Characterizing global satellite-based indicators for coastal vulnerability to erosion management as exemplified by a regional level analysis from Northeast Brazil.

The detection of coastal vulnerability to erosion is crucial for decision-making regarding the economy, ecology, health, security, among other issues. Most of the studies gather a large data set about physical and anthropogenic interference's on the vulnerability of coastal erosion regions around the world. However, for developing nations like Brazil, with extensive shoreline, it is challenging to develop and maintain an in situ infrastructure to offer a systematical scientific data set. In this context, several methods like Coastal Vulnerability Index (CVI) for monitoring the dynamic behavior of coastal systems require in situ collected data. Therefore, this contribution explores the use of global open source satellite-based indicators to assess coastal vulnerability to erosion at a regional level adopting an uncorrelated orthogonal basis set of Principal Component Analysis (PCA). For this, the data set covers many spheres of the environment like biophysical and social factors, adopting the Pernambuco State's coast, Brazil, as a case study. The results showed the direct relationship between a high level of urbanization and low vegetation with the high coastal vulnerability to erosion. PC1 revealed built-up and surface temperature vary inversely to the soil organic carbon and vegetation cover along about 20 km (≈10% of the shoreline extension). The hotspots were in the urban cluster (Paulista, Olinda, Recife, and Jaboatao dos Guararapes), combined with high shoreline change around -2 m/yr. PC2 showed the natural action of wind on wave heights combined with sediment removal and the backshore settlement along 10 km of extension (≈5.5% of the shoreline), with the highly vulnerable sites concentrated in Itamaraca Island and C. S. Agostinho. This approach benefits from the multi-satellite and multi-resolution data sets integration to unravel the statistical influence of each variable able to guide stakeholders.