*4.1. Subsidence in Deltas*

Dating of modern deltas reveals that many major deltas worldwide were formed during the Holocene between 8,500 and 6,500 years before the present. The Fraser (Canada), Mississippi (United States), and Nile (Egypt) River deltas, to cite a few, were all formed within that period [59]. Accordingly, it has been suggested that following the initial rapid SLR during deglaciation, Holocene deltaic sequences began to accumulate as the rate of fluvial sediment input exceeded the declining rate of SLR in continental margins [59]. Compaction of the strata deposited during the Holocene has been described as a major driver of natural land subsidence in these dynamic environments [11,60]. In the case of the Magdalena River, we argue that downward trends in vertical land motion (VLM) reflect ongoing subsidence associated with modern fluvio-lacustrine sediments [26,61] that overlie relict sediments from a former delta that migrated westward after the mid-Holocene [21]. Low magnitude seismological activity associated with the SMBF indicates that VLM estimates reported in this work (Figure 5), at least during the period of observation (2007– 2021), are not earthquake-driven, which suggests that one of the primary drivers of this process is sediment compaction of organic and mud deposits.

VLM in the study site is closely related to the Holocene history of the Magdalena delta and the supply of sediment from the drainage basin. Natural drainage displacements and levee constructions since the 1920s [22] have progressively diminished the amount of sediment delivered by the Magdalena River to the study area [22], disrupting the balance between sedimentation and natural compaction of sediments. More recently, the construction of the highway between Barranquilla and Santa Marta in the 1950s hampered the interchange of water between lagoons and ocean [23], indirectly causing subsidence by disturbing the mangrove forest.

Compared to previous assessments of RSL change for the region, which indicate current rising sea levels of approximately 0.5 cm/yr [32], local subsidence values of up to −1.5 cm/yr highlight the relevance of incorporating subsidence in the prediction of SLR and coastline positions. Earlier work aiming to correlate coastline erosion to sealevel rise, as measured from historical tide gauges, revealed that the average coastline change is two orders of magnitude greater than the rate of the sea-level rise [62]. Thus, by increasing the rate of local RSL, subsidence might increase future flooding risk associated with storms [55,63], resulting in the possible drowning of the deltaic barrier island.
