Symptomatic lithospheric drips triggering fast topographic rise and crustal deformation in the Central Andes

The study investigates whether rapid, localized subsidence-uplift cycles and internal crustal deformation in Central Andean hinterland basins are surface expressions of lithospheric foundering via drip-like downwellings. Prior conceptual models and seismic tomography indicate drip-like high wave speed anomalies beneath several plateaus worldwide, including the Puna Plateau. However, the surface tectonic response to such foundering is poorly constrained. The authors hypothesize that the Arizaro and nearby Puna basins experienced drip-induced deformation and topographic change, and they design analogue experiments to test how mantle lithosphere instabilities couple to the crust to generate basin subsidence, uplift, and shortening.

The paper references global evidence of lithospheric dripping/delamination from seismic tomography and geology beneath regions such as the Colorado Plateau, Sierra Nevada, Altiplano, Tibet, Carpathians, and Anatolia. Drip tectonics has been linked to intraplate magmatism (Mongolian Plateau, Siberian, North China, and Wyoming cratons). In the Central Andes, prior work proposed lithospheric removal to explain non-subduction-related magmatism, deformation, and pulsed uplift, and specifically suggested the Miocene Arizaro Basin as a drip manifestation based on rapid subsidence followed by uplift and internal shortening not explained by flexure or local extension. Studies also inferred lithospheric dripping beneath southern Arizaro and Salar de Atacama. Despite these proposals, direct links between geodynamic predictions and observed surface deformation were limited, motivating the analogue approach.

The authors performed three-dimensional scaled analogue experiments in a 25 × 25 × 21.5 cm plexiglass box filled with PDMS representing the sub-lithospheric mantle. The mantle lithosphere was a 2 cm-thick PDMS–modeling clay mixture with higher density and viscosity than the underlying PDMS. The brittle upper crust was a 2 mm-thick granular layer (silica spheres and ceramic E-spheres or sand); in EXP-3, a 5 mm-thick viscous lower crust (PDMS–clay mix) was added between the brittle crust and mantle lithosphere. Two methods initiated a localized instability: Method 1 (EXP-1, EXP-3) pushed a 2.5 cm diameter hemisphere into the mantle lithosphere from above before surface leveling; Method 2 (EXP-2) inserted the hemisphere into the PDMS, then placed the mantle lithosphere on top. The density contrast and minor thickness undulations drove Rayleigh–Taylor-type drips. Edges were lubricated with petroleum jelly to minimize sidewall drag; the model had a free surface. Imaging and measurements used synchronized cameras: two top-view cameras for Stereo Digital Image Correlation (SDIC) and digital photogrammetry to compute cumulative surface strain and topography (vertical precision ±0.0945 mm), and a side-view camera for 2D Particle Image Velocimetry (PIV) to track mantle flow and drip descent. Black sand served as crustal tracers; dispersed clay particles within recycled PDMS provided PIV markers. Scaling: length scale L = 2.5 × 10⁻⁷ based on 20 mm model mantle lithosphere to 80 km natural thickness; density scale 0.33; gravity scale 1; viscosity scale 1.5 × 10⁻¹⁷; time scale T = 1.82 × 10⁻¹⁰ so 1.59 h ≈ 1 Myr. Upper crust cohesion was negligible with angle of internal friction ~40° (EXP-1/3) and slightly lower in EXP-2. Experiments:

• EXP-1 (Method 1, no lower crust): tracked drip evolution, surface topography, and strain over 40 h (~25.2 Myr).
• EXP-2 (Method 2, no lower crust): same layer properties (except upper crust density/friction), weaker coupling between perturbation and lithosphere.
• EXP-3 (Method 1 + viscous lower crust): same as EXP-1 but with lower crust, testing its effect on coupling and surface response.

