Assessment of coastal Strandveld integrity using WorldView-2 imagery in False Bay, South Africa

Abstract

The coastal zone as interface between land and sea faces much pressure from human activities. These pressures make it difficult for the coastal zones to fulfil their natural functions, so threatening the state of coastal environments and making them vulnerable to coastal disasters and degradation. This study aimed to test whether high resolution remote sensing imagery can be used to assess the integrity of coastal Strandveld vegetation at the high spatial resolution required as baseline information for local environmental management. The study focused on the Cape Flats Dune Strandveld vegetation in northern False Bay near Cape Town, South Africa. The approach modifies and adopts a method developed by Lück-Vogel et al. (2013) for a remote sensing derived habitat intactness assessment for the Sandveld region at South Africa¿s west coast. A regression analysis was performed on the eight spectral bands of two WorldView-2 images and five image derivatives to evaluate the most suitable bands for producing an ecosystem integrity index. Based on the results, only the RED and the NIR bands were required to perform a decision tree classification for the integrity classification, which significantly simplifies the application and opens the way for transfer to sensors with similar spectral bands. The overall accuracy of the results was 80.5% with a kappa value of 0.75. This means that this approach provides results accurate enough for coastal management and conservation applications at a local scale. A further finding is the importance of seasonality to delineate natural and alien vegetation accurately.