Impacts of Climate Change and Human Activities on Small Island Nations: A Data-Driven Approach to Disentangling Coastal Changes

Small Island Developing States (SIDS) are a group of 58 nations identified by the United Nations as facing unique sustainability challenges, including high exposure to climate change, lack of data, and limited resources. The effects of climate change are already observed in SIDS, notably an increase in the magnitude and frequency of natural disasters, marine biodiversity loss, ocean acidification, coral bleaching, sea-level rise, and coastal erosion. The coastal zone is considered to be the main economic, environmental, and cultural resource of SIDS. Monitoring coastal changes is therefore essential to protect communities, biodiversity, natural landscapes, and their economies, as well as to help them adapt to and mitigate against climate change.

This talk presents the use of remote sensing data to monitor and analyse the evolution of small islands. Open-access satellite missions, namely Landsat (NASA) and Sentinel (ESA), provide imagery with spatial resolutions of 10 to 60 metres and temporal resolutions of 5 to 16 days. These capabilities enable the retrieval of high-temporal-resolution time series of coastline positions across islands worldwide.

A specific focus is placed on the Maldives (Indian Ocean) due to its low elevation and extensive human interventions. Existing literature lacks a comprehensive understanding of the patterns of coastal changes, as well as the main anthropogenic and environmental drivers involved, which operate across diverse temporal (e.g., daily, seasonal, multi-decadal) and spatial scales (e.g., site-specific or atoll-wide). Maldivian coastlines are not systematically or frequently monitored, such that sub and interannual variability and the geomorphological responses to climate forcings, such as the Indian Monsoon and the Indian Ocean Dipole, are not understood.

To address this research gap, a data-driven framework was developed, leveraging remote sensing, in situ measurements, and open-access databases. This framework quantifies and disentangles coastal changes through three steps: firstly, an image segmentation algorithm that exploits characteristic spectral features was developed to extract the shape of small islands over time, providing reliable monthly time series of coastline positions. Secondly, time series decomposition was conducted, separating time series into trend, seasonality, and residuals. Each component was analysed separately using different methods: trend analysis to investigate the impacts of human activities (e.g., land reclamation, sand mining, shoreline armouring) and climate change (e.g., coral growth, sea-level rise) on natural coastal responses; seasonality analysis to explore sub- and inter-annual drivers, including the Indian Monsoon and the Indian Ocean Dipole; and residual analysis to quantify lagged effects of met-ocean conditions (e.g., waves, sea level) through causal inference methods. Thirdly, results are generated for several hundreds of islands, identifying regional patterns across atolls and illustrating how met-ocean conditions influence coastlines on a larger scale.

Overall, this novel approach integrates advanced remote sensing and data science techniques, providing an unprecedented analysis of coastal dynamics at both fine and large scales, supporting better-informed environmental policies, urban planning, and marine infrastructure development to ensure sustainable management and resilience of these ecosystems.