Predication of coastal erosion by using radar data

Waves are the main factor inducing coastal erosion. The main objectives of this study are to investigate coastal erosion in a short period. Polarised TOPSAR and ERS-1 data are used for this purpose. ERS-1 data was taken in 8 August 1993 and polarised TOPSAR data was taken in 6 December 1996. The wave spectra refraction patterns are modeled from TOPSAR and ERS-1 data. The simulations of quasi-linear model and radon transform were used to model the shoreline change. The rate of shoreline change was modeled with a different date. The vectors of shoreline were used to find also the rate of shoreline change. The different methods used to detect shoreline change were compared by using a statistical model. The results show that the wave spectra extracted from ERS-1 ranged from 25 m to 200 m and moved from the southwest direction. The wave spectra extracted from polarised TOPSAR data show that a wave spectra wavelength is shorter compare to ERS-1 wave data. The wavelength modeled from polarised TOPSAR data ranged from 20 m to 170 m and moved from northeast direction. The different wave spectra patterns induced a different pattern of wave refraction. The rate of erosion modeled from ERS-1 data is less than 1 m/y. The rate of erosion modeled from TOPSAR data is 1.4 m/y. The statistical model shows that radon transfer model has a good correlation with wave spectra model and vectors of shoreline rate change. The coastal erosion occurred due to the change of wave spectra refraction pattern. The radon transform could be used as new method to detect a shoreline change from remotely sensed data