Mangrove mapping using Sentinel-1 data for improved decision support on sustainable conservation and restoration interventions in the Keta Lagoon Complex Ramsar Site, Ghana

Abstract

Despite the conservation importance of the Keta Lagoon Complex Ramsar Site (KLCRS), obtaining information on the extent and distribution of mangroves is challenging due to the unavailability of optical satellite data. This research explored Sentinel-1 radar data to provide information on mangrove distribution in the KLCRS. Global positioning system points from 443 training and 196 validation sites were used. In addition, focus group discussions and key informant interviews were used to corroborate information on mangrove distribution. Sentinel-1 data were processed for backscatter coefficients and two backscatter derivatives. These were stacked into a four-layer image composite and classified using a support vector machine. An overall classification accuracy of 89.28% was obtained. In addition, user and producer accuracies of 100 and 97.3% respectively were obtained for the mangrove class. The results show that mangroves occupy a total area of 41.02 km2 in the KLCRS and are mostly found around the Salo, Bomigo, Anyanui and Dzita communities. This study demonstrates the possibility of using Sentinel-1 imagery to map mangroves within the KLCRS. Thus, this study serves as a guideline for other data-constrained mangrove landscapes to map and monitor mangroves for conservation and restoration actions.