A QUANTITATIVE COMPARISON OF HORIZONTAL UNCERTAINTY IN ORTHOIMAGE- AND LIDAR-ESTABLISHED ESTUARINE SHORELINES IN NORTHEAST FLORIDA

Rising sea levels have become a threat to coastlines globally. As a result, there is an increased need to understand what future coastline elevations will be to better determine at-risk areas. While the most accurate measurements of shorelines today are from high-resolution techniques (e.g., LIDAR), historic quantitative data can still be used to assess rates of coastal change over time with a defined level of certainty. With the longevity of aerial photographs over the only relatively recent use of LIDAR, historical data can provide greater breadth to determining rates of shoreline migration despite lack of hard quantitative elevation data. This study specifically focuses on reconciling visual shoreline indicators of elevation from aerial photographs with more accurate LIDAR techniques. Our study area is the Guana Tolomato Matanzas National Estuarine Research Reserve in northeast Florida that contains a range of estuarine shoreline types, including salt marsh, mangrove, mudflats, and sandy beaches. We compare 0.25-m pixel resolution false color infrared digital orthorectified quarter quad imagery from February 2009 with 1-m pixel resolution LIDAR data collected in February 2008. The mean high water (MHW) NAVD88 shoreline was extracted from LIDAR data and compared with visual MHW proxies, which are assumed to be the edge of live marsh and mangrove vegetation or the wet-dry line on sandy shorelines. Using AMBUR (Analyzing Moving Boundaries Using R) program, statistical uncertainties are determined for these various categories of shorelines. Generally, we find that the shoreline positions are the same within uncertainty of the image orthorectification (5 m). However, within this uncertainty buffer, 1-3 meter offsets towards interiors of marshes and mangroves of the LIDAR shoreline are common, with the degree of offset increasing in areas outside of active boat traffic. This may be due to variations in sediment type or wave energy that influence subtle topography. These results suggest that edge of vegetation can serve as a MHW proxy in this region within the uncertainty of orthoimage rectification, which ranges here from 5 meters in 2009 imagery to 7-15 meters in older data sets.