• Drip morphology and kinematics: All experiments formed a bulbous drip head with a thinning neck, generating mantle return flow. In EXP-1, descent was steady to ~18 h, paused, then slowed, with the head reaching the bottom by ~30 h. In EXP-2, the drip sank continuously, touching the bottom at ~25 h, with a narrower neck (~1/3 width of EXP-1). In EXP-3, descent occurred in two stages, slower than EXP-1, touching the bottom near ~37 h; the neck was ~2× wider than in EXP-1, and some lower crust was entrained.
• Surface topography and strain:
  • EXP-1 (symptomatic): Developed an approximately circular basin with central subsidence ~1.5 mm (~6 km scaled), peripheral uplift (>1 mm), and basin radius evolving from ~6 cm to ~1.5 cm as shortening progressed and dynamic support waned. Cumulative surface strain reached ~15% with wrinkle-like fold-and-thrust features indicating interior shortening; extension developed at basin margins.
  • EXP-2 (asymptomatic): Rapid initial depression formed before recording, followed by limited central depression persisting (~0.5 mm) and progressive topographic recovery. Horizontal convergence and cumulative strain remained negligible (<2% early; no appreciable crustal deformation at 40 h) despite active mantle dynamics.
  • EXP-3 (symptomatic with lower crust): Large, well-defined depression formed; cumulative strain reached ~20% with pronounced wrinkle-fold structures and marginal extension beginning ~10 h. Subsidence lasted longer and uplift rate was greater than in EXP-1.
• Symptomatic vs. asymptomatic classification: Well-coupled drips (EXP-1, EXP-3) produce interior shortening and fold-and-thrust structures; poorly coupled drips (EXP-2) yield transient subsidence and recovery with little-to-no shortening.
• Comparison to Central Andes: Experimental relative elevation curves (EXP-1, EXP-3) match three-stage Arizaro Basin evolution: Stage 1 accelerated subsidence (~1 mm model ≈ ~4 km natural) during initial strong downwelling; Stage 2 recovery as drip thins/necks; Stage 3 additional uplift due to continued lithosphere removal along the neck. Experimental basin widths scale to ~60–240 km, consistent with the 75–100 km diameter Arizaro Basin. Internal shortening and ridge-like structures (e.g., Sierra de Macon analogue) match observed basin-internal thrusting and folding. The presence of multiple oval high P-wave speed anomalies (A–D) beneath southern Puna aligns with multiple drips potentially beneath Arizaro, Salar de Atacama, and Salinas Grandes. Geophysical inferences of shallow Moho (~42 km) are consistent with modeled crustal removal.

The experiments directly link mantle lithospheric drip dynamics to observed surface tectonics in the Central Andes by showing that well-coupled (symptomatic) drips produce circular basins with rapid subsidence followed by uplift and coeval interior shortening and marginal extension—features documented in the Arizaro Basin stratigraphy and structure. The three-stage elevation history from the models reproduces the timing pattern inferred for Arizaro, including an early rapid subsidence phase, later slowing and lacustrine deposition, and subsequent uplift/inversion associated with ridge formation (Sierra de Macon). Tomographic high-velocity anomalies A–D beneath the southern Puna are consistent in size and geometry with drip heads. Gravity data indicating dense lithosphere beneath Arizaro–Atacama and a shallow Moho support lithosphere and lower crust removal, as reproduced in EXP-3. EXP-2 demonstrates that some drips may be asymptomatic, creating transient basins without detectable crustal shortening, providing an explanation for basins lacking internal contractional structures elsewhere. Overall, the models highlight that non-subduction mantle processes (lithospheric dripping) can be primary drivers of surface tectonics and plateau topography pulses in the Central Andes.

Three-dimensional analogue experiments demonstrate that lithospheric drips can be either symptomatic—well-coupled to the mantle lithosphere and crust, generating rapid subsidence, later uplift, and interior shortening—or asymptomatic, producing transient topography with minimal crustal deformation. The symptomatic cases reproduce the geometry, kinematics, and timing of the Arizaro Basin and nearby Puna basins, suggesting that these features formed due to lithospheric dripping rather than solely subduction-related processes. Multiple tomographically imaged high-velocity anomalies beneath southern Puna are consistent with drip heads influencing basin evolution. The study underscores the significant role of drip-driven, non-subduction geodynamics in surface tectonics and plateau uplift. Future work should incorporate thermal and compositional variations, explore a broader parameter space of rheologies and coupling, and integrate additional geological and geophysical constraints to refine the spatiotemporal evolution of dripping beneath the Central Andes.

Analogue models simplify Earth’s rheology and thermal structure: thermal effects were not explicitly modeled, and rheologies used discrete layers with non-Newtonian but simplified parameters. Box boundaries, though lubricated, still impose sidewall drag; imaging-derived mantle velocities are volume-integrated. Scaling assumes constant viscosity contrasts and densities; natural heterogeneity in composition, temperature, and hydration is not fully captured. Only specific initialization methods and perturbation geometries were tested, and results are sensitive to initial coupling between perturbation and lithosphere. The temporal correlation with Andean events is qualitative due to scaling uncertainties.