Coastline Changes Extracted From SAR Images

Below is an example result obtained by the University of Miami's remote sensing team in the hours after the landfall of Hurricane Ian.

The animated GIFs show the COSMO-SkyMed SAR image from September 29, 11:22 UTC (07:22 local time) with extracted coastlines from an algorithm developed at CSTARS (red lines). Most of the coastlines follow the strongest contrast between land and water very closely, where the human eye would expect them to be. If we remove the SAR image and superimpose just the red lines on the Google Earth background, they do not align well with the locations and shapes of some islands and sandbanks in the optical image, indicating significant changes. This shows that SAR images can be used to detect and quantify coastline changes very soon after a hurricane landfall. However, the Google Earth image is from January 2021, so some of the detected changes may have occurred in the months before the hurricane landfall.

Ideally, we should have acquired two SAR images right before and after the landfall, but changing hurricane path predictions and limited SAR imaging capacities can make it difficult to achieve a perfect coverage of the actual hurricane landfall area in space and time.

‍