A continuous monitoring approach for ecosystems based on time series analyses: A proposal with a case study for the mangroves of Marismas Nacionales, Mexico
Keywords: mangroves, Marismas Nacionales, time series, Sentinel-2, NDVI, Mann-Kendal test, trends
Abstract. The mangrove ecosystem stands out for the numerous environmental services it provides, in addition to being one of the most vulnerable ecosystems to climate change. Marismas Nacionales is located in northwestern Mexico and hosts one of the most extensive areas of continuous mangrove ecosystem along the Pacific North coast of Mexico. Although vital, this mangrove ecosystem faces multiple pressures and is shifting inland in response to climate change —making ongoing monitoring of its size and health essential. Aiming for monitoring the health of mangrove and disturbed mangrove areas, a time series of monthly NDVI composites derived from Sentinel-2 imagery (10 m spatial resolution) was analyzed for the period 2019–2024. At the pixel level, the Mann-Kendall test was applied to determine significant trends. An inspection of the Z-statistic was conducted to identify gradual and relevant changes; when possible, these changes were validated with field data and high-resolution imagery. The results revealed heterogeneity in the behavior of pixel time series, reflected in the values of the Z-statistic. This heterogeneity is due to the fact that mangroves are subject to different change factors depending on their spatial location. Among the key factors identified were hurricane-induced damage, land-use change, inland colonization, recovery driven by restoration efforts, post-hurricane vegetation rebound, and losses resulting from coastal erosion. Across the 80,959 hectares encompassed by the study area, 47.08% exhibited a significant negative trend, 20.19% a non-significant negative trend, 18.07% a non-significant positive trend, and 14.66% a significant positive trend. The analyzed data revealed the dynamic nature of the mangrove ecosystem in response to various change factors, and the proposed method could serve as a foundation for integration into the national products generated by Mexico’s Mangrove Monitoring System.
