Challenges in Estimating Long-Term Insar-Derived Land Subsidence in The Mekong Delta, Vietnam

The Mekong Delta in Vietnam is a low-lying, densely populated, and agriculturally important region, making it highly vulnerable to subsidence and sea-level rise. Recent years have seen accelerated subsidence, reaching several centimeters per year, primarily due to intensive groundwater pumping. Previous large-scale InSAR studies have been limited to short timeframes, applied linear assumptions and simplified LOS-to-vertical projections, potentially misrepresenting actual subsidence dynamics. In this study, we assessed the feasibility of long-term (2015-2023) subsidence monitoring using Sentinel-1 InSAR, processing a total of 708 descending-mode images with Persistent Scatterer (PS) InSAR and Small Baseline Subset (SBAS) InSAR. Unlike earlier work, our analysis explicitly addresses the challenges of processing large datasets in a coherence-limited delta environment and provides the first systematic comparison of PS and SBAS performance over an eight-year period. We show that while PS InSAR identified ~564,000 points, primarily in stable urban areas, it largely failed in non-urbanised regions. In contrast, SBAS InSAR yielded ~1,200,000 coherent points, with superior sensitivity to spatially variable subsidence across the delta. By linking InSAR point densities to land cover classes, we further demonstrate that SBAS is particularly effective in vegetated and agricultural areas, whereas PS points remain concentrated in built-up zones. These findings demonstrate that SBAS InSAR is more robust for delta-wide, long-term monitoring. Given the absence of in-situ validation over the studied period (permanent GNSS stations, levelling benchmarks, extensometers), InSAR currently remains the only comprehensive method for tracking deformation across the Mekong Delta. Our findings establish both the feasibility and the methodological requirements for reliable long-term InSAR-based monitoring in subsiding deltas, and highlight the need for future integration with independent ground-based measurements.