Impact of satellite derived wind and LULC on coastal inundation: a sensitivity study for the east coast of India
摘要
With the recent rise in cyclone intensity and frequency, it’s crucial to improve the modelling techniques to better predict coastal flooding during storm surge events. The study uses a stand-alone Advanced Circulation (ADCIRC) model to assess the impact of satellite-derived winds and Land use/land cover (LULC) to compute storm tides (ST) and associated coastal inundation. Simulations are carried out for four cyclones during 2018–2020 that made landfall over different locations along the east coast of India. Winds obtained from merged satellite product of Cross Calibrated Multi-Platform, generated winds using JTWC best-track information along with GFS winds are considered. It is observed that inundation from the JTWC best track is able to compute coastal inundation better than the satellite-derived winds. An improvement of up to 80% in the simulated inundated area is observed when the temporal frequency of cyclonic wind is increased from 6 h to 3 h, compared to simulations using 6-hourly winds. This signifies that high-temporal satellite winds of less than 3 h are highly desirable, especially during the cyclone landfall period, for realistic simulation of ST and inland inundation. Additionally, LULC obtained from MODIS, Sentinel-2 and Resourcesat-2 satellites are considered for finding the optimum LULC required for the better computation of coastal inundation. It is found that LULC from Resourcesat-2 performs much better in terms of wind speed with the lowest RMSE ranging from 2 to 2.4 m/s with respect to Automatic Surface Observation System measurements. Our analysis suggests that although the spatially averaged winds over the mangroves dominated area reduced by only 2–5%, the corresponding coastal inundation reduced by 14–20%, signifying the importance of wind-induced surface stress (modified by land cover) and bottom stress (due to increased friction from mangroves). The study highlights the importance of high-temporal-resolution winds and realistic LULC representation in ST modelling, providing useful information for enhancing coastal flood risk estimation and disaster preparedness planning.