of shoreline positions interpreted from remote imagery and represented as line features, and/or island areas represented as polygons. However, the effects of image resolution and scale of interpretation have rarely been considered when examining the results of such studies. In order to examine the control of image resolution and scale of interpretation on the accuracy of shoreline positions, repeated digitisation of islands was undertaken under a range of conditions similar to those applied in an operational setting. High-resolution (40 cm) satellite imagery was resampled to coarser resolutions and islands repeatedly digitised at a range of scales. Results quantify seemingly intuitive relationships, whereby shorelines digitised from coarse imagery and at small scales are less accurate than those digitised from high-resolution imagery at large scales. Results of this study provide a useful guide for selecting imagery for shoreline change studies and understanding the positional uncertainty of vector objects within a GIS.