This work analyses the geomorphology, evolution and current kinematics of the emerged Pinoso Diapir (Eastern Betics, Spain), expressed as a salt fountain 333 m in local relief, covering 7.9 km2, and with an estimated volume of 791 Mm3. The diapir is developed in the Triassic Keuper Facies, a clay- and gypsum-rich evaporite formation with low halite purity, and consequently high dynamic viscosity and low susceptibility to dissolution. The salt extrusion, covered by a thick and resistant sealing caprock, displays: (1) a summit bulge above the feeder with internal drainage, riddled by large and deep caprock collapse sinkholes rooted in a perched salt karst; (2) steep salt sheets (i.e., namakiersi) in the flanks dominated by a fluvial landscape with radial drainage network; and (3) an extraordinary large landslide complex (ca. 81 Mm3) induced by fluvial undermining. The salt dome is surrounded by a ring-shaped salt-withdrawal basin that controls a peculiar subcircular drainage. It is proposed that salt breakout was initiated in Plio-Quaternary times by erosional unroofing, and that centripetal salt flow towards the diapir is mainly driven by differential overburden loading. Outcrops reveal that basal sliding plays a major role in salt sheet advance, over frontal rolling. A minimum long-term salt sheet advance rate of ≥1 m/ky has been estimated using luminesce ages from Quaternary alluvium overridden by the namakier. The contrasting morpho-hydrological differences between salt fountains developed in low- and high-purity salt formations are discussed. Solution mining operating since 1973 has not played a significant role in sinkhole development, as supported by historical imagery and geochronological data from a relict sinkhole. However, the progressive closure of the caverns by salt flow causes progressive subsidence in the dome crest at rates of around 1 cm/yr, which may have reversed the natural diapiric rising trend.
Extreme rainfall is a primary trigger of landslides. Yet, event-based inventories from dryland sedimentary landscapes remain scarce, and the spatial relationships between gridded hydrometeorological variables and landsliding in such settings remain poorly constrained. Well-documented rainfall-triggered events provide benchmarks for testing how dryland sedimentary landscapes respond to extreme hydrometeorological forcing. We examined landslides triggered by the March–April 2017 Comodoro Rivadavia storm in the Pampa de Salamanca, eastern Patagonia, where ~400 mm of rainfall caused catastrophic flooding. Using satellite imagery, we mapped 2555 predominantly shallow landslides across a 5660 km2 dryland region underlain by weakly lithified sediments. Analysis of hourly ERA5-Land precipitation and runoff for 1950–2025, using 1–72 h accumulations, shows that multi-day rainfall and runoff were exceptional relative to historical annual maxima and modeled return levels. Landslide-bearing cells were disproportionately located in areas where multi-day rainfall and runoff exceeded annual maxima in other years. At the grid scale, slope-normalized landslide density was weakly and non-significantly associated with precipitation and runoff maxima, indicating that these variables alone did not resolve the spatial distribution of failures. Runoff showed partial spatial correspondence with landslide clusters, particularly where elevated runoff coincided with higher terrain and lithostratigraphic units. These patterns are consistent with the rainfall event providing regional triggering conditions, whereas terrain- and lithology-conditioned hydrological response may have influenced where failures clustered. Results provide an event-based inventory and show that multi-day hydrometeorological rarity, interpreted alongside terrain, lithostratigraphy, vegetation and soil moisture, offers informative regional context for landslide clustering beyond rainfall-intensity maxima alone.
