Abstract
Complex changes in coastlines are increasing with climate, sea level, and human impacts. Remote Sensing (RS) and Geographic Information Systems (GIS) provide critical information to rapidly and precisely monitor environmental changes in coastal areas and to understand and respond to environmental, economic, and social impacts. This study was aimed at determining the temporal changes in the coastline of the Seyhan Basin, which is one of the basins significantly affected by climate change and drought in Turkiye. In this context, approximately 50 km of coastline was automatically extracted on the Google Earth Engine (GEE) platform using Landsat satellite images from 1985–2023. This coastline was divided into 3 different regions, and spatial analysis was performed with different statistical proportioning techniques (EPR, LRR, NSM, SCE, and WLR) according to years with the Digital Shoreline Analysis System (DSAS) tool. In addition, to determine whether there is a statistically significant difference between the results obtained from the different methods used to determine the coastal change, the Kruskal-Wallis H test and ANOVA test were applied by min-max normalization. The amounts of erosion and deposition found according to different methods vary by region. Statistical differences were found between the methods used, varying by region. In general, NSM and EPR methods provided similar results in determining coastal changes, while other methods differed by region. In the study, the Kalman filtering model was also used to predict the coastline for the years 2033 and 2043 and to identify areas that are vulnerable to erosion and deposition on the future coastline. Comparisons were made to determine the performance of Kalman filtering. In the 10-year and 20-year future forecasts for determining the coastline for the years 2033 and 2043 with the Kalman filtering model, it was determined that the excessive prediction time negatively affected the performance in determining the coastal boundary changes.