Earth Surface Processes and Landforms最新文献

The São Francisco River in Northeast Brazil has seen hydrological and morphological changes due to extensive damming and climate change over the past century. In this study, we examine the influence of human activities and natural fluctuations in precipitation on the hydrological patterns of the basin and the morphological responses of the lower course of the river (LOW-SF) to these alterations over a span of several decades. The findings indicate a decrease in water release by 41% from 1995 to 2013 and 54% from 2013 to 2018, solely attributed to human actions. Furthermore, the operation of the reservoirs of the large dams resulted in a reduction in hydrological seasonality. The changing hydrological regime caused morphological changes that resulted in an expansion of the exposed subaerial fluvial bars in the LOW-SF and a reduction in channel width. As a result, the abandonment of small secondary channels occurred, leading to the cessation of inundation in previously buried elevated portions of bars, even during certain seasons. Another important factor was the spread of morphological changes in the LOW-SF, which started from the areas farthest from the last dam in the series of large dams, the Xingó Dam, and spread to the nearby regions. This is due to the lack of major tributaries in the semiarid region of the LOW-SF. The integrated assessment presented in this study illustrates both natural and anthropogenic influences. Moreover, in light of projected declines in precipitation, it is anticipated that natural phenomena could result in a substantial 73% decrease in water flow by the mid-20th century. This climatic scenario will lead to increased utilization of hydroelectric plants and more stringent control of water flow downstream of the dam cascade, intensifying the already documented adverse effects and posing the possibility of novel morphological adaptations.

Glacial lake outburst floods (GLOFs) are sudden, and often hazardous, floods occurring upon the failure of a glacial lake dam or moraine. A GLOF occurred at Sulzenau Lake (Tyrol, Austria) in August 2017 due to a partial moraine and dam failure, damaging nearby infrastructure. Due to the ongoing retreat of Sulzenau Glacier, the areal extent, depth, water volume, and shoreline configuration of Sulzenau Lake fluctuate over both short- and long-term periods. Here, we used remote sensing data to create a detailed geomorphological overview of the valley, analyse the lake's evolution since 2009, and characterize the conditions leading to the 2017 dam failure. Using optical remote sensing imagery, we generated detailed pre- and post-event geomorphological maps of Sulzenau Lake and areas impacted by the GLOF to characterize erosional and depositional zones. We employed the Normalized Difference Water Index (NDWI) and mapped the post-event boulder distribution. Based on multi-temporal mapping, we calculated water volumes, analysed changes in lake and glacier surfaces since 1970, and compared them with ERA-5 meteorological data. Lake growth was primarily due to rising temperatures and glacier retreat. In 2017, both precipitation and air temperatures in the Sulzenau Valley exceeded the 1991–2021 averages, with precipitation 14.8% higher and air temperatures 0.35°C above the 30-year mean. Ice velocities for Sulzenau Glacier reached 170 m/year during 2015–2022. By modelling flow conditions required for observed boulder movements during the GLOF, we constrained the peak discharge to 150–200 m³/s. No significant pre-2017 GLOF activity or meteorological anomalies were detected. Accordingly, we attribute the GLOF and dam failure to an increased meltwater flux and increased precipitation, possibly augmented by subglacial/englacial lake drainage. The 2017 Sulzenau Valley GLOF is a pertinent example of environmental changes and associated hazards in high-mountain glacial environments due to global warming.

