BIHAMTA TOOSI N., SAFIANIAN A., FAKHERAN S., POURMANAFI S., GINZLER C., T. WASER L., Remote Sensing, Vol. 12, PP. 2684, 2020.

Abstract:

Mangrove forests grow in the inter-tidal areas along coastlines, rivers, and tidal lands.They are highly productive ecosystems and provide numerous ecological and economic goodsand services for humans. In order to develop programs for applying guided conservation andenhancing ecosystem management, accurate and regularly updated maps on their distribution, extent,and species composition are needed. Recent advances in remote sensing techniques have made itpossible to gather the required information about mangrove ecosystems. Since costs are a limitingfactor in generating land cover maps, the latest remote sensing techniques are advantageous. In thisstudy, we investigated the potential of combining Sentinel-2 and Worldview-2 data to classify eightland cover classes in a mangrove ecosystem in Iran with an area of 768 km2. The upscaling approachcomprises (i) extraction of reflectance values from Worldview-2 images, (ii) segmentation based onspectral and spatial features, and (iii) wall-to-wall prediction of the land cover based on Sentinel-2images. We used an upscaling approach to minimize the costs of commercial satellite images forcollecting reference data and to focus on freely available satellite data for mapping land cover classesof mangrove ecosystems. The approach resulted in a 65.5% overall accuracy and a kappa coefficientof 0.63, and it produced the highest accuracies for deep water and closed mangrove canopy cover.Mapping accuracies improved with this approach, resulting in medium overall accuracy even thoughthe user ’s accuracy of some classes, such as tidal zone and shallow water, was low. Conservation andsustainable management in these ecosystems can be improved in the future.

Keywords:ecosystem; mangrove; random forest; Sentinel-2; upscaling; Worldview-2