Earth Surface Processes and Landforms最新文献

The Vychegda River valley, a tributary of the Severnaya Dvina, has been studied for over a century in the context of reconstructing the extent of the last glaciation and its impact on the surrounding landscapes. Flowing westward toward the former ice margin, the Vychegda River has led many researchers to hypothesize the repeated formation of ice-dammed lakes that may have spread across the valley and overflowed into the neighbouring Kama River basin. Some have also proposed that the tongue of the last ice sheet penetrated into the lower Vychegda valley.

To test these hypotheses, we investigated all major bank exposures along a 400-km stretch of the river's lower course, made two coring profiles, and obtained four dozen new OSL and ¹⁴C dates from mostly alluvial terrace deposits. At the confluence with the Yarenga River, where a moraine ridge—previously interpreted by some authors as the terminal moraine of the last glaciation—is located, we discovered a river terrace dated to about 140 ka. This suggests that the ridge corresponds to the penultimate glaciation rather than the last one. Homogenous sands, previously interpreted as lacustrine deposits of various ages, are now reinterpreted as aeolian deposits formed during the late MIS 2 to early Holocene, overlying the river terraces.

We conclude that during the Late Pleistocene, neither glacial tongues nor proglacial lakes reached the Vychegda valley, allowing fluvial conditions to persist throughout its extent. After removing the aeolian cover, the MIS 2 river terraces are found to be at the same elevation as, or even slightly lower than, the Holocene floodplain. This indicates a slight tendency towards aggradation in the valley during MIS 2 and the Holocene, except for the mouth reach, where a 3–4 m post-LGM incision occurred due to the glacial isostatic compensation following the retreat of the last Scandinavian Ice Sheet.

Undercut streambanks are a physical element of habitat formed by bank erosion processes. They offer overhead cover for fish and other biota, yet little is known about their frequency, geometry (width, length and area) or spatial distribution. This study used field data from the gravel-bed portion of a fluvial system with riffle-pool-bar bedforms in the northern California Coast Ranges to investigate (1) the frequency and geometry of undercuts, (2) the relation between reach-scale fluvial variables and undercut frequency and geometry and (3) the spatial distribution of undercut streambanks. Results showed that undercut streambank frequency was 60% higher and median undercut widths were 15% higher in the relatively wetter coastal zone supporting a conifer forest. The longest undercut lengths were similarly present in this zone. These channel reaches had a higher frequency of large wood than in drier reaches farther inland. Nonetheless, undercut widths were relatively consistent in comparison to the variability of undercut lengths and areas. Probability density function (PDF) statistics illustrate the magnitude-frequency distribution of undercut area. The power-law exponent of the linear portion of this curve, representing the smaller number of undercuts with larger areas, equals −1.54, a dimensionless metric useful for comparison to other stream environments. Significantly, variability in the frequency and the geometry of undercuts results from their spatial distribution relative to geomorphic elements within the fluvial system. The majority of the undercuts were present in floodplain sediment, on the opposite side of the channel from bar bedforms and adjacent to pools and large wood. As components of physical habitat, undercuts were more closely spaced than pools, suggesting they are significant in contributing to the physical heterogeneity of gravel-bed streams. Process-based investigations are required to understand geomorphic and ecological interactions that promote the formation and longevity of undercuts in fluvial systems where habitat restoration is urgent.

A river is considered supply limited, or hungry, when its capacity to transport sediment (strongly) exceeds the supply of sediment. Common consequences include erosion of bed material and channel incision, as observed in the heavily engineered river Meuse in the Netherlands, where these processes have contributed to the formation of scour holes. This study establishes a sediment budget for the Dutch part of the river Meuse, aiming to better understand the drivers of channel incision, and to assess the potential of mitigation by removing bank protection. We used multibeam data to track riverbed elevation changes, laser altimetry data describing banks and floodplains, a detailed dredging and nourishment database, recent bed material samples and hydrodynamic modelling results to set up a sediment budget for a 250 km stretch of the river from 2011 to 2019. Our findings reveal that in the Dutch Meuse River, annual sediment extraction volumes frequently exceed natural sediment loads several times. Rather than reduced inputs from tributaries, sediment mining and barrages reducing sediment connectivity control the hunger of the river. Removing bank protection alleviates supply-limited conditions, at least temporarily. The comprehensive sediment budget obtained in this study offers a knowledge base for sustainable river management, highlighting the importance of international collaboration in sediment management efforts.