Severe wildfire may alter steep mountain streams by increasing peak discharges, elevating sediment and wood inputs into channels, and increasing susceptibility to landslides and debris flows. In the Pacific Northwest, where mean annual precipitation is high and mean fire-return intervals range from decades to centuries, understanding of steep stream response to fire is limited. We evaluate the hydrologic and geomorphic response of ~100-m-long steep stream reaches to the large-scale and severe 2020 fires in the Western Cascade Range, Oregon. In the two runoff seasons after the fires, peak flows in burned reaches were below the 2-year recurrence interval flood, a level sufficient to mobilize the median grain size of bed material, but not large enough to mobilize coarser material and reorganize channel morphology. Sediment inputs to study streams consisted of two road-fill failure landslides, slumps, sheetwash, and minor bank erosion; precipitation thresholds to trigger debris flows were not exceeded in our sites. There was a 50% increase in the number of large wood pieces in burned reaches after the fires. Changes in fluxes of water, sediment, and wood induced shifts in the balance of sediment supply to transport capacity, initiating a sequence of sediment aggradation and bed-material fining followed by erosion and bed-material coarsening. Gross channel form showed resilience to change, and an unburned reference reach exhibited little morphologic change. Post-fire recruitment of large wood will likely have long-term implications for channel morphology and habitat heterogeneity. Below-average precipitation during the study period, combined with an absence of extreme precipitation events, was an important control on channel responses. Climate change may have a complex effect on stream response to wildfire by increasing the propensity for both drought and extreme rain events and by altering vegetation recovery patterns.

Earthquakes are vibrations that occur on the surface of earth, generating fires, ground shaking, tsunamis, landslides and cracks. These incidents can cause severe damage and loss of life. Accurate earthquake forecasts are critical for anticipating and mitigating these hazards, which can avoid damage to buildings and infrastructure and save lives. To address the challenges given by earthquakes probabilistic nature, this paper presents a hybrid SARIMA–XGBoost approach to earthquake magnitude prediction. The suggested technique consists of a two-step process: an exploration phase that uses exploratory data analysis, which includes descriptive statistics and data visualisation, and a prediction phase that focusses on forecasting future earthquakes. Using a large significant earthquake dataset spanning 1965–2023, the study intends to gain insights and lessons for more effective earthquake prediction methods. Further, in a comparison analysis, the results of SARIMA-XGBoost model are compared to those of traditional ARIMA and SARIMA models. The results highlight the superior performance of the hybrid SARIMA–XGBoost model, showcasing a mean absolute error (MAE) of 0.038, a mean squared error (MSE) of 0.0040, and a root mean squared error (RMSE) of 0.068. These metrics collectively underscore the model's enhanced accuracy in forecasting earthquake magnitudes. The notably low values of MAE, MSE and RMSE indicate that our hybrid approach significantly improves prediction accuracy compared to alternative models. By integrating SARIMA's time series (TS) analysis with XGBoost's machine learning (ML) capabilities, the hybrid model reduces forecasting errors more effectively, demonstrating its clear advantage in precision.

Rock glaciers are characteristic landforms in alpine environments originating from the movement of permanently frozen ground. Hereby, rock glacier velocity (RGV) is an important parameter for understanding the effects of climate change on mountain permafrost. Although understanding of rock glacier dynamics has increased during the last decades, linking small-scale surface kinematics to sub-surface properties and heterogeneities remains a challenge. To address this gap, we conducted geophysical surveys (electrical resistivity tomography [ERT] and ground-penetrating radar [GPR]) along two profile lines of 450 and 950 m in length on a rock glacier in the Central Swiss Alps. Additionally, RGV was derived from Sentinel-1 differential synthetic aperture radar interferometry (DInSAR) to quantify annual east–west displacement and elevation change as well as seasonal acceleration during the snow-free summer months with a ground sampling distance of 5 m. Our results show that movement angle and seasonality are highly associated with different patterns in sub-surface structure. These different movement patterns are linked to subunits of different morphological origins. Thereby, we can upscale the geophysical results based on the DInSAR surface movement parameters and outline an area within the study site probably affected by ice of glacial origin. Hence, DInSAR movement angle and seasonality can help to bring local sub-surface information derived from time-consuming geophysical investigations into the spatial domain. In this way, a better understanding of the current morphodynamics as well as the past and future evolution of the landform can be reached. Applying the approach to other sites with available geophysical investigations could enhance our knowledge about systematic links between surface kinematics and the internal structure of rock glaciers and other ice-rich glacial and peri