Time series monitoring and analysis of Pakistan’s mangrove using Sentinel-2 data

Abstract

Sustainable Development Goal-14 (SDG-14) directly demands the protection, conservation, restoration, and sustainable management of the global mangrove ecosystem. Assessment of the development Pakistan has made toward the sustainable management of mangrove ecosystems necessitates the remote sensing-based evaluation of national-level mangrove cover. Using Google Earth Engine (GEE) for geoprocessing 12,000+ 10-m high-spatial resolution Sentinel-2 time-series images (2016–23) and applying random forest (RF) classifier, the current research provides the latest spatial distribution of mangroves along Pakistan’s coastline and changes observed for a duration of 8 years. Additionally, this research provides the first spatiotemporal health assessment of Pakistan’s national mangrove cover as well. Rational analysis of the results indicated splitting the entire timeline based on two seasons (Jan–Jun and Jul–Dec). Results revealed an overall increase of 1,210 km2 (2023) in the mangrove cover on the national-level; a 3.42 km2 average annual increase from 2016 (1,186 km2). Mangrove gain/loss assessment based on the land use land cover (LULC) transition matrix illustrated 223 km2 gain and 199 km2 loss; a 24 km2 net gain. Declines of 20.28% and 7.91% were found in maximum- and mean-NDVI (2016-23), depicting the deteriorating mangrove health conditions. Likewise, significant Sen’s slope analysis (p < 0.05) indicated that 88.8% of all the mangrove-NDVI pixels exhibited an overall decrease, whereas 11.2% pixels showed an overall increase (2016-23). It was concluded that despite showing a growth in the extent, mangrove cover in Pakistan has shown a decline in health, primarily due to deforestation for urban operations and sea-level rise, still making them vulnerable and potentially leading to a disrupted ecosystem, including carbon release in the atmosphere. This study will assist in the formulation of mangrove conservation and management strategies, whereas future research can explore the potentials of land surface temperature (LST) and evapotranspiration in combination to the NDVI for an in-depth analysis of the health status of mangroves.