Supergene minerals preserve a record of protracted exposure and weathering, rendering them valuable for understanding and reconstructing continental surface evolution and palaeoclimatic history. Determining when and how these minerals precipitated is fundamental for reconstructing the timing, nature, and controlling factor of weathering processes recorded in weathering profiles. In this study, we investigate a well-preserved, 5-m-thick lateritic ferruginous duricrust developed on low-relief uplands (~1100-m elevation) of the Brazilian Central Plateau (BCP). The BCP represents a high-standing postorogenic surface in southeastern Brazil, where numerous geochronological data provide a framework for regional comparison. Using (U–Th)/He geochronology on 100 haematite and goethite grains from nodular, pisolitic, and protopisolitic facies at three depths (~0.5 to ~5 m), we provide new constraints on the timing and possible controls of discrete weathering episodes in the BCP. Two successive weathering phases during the Cenozoic Era were identified. The older phase, recorded predominantly by haematite, occurred between ca. 35 and 24 Ma (Late Eocene–Oligocene) under seasonally contrasted tropical conditions. The younger phase, dated between ca. 17 and 8 Ma (Middle to Late Miocene), is characterized by widespread goethite precipitation under more humid and cooler climatic conditions that influenced the entire profile. These findings are consistent with (U–Th)/He datasets from nearby sites and confirm the spatial extent and synchronicity of these weathering events across the BCP. Comparison with geochronological data from different lithologies reveals a strong control of basement composition on weathering style and age distribution: duricrusts developed over igneous and sedimentary rocks yield well-clustered ages, whereas profiles over complex lithologies, such as cangas formed on Banded Iron Formations, show scattered and broadly distributed ages. This study contributes to refining the understanding of Cenozoic weathering dynamics and long-term landscape evolution across the BCP.

High-elevation regions are prone to severe natural disasters, particularly landslides, which pose significant risks to both infrastructure and human life. Understanding and predicting landslide susceptibility is crucial for effective disaster management and mitigation efforts. This study aims to enhance landslide susceptibility mapping by developing and comparing the performance of an extended random forest (ERF) architecture with that of an artificial neural network (ANN). The primary goal is to enhance prediction accuracy and reliability by applying advanced machine learning (ML) techniques and integrating data. We employed a comprehensive dataset comprising 140 landslides and nine major landslide conditioning factors. The dataset was partitioned into training (70%) and testing (30%). Feature selection was conducted using recursive feature elimination (RFE) to identify the most influential variables. The ERF architecture integrated predictions from five robust ML algorithms using appropriate weighting to optimise ensemble performance. The proposed models demonstrated high robustness, with the ERF achieving a prediction accuracy of 0.97, significantly outperforming the ANN (0.91). The hybrid ensemble architecture showed superior performance because of its robust training phase and integration of multiple strong learners. Validation was performed using persistent scatterer interferometry (PS-InSAR) from 2023 to assess deformation. PS-InSAR validation revealed significant deformation velocities, confirming the model's ability to identify high-susceptibility zones. These findings contribute to the field by providing a robust framework for landslide prediction and mitigation, particularly in high-risk regions.

The Brazilian coastal zone underwent significant changes through climatic variations and relative sea level (RSL) fluctuations during the Late Pleistocene and Holocene, resulting in notable geomorphological alterations. These transformations led to the formation of estuarine-lagoonal environments in the lowlands of the Brazilian coastal plains. With the RSL falling and coastline regression, depressed areas experienced siliciclastic and organic sedimentation, promoting the development of peatlands, which are wetland ecosystems dominated by Histosols. This study examines the depositional dynamics of a broad forested peatland, surrounded by hydromorphic mineral soils and adjacent to a large river, located 25 km from the coastline, which initially suggests a fluvial origin with no connection to estuarine environments in SE Brazil. Profiles of Histosols from the mire and Cambisols from the surrounding lower terraces were analysed for granulometry, and radiocarbon dating (¹⁴C) and optically stimulated luminescence (OSL) dating were performed. The geomorphological model indicates that, approximately 120,000 years ago, a rise in the RSL submerged the coastal plain, depositing sediments of the Marine Isotope Stage (MIS 5). Around 18 000 years ago (MIS 2), a fall in the RSL of over 100 m caused intense landscape dissection and the formation of incised valleys. During MIS 1, 7000 years ago, estuarine-lagoonal sedimentation occurred in the incised valleys. Subsequent regression and sediment filling transformed paleolagoons into lakes, fostering terrestrialisation and paludisation. At the southwestern edge of the peatland, high gravel content indicates colluvial deposition. The lower terrace near the peatland was dated to 5500 years ago and provided the surface for the deposition of the shell mound (sambaqui), which consists mainly of oyster shells, evidencing human occupation (shell mound builders) and indicating an ancient estuarine ecosystem surrounding the Brazilian coastal plain at that time.

