A cloud computing-based approach to mapping mangrove erosion and progradation: Case studies from the Sundarbans and French Guiana. (5th January 2021)

• Record Type: Journal Article
• Authors: Bhargava, Radhika; Sarkar, Dipto; Friess, Daniel A.
• Is Part Of: Estuarine, coastal and shelf science. Volume 248(2021)
• Publication Date: 2021-01-05
• Subjects: mangrove loss -- coastal erosion -- coastal accretion -- Google Earth Engine -- water classification; Estuarine oceanography -- Periodicals; Coasts -- Periodicals; Estuarine biology -- Periodicals; Seashore biology -- Periodicals; Coasts; Estuarine biology; Estuarine oceanography; Seashore biology; Periodicals; 551.461805
• DOI: 10.1016/j.ecss.2020.106798
• Languages: English
• ISSNs: 0272-7714

Abstract

Abstract: Erosion and progradation are natural dynamics that take place across mangrove landscapes, but anthropogenic pressure may disturb these dynamics and cause increased erosion. The increased erosion of mangrove coastlines leads to mangrove degradation and loss of ecosystem services. However, to map the extent of erosion and progradation in dynamic landscapes is often time and data consuming and challenging to perform at a large scale. This study uses cloud computing-based remote sensing analysis in Google Earth Engine to map the extent of mangrove shoreline dynamics in two contrasting hotspots: the Sundarbans mangrove complex across India and Bangladesh, and the coast of French Guiana. Mapping was performed between 1984 and 2018 using available data on mangrove distribution and surface water change, classifying change of states as permanent, seasonal and ephemeral. Between 1984 and 2018, erosion and progradation accounted for 24.55% and 12.52% of total change, respectively. Contrastingly, in French Guiana erosion and progradation accounted for 9.53% and 4.48% of total change, respectively. The Sundarbans experienced more permanent changes as compared to French Guiana. The permanent loss of mangrove forests in the Sundarbans can be attributed to the reduction in sediment supply and the temporary loss in French Guiana to cyclical patterns of migrating mudbanks. The methods presented in this study performed with an overall accuracy above 90% in different settings of both the case studies. This study provides an important baseline of mangrove erosion and progradation extent and an improved method to analyse coastal dynamics in mangrove landscapes. Graphical abstract: Image 1 Highlights: Erosion is an important cause of mangrove loss in the Sundarbans and French Guiana. Cloud computing-based approach allows rapid and accurate quantification of shoreline dynamics. The Sundarbans experienced a higher permanent loss of mangrove forests compared to French Guiana. In both case studies, there were large amounts of temporary changes. Cloud computing-based approaches can provide accurate, regularly updated outputs for managers.