Flood Variability Determines the Location of Lobe‐Scale Avulsions on Deltas: Madagascar

Abstract

River deltas grow through repeated lobe-scale avulsions, which often occur at a location that correlates with the backwater lengthscale. Competing hypotheses attribute the avulsion node origin to either the morphodynamic feedbacks caused by natural flood discharge variability (backwater hypothesis) or to the prograding delta lobe geometry (geometric hypothesis). Here, using theory, historical flood records, and remotely sensed elevation data, we analyzed five lobe-scale delta avulsions in Madagascar, captured by Landsat imagery. Avulsion lengths were 5-55 km, distances significantly longer than the backwater lengthscale and inconsistent with the geometric hypothesis. We show that the steep, silt-bedded rivers of Madagascar have flood-induced bed scour, driven by backwater hydrodynamics, that propagates farther upstream than the backwater lengthscale. The avulsion lengths are 3.1 ± 1.5 times the predicted flood scour lengths, similar to low-gradient deltas, and consistent with backwater hypothesis. Results demonstrate that erosion initiated by nonuniform flows in the backwater zone is a primary control on delta avulsion locations. Plain Language Summary River deltas grow through abrupt channel shifts, called avulsions, which pose a threat to life and property, but we do not understand why avulsions occur where they do. One hypothesis is that flood discharge variability creates a preferential zone of sediment accumulation within the so-called backwater zone of coastal rivers, which becomes the locus of avulsion. The rationale is that the flow acceleration and deceleration within the backwater zone during different sized floods creates nonoverlapping patterns of erosion and deposition resulting in a peak in sediment accumulation. If this hypothesis is correct, then avulsion sites should be farther upstream, relative to the backwater lengthscale, in steep rivers with fine-grained sediment beds that have more significant bed scour during large floods. By analyzing historical avulsions on Madagascar, we found that avulsions occurred far upstream due to the large floods and fine-grained sediment, in support of the flood discharge variability hypothesis. Our work highlights that changes in flood regime and sediment grain size of coastal rivers due to land use and climate changes can influence avulsion location.

Climate‐Change Controls on River Delta Avulsion Location and Frequency

“…In contrast to fan deltas, lowland deltas commonly feature avulsion nodes on unconfined plains without a topographic slope break (Brooke et al., 2020; Ganti et al., 2014). Avulsion nodes on lowland deltas have been documented to shift with movement of the shoreline (Ganti, Chadwick, Hassenruck‐Gudipati, Fuller, & Lamb, 2016; Ganti et al., 2014) to maintain a constant avulsion length that scales with the backwater length‐scale (Figure 1e) (Chatanantavet et al., 2012; Jerolmack & Swenson, 2007), where is the ratio of bankfull river channel depth, , to channel‐bed slope (Lamb et al., 2012; Paola & Mohrig, 1996).…”

Where rivers jump course

“…In accordance with emerging theory ( 30 ), we hypothesized that the longitudinal extent of flood-driven scours in these rivers is more pronounced than in backwater-scaled deltas (fig. S4), which diminishes sedimentation within the backwater reach and thus causes LA≫Lbtrue¯ ( 31 ). To test this hypothesis, we estimated the dimensionless flood duration, defined as Te*=tscour/tadj, where t scour is the typical bankfull-overtopping flood duration and t adj is a bed-adjustment time scale ( 11 ), globally by simulated monthly water and suspended sediment discharges from 1980 to 2020 C.E.…”

“…S6). For rivers with Te*>1, the longitudinal extent of flood-driven erosion is expected to extend upstream of the backwater zone to a distance approximated by Lbtrue¯Te* ( 31 , 33 ) (fig. S4).…”

Effect of Sea‐Level Change on River Avulsions and Stratigraphy for an Experimental Lowland Delta

“…The distribution of avulsion lengths from CROF18 is approximately constant and scales with the backwater length during sea‐level rise despite major changes in shoreline position. Avulsion lengths reflect a preferential avulsion node within the backwater zone ($${L}_{A}\le {L}_{b,Bresse}=1.8$$ m) that is upstream of the zone of prominent scour during high flows $$({L}_{A}\ge {L}_{d,Bresse}=0.9\,\mathrm{m})$$, consistent with theory (Bresse, 1860; Brooke et al., 2020; Lamb et al., 2012). Dimensionless avulsion length values cluster in the range of $$0.5< {L}_{A}^{\ast }< 1$$ similar to field data and the CROF16 experiment (Figure 10a).…”

“…Previously, backwater hydrodynamics and backfilling—the latter termed “morphodynamic backwater” by Hoyal and Sheets (2009)—have been viewed as competing hypotheses to explain how avulsions occur (cf. Brooke et al., 2020; Ratliff et al., 2021; Zheng et al., 2019). On the contrary, we find that these ideas are not in competition.…”