Predicting physical and chemical erosion rate responses to climate change are an ongoing challenge in geomorphology. A promising approach for investigating this is by measuring transient variations in physical and chemical erosion rates during climatically variable time periods, which can be accomplished by measuring cosmogenic nuclide concentrations and chemical depletion in sedimentary deposits. Interpreting such measurements warrants applying landscape evolution models that track variations in topography, cosmogenic nuclide concentrations and chemical depletion in soils. We applied a recently developed model that tracks these quantities at Little Lake, Oregon. Previous studies documented variations in cosmogenic nuclide concentrations and chemical depletion in paleo-lake sediments from 50 ka BP to the present, a time interval that includes cooling before the Last Glacial Maximum and warming after it. We extended the model by adding climate-sensitive parameterizations for mineral dissolution, soil transport by frost heave and soil production by frost cracking. We conducted simulations driven by a paleo-temperature time series applicable to Little Lake. Simulations showed that a shift to frost heave, frost cracking and temperature-controlled mineral weathering and alteration elevated ¹⁰Be-inferred denudation rates and lowered chemical depletion fraction (CDF) values comparable to those observed in cores from paleo-Little Lake. In contrast, introducing a lake with no changes to process operation led to a decline in denudation rates. No single climate-sensitive process could reproduce both high inferred denudation rates and low CDF, indicating that all of the climate-sensitive processes modelled in our simulations are needed to explain observed values. Modelled denudation rates increased when the connection between frost cracking intensity and maximum soil production rate was strengthened. The integration of climate-sensitive processes showed that a handoff from biotically driven processes to frost-driven ones could induce large, detectable changes in both inferred denudation rate from ¹⁰Be and CDF, signalling the potential for globally heterogeneous climate-denudation rate linkages.

The Three Gorges Dam (TGD) and Xiaolangdi Dam have triggered substantial morphological changes along the middle Yangtze River and the lower Yellow River (LYR), respectively. Yet a comprehensive comparison of these changes is lacking, which hinders our understanding of large rivers' response to mega-dam operation. Based on systematic data of hydrology, sediment and cross-sectional surveys over the last two decades, we compared the water-sediment conditions and the morphological changes of the two rivers. Results showed that sediment load decreased by about 90% in both rivers while discharge reduction was minor. Both rivers underwent degradation, widening, decreased width-to-depth ratio (W/H), reduced slope and bed coarsening, but with different magnitudes, rates and spatial extents. The average incision depth of thalweg was 4.2 and 4.6 m, and the maximum was 24.9 and 11.9 m at the Yichang-Chenglingji Reach (~380 km) in the middle Yangtze River and the LYR (~730 km), respectively. The upper Jingjiang reach experienced the most severe degradation within the middle Yangtze River, with no temporal declining trend of the erosion rate. Erosion centers were primarily located in the braided reach of the LYR. The erosion rate at this reach decreased recently after about 2017, while it increased at the downstream transitional and meandering reaches. The changes in the LYR agree with Lane's balance, and W/H was negatively correlated with water discharge multiplied by slope. Heterogeneous changes along the channels prohibited the application of simplified, unified frameworks for channel adjustment to regulated water and sediment.

Geomorphology plays a pivotal role in Earth sciences by unravelling the dynamic interplay among landforms, natural processes and anthropogenic influences. With the development of Earth observation techniques, remote sensing has revolutionised geomorphology by enabling large-scale, high-resolution analysis of Earth surface processes and landforms. These help to achieve more accurate, scalable and process-oriented geomorphological interpretations of surface morphology, spatial patterns, and evolution processes across a range of spatial and temporal scales. This Special Issue of ‘Remote Sensing Applications in Geomorphology’ highlights the applications of remote sensing in geomorphological research. The contributions of this Special Issue can be summarised into four aspects according to their methodological innovations and insights. They are modelling and representation of landform morphology, extraction and mapping of geomorphic features, exploration and revelation of landform processes, and those particularly empowered by AI. Each aspect effectively demonstrates the application of remote sensing data and methodologies in geomorphological research. Moreover, potential perspectives for the remote sensing application in geomorphological research are discussed from the above four aspects to achieve a better understanding of future development.

The reconstruction of ice-sheet dynamics can be achieved by analysing changes in the morphology of landforms that form an evolutionary sequence, in which individual features may either align with or be transverse to the direction of ice flow. Over time, the transformation of these landforms in response to ice-sheet processes demonstrates reorientation and successive changes in the relationship between processes and forms, typically in the direction of ice flow. This paper explores the evolving structure of a landsystem and the dynamics of the lobate margin of the Scandinavian Ice Sheet, along with its palaeoglaciological conditions during the Late Weichselian glaciation near Rajgród (NE Poland), where ridges transverse to the ice-sheet flow have been identified. These features include large-scale ribs, highly transformed ribs with superimposed smaller ribbed moraine and controlled moraine. The identified ribbed bedforms represent incipient linear and transitional types (characterised by up-ice-pointing horns), reflecting changes in the stick–slip sliding behaviour at the base of the ice-sheet lobe.

The zone of controlled moraine suggests the presence of debris-rich ice, likely frozen to its bed, where permafrost may have extended from the proglacial area. The occurrence of small-scale ribs and the zone of controlled moraine further confirm the patchwork system and band-like pattern of increased basal shear stress, which would have decelerated ice-flow velocity within the ice lobe formed by the Mazury Palaeo-ice Stream. These landforms, along with flute-like features, were superimposed on pre-existing large-scale ribs, indicating the potential for multiple slowdowns in ice flow, interrupted by phases of faster movement. Consequently, the landforms represent a complex palimpsest in the lobate margin of the ice sheet, which evolved during the waning stages of the Weichselian deglaciation, reflecting the changing dynamics of compressional ice flow. Lastly, the doughnut-shaped and hummocky moraine most likely developed under differential redistribution of supraglacial debris, resulting from the down-wasting of stagnant ice.