Assessing Changes and Quantifying Loss in African Mangrove Forests Using SAR/Optical Data Fusion
Abstract
Mangrove forests are found in inter-tidal zones of tropical regions around the world and provide important ecological and economic benefits - they are considered carbon sequesters, habitats for flora and fauna, and natural barriers to hurricanes and tsunamis. Wood from mangrove forests are used as fuel and building materials in surrounding coastal communities, therefore promoting local livelihoods. Despite the importance of these ecosystems, mangrove forests have historically been degraded in natural processes such as severe weather, and anthropogenic factors like conversion to agriculture and aquaculture.
This study assesses change in mangrove forests in Nigeria and Mozambique from 2015 to 2018 using SAR and optical data fusion. Due to frequent cloud cover over the study area, SAR and optical data is fused to obtain gap-free imagery without clouds. Landsat-8 OLI and Sentinel-1 imagery is fused with TensorFlow, an open source platform used in developing machine learning models. The resulting images are classified to discriminate mangrove forest cover from other land cover types, and change is estimated using image differencing. Understanding the rates and magnitude of mangrove change across space and time can aid in identifying priority areas for forest regeneration, and can help construct sustainable management practices for the future.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFMGC51F1130S
- Keywords:
-
- 1616 Climate variability;
- GLOBAL CHANGE;
- 1632 Land cover change;
- GLOBAL CHANGE;
- 1640 Remote sensing;
- GLOBAL CHANGE;
- 1655 Water cycles;
- GLOBAL CHANGE