Contribution of campaign GNSS toward parsing subsidence rates by time and depth in coastal Bangladesh

Abstract

Coastal regions are vulnerable to rising seas, increasing storm magnitude, and decimation of ecologically-fragile areas. Deltas are particularly sensitive to the balance between sea-level rise, land subsidence and sedimentation that determine relative elevation. Bangladesh has been highlighted as being at risk from sea-level rise. Integrating measurements from different methods can approach a more complete understanding of factors controlling areally and temporally varying subsidence rates. To augment our compilation of rates from stratigraphic wells, historic buildings, vertical strainmeters, RSET-MH, and continuous Global Navigation Satellite System, we resurveyed 48 geodetic monuments in coastal Bangladesh ∼18 years after the monuments were installed. A later resurvey of 4 sites showed that some sites with higher subsidence may be unstable, but we consider the subsidence pattern of all the sites. Sites with rates <2 mm/yr overlie thin (≤35 m), sandy Holocene deposits located along interfluves between the main paleo-river valleys. As Holocene strata thicken seaward and become muddier, subsidence rates increase to 20–25 mm/y. Sites in incised valleys of the Ganges, Brahmaputra and Meghna Rivers, with Holocene sediments >100 m show subsidence rates of 20 ± 10 mm/y, with a slight seaward increase. Overall, subsidence rates increase with Holocene sediment thickness and the seaward shift from sandy to muddy sediments. Together with earlier measurements, we parse the different rates and mechanisms of subsidence. Earlier models show 2–3 mm/yr correspond to deep processes, such as isostasy. Within the shallow Holocene (<10 m), we estimate 5–8 mm/yr of subsidence from shallow, edaphic effects (tree roots, burrows, organic matter decomposition) and shallow (≤10 m) sediment consolidation on short timescales. Below this, we estimate 3–6 mm/yr from compaction of the upper Holocene strata, with 2–5 mm/yr occurring in deeper Holocene strata. Subsidence rates in areas of active sedimentation, such as rice fields and mangrove forests, are greater than buildings and structures with deep foundations. Subsidence on timescales >300 y, which do not include edaphic effects, are up to ∼5 mm/y. We note subsidence can be offset by active deltaic sedimentation, and does not necessarily indicate elevation loss. Collectively, the integration of these approaches allows us to begin quantifying the varied contributions to land subsidence from edaphic effects, Holocene sediment compaction, lithology, and time. Similar factors may contribute to the highly variable subsidence rates observed at other deltas worldwide.