Satellite observations of 16 meandering rivers across the Amazon Basin showed that sediment-rich rivers are dominated by downstream-rotating meanders, while the few upstream-rotating meanders — bends whose migration concentrates upstream of the point of maximum curvature — occur on rivers with low sediment loads, tentatively attributed to slow point-bar growth. This attribution, however, was never isolated from competing mechanisms (spatially variable bank-material erodibility, backwater effects following meander cutoffs, and valley orientation of tightly curved composite bends), and it rested on suspended-sediment flux as a proxy because no direct bed-material flux data exist. The open problem is to determine what physically initiates upstream-rotating meanders — in particular whether low coarse-bed-material supply and the consequent slow construction of point bars is the cause.

Background

A meandering river is a single-thread channel whose planform winds across its floodplain; its sinuosity is the ratio of channel length along the centerline to the straight-line valley distance. As a bend migrates laterally, sediment eroded from the outer bank is partly deposited on the inner bank as a point bar — an emergent shoal that deflects flow toward the outer bank, sustaining further erosion and bend growth. Classical theory and field work predict that the local migration rate peaks just downstream of the bend's point of maximum curvature (a spatial phase lag of roughly 2–4 channel widths, confirmed globally by Hu et al., 2025), which produces a prevailing downstream-rotating style of bend growth. Individual bends can be classified by migration regime: extensional (migration at the curvature maximum, typical of young bends), downstream-rotating (migration concentrated downstream of the curvature maximum), and upstream-rotating (migration concentrated upstream of it). Rivers transport sediment in two modes that need to be distinguished: suspended load (fine particles carried in the water column) and bed material (coarser sand and gravel that travels near the bed and is the sediment actually available for building point bars). Linear bar–bend theory — a linearized mathematical treatment of how alternate bars (migrating sediment deposits that shift periodically from one bank to the other) interact with bend curvature (Seminara, Zolezzi, Tubino & Zardi, 2001) — predicts that in channels wider than a resonant width-to-depth threshold (the channel aspect ratio at which the bar–bend response switches from damped to freely resonant behavior), morphodynamic influences propagate upstream, so upstream-directed behavior is not merely an anomaly but a theoretically meaningful regime; the long-standing mechanistic debate between 'bar push' (inner-bank deposition forces flow diversion and outer-bank erosion) and 'bank pull' (outer-bank erosion induces inner-bank deposition) remains unresolved (Eke et al., 2014; van de Lageweg et al., 2014), with a first quasi-global empirical test only very recently (Nagel, Darby, Leyland & Lazarus, 2026). In the Amazon Basin, tributaries draining the Andes carry sediment loads one to two orders of magnitude larger than lowland rivers draining cratonic shields (the ancient, tectonically stable, low-relief cores of continents, which yield little sediment), making the basin a natural laboratory. Constantine, Dunne, Ahmed, Legleiter & Lazarus (2014) showed basin-wide that higher suspended-sediment supply drives faster meander migration and more frequent cutoffs. Ahmed, Constantine & Dunne (2019) then analyzed Landsat satellite time series (1984–2014) of 16 Amazonian meandering reaches and found that sediment-rich rivers increase sinuosity faster and are dominated by downstream-rotating meanders with large, fully expanding point bars, whereas the few upstream-rotating meanders occur on sediment-poor rivers and show deposition over the bar head. They hypothesized that low sediment supply, by starving point-bar growth, initiates upstream rotation — but could not isolate this mechanism from three previously proposed alternatives: spatial variations in bank-material erodibility (and finer-scale floodplain alluvium heterogeneity, unresolvable from Landsat imagery), backwater effects created by meander cutoffs (the backwater effect is the upstream rise and flow deceleration caused by a downstream obstruction or confluence, here a newly shortened channel), and the upstream valley orientation of tightly curved composite bends. Because bed-material flux is unmeasured on these rivers, they used total suspended sediment flux normalized by bankfull channel width (the channel width at the flow stage that just fills the channel before spilling onto the floodplain) as a proxy, leaving the relationship between coarse sediment supply and bar construction untested.

Problem Statement

What physical mechanism initiates upstream-rotating meanders — bends whose lateral migration is concentrated upstream of the point of maximum channel curvature — in the low-sediment-supply rivers of the Amazon Basin? A complete answer must (1) test the hypothesis of Ahmed et al. (2019) that low supply of coarse bed material, and the consequent slow growth of point bars, causes bends to rotate upstream, using evidence that links coarse bed-material supply and point-bar construction dynamics — not merely suspended-sediment flux correlations — to rotation direction, by whatever combination of field, experimental, or numerical evidence is adequately justified; and (2) isolate this hypothesized mechanism from the previously proposed alternatives — spatial variations in bank-material (floodplain alluvium) erodibility, backwater effects associated with meander cutoffs, and the upstream valley orientation of tightly curved composite bends — for example through paired field reaches, controlled flume experiments, or numerical models in which sediment supply, bank erodibility, and geometry are varied independently. The question is answered when the dominant cause (or causes) of upstream rotation is identified with evidence that discriminates among these mechanisms.

References

  1. Ahmed, J., Constantine, J. A., Dunne, T. (2019). The role of sediment supply in the adjustment of channel sinuosity across the Amazon Basin. Geology 47(9), 807–810. doi:10.1130/g46319.1, https://doi.org/10.1130/g46319.1
  2. Constantine, J. A., Dunne, T., Ahmed, J., Legleiter, C., Lazarus, E. D. (2014). Sediment supply as a driver of river meandering and floodplain evolution in the Amazon Basin. Nature Geoscience 7, 899–903. doi:10.1038/ngeo2282, https://doi.org/10.1038/ngeo2282
  3. Greenberg, E., Ganti, V. (2024). The pace of global river meandering influenced by fluvial sediment supply. Earth and Planetary Science Letters 634, 118674. doi:10.1016/j.epsl.2024.118674, https://doi.org/10.1016/j.epsl.2024.118674
  4. Wu, C., Kim, W., Herring, R., Cardenas, B. T., Dong, T. Y., Ma, H., Moodie, A., Nittrouer, J. A., Tsai, F., Li, A. (2023). Lowland river sinuosity on Earth and Mars set by the pace of meandering and avulsion. Nature Geoscience 16, 747–753. doi:10.1038/s41561-023-01231-1, https://doi.org/10.1038/s41561-023-01231-1
  5. Cheng, Y., Chen, D., Li, Z., Finotello, A., Xia, J., Gao, P., Yue, Y., He, Y., Guo, X. (2026). Scaling of River Meander Skewing and Fattening Modulated by Environmental Conditions. Journal of Geophysical Research: Earth Surface 131(6), e2025JF008904. doi:10.1029/2025JF008904, https://doi.org/10.1029/2025JF008904
  6. Li, Y., Edmonds, D. A., Konsoer, K. M., Constantine, J. A., Jing, Y., Shoemaker, D. T., Islam, M. M. (2026). Floodplain slope organizes chute and neck meander cutoffs across rivers in the Amazon basin. Science Advances 12(21), eaef6658. doi:10.1126/sciadv.aef6658, https://doi.org/10.1126/sciadv.aef6658
  7. Abad, J. D., Motta, D., Guerrero, L., Paredes, M., Kuroiwa, J. M., Garcia, M. H. (2023). Hydrogeomorphology of Asymmetric Meandering Channels: Experiments and Field Evidence. Water Resources Research 59(7), e2022WR033904. doi:10.1029/2022WR033904, https://doi.org/10.1029/2022WR033904
  8. Hu, J.-C., Xia, J., Deng, S. (2025). Analysis of planform migration characteristics and lag effects in meandering rivers using remote sensing image interpretation. Science China Earth Sciences 68(3), 824–835. doi:10.1007/s11430-024-1483-7, https://doi.org/10.1007/s11430-024-1483-7
  9. Nagel, G. W., Darby, S. E., Leyland, J. (2023). The use of satellite remote sensing for exploring river meander migration. Earth-Science Reviews 247, 104607. doi:10.1016/j.earscirev.2023.104607, https://doi.org/10.1016/j.earscirev.2023.104607
  10. Nagel, G. W. (2025). Global detection and analysis of surface water and river meander dynamics using cloud-based remote sensing. PhD thesis, University of Southampton. https://eprints.soton.ac.uk/508230/
  11. Nagel, G. W., Darby, S. E., Leyland, J., Lazarus, E. D. (2026). Bank Pull or Bar Push: Who Leads the Dance of Meander Migration? EGUsphere preprint. doi:10.5194/egusphere-2026-4621, https://doi.org/10.5194/egusphere-2026-4621
  12. Seminara, G., Zolezzi, M., Tubino, M., Zardi, D. (2001). Downstream and upstream influence in river meandering. Part 2. Planimetric development. Journal of Fluid Mechanics 438, 213–230. doi:10.1017/s0022112001004281, https://doi.org/10.1017/s0022112001004281
  13. Eke, E., Parker, G., Shimizu, Y. (2014). Numerical modeling of erosional and depositional bank processes in migrating river bends with self-formed width: Morphodynamics of bar push and bank pull. Journal of Geophysical Research: Earth Surface 119(7), 1455–1483. doi:10.1002/2013JF003020, https://doi.org/10.1002/2013JF003020
  14. van de Lageweg, W. I., van Dijk, W. M., Baar, A., Rutten, J., Kleinhans, M. G. (2014). Bank pull or bar push: What drives scroll-bar formation in meandering rivers? Geology 42(4), 319–322. doi:10.1130/g35192.1, https://doi.org/10.1130/g35192.1
  15. Finotello, A., Ielpi, A., Lapôtre, M. G. A., Lazarus, E. D., Ghinassi, M., Carniello, L., Favaro, S., Tognin, D., D'Alpaos, A. (2024). Vegetation enhances curvature-driven dynamics in meandering rivers. Nature Communications 15, 1968. doi:10.1038/s41467-024-46292-x, https://doi.org/10.1038/s41467-024-46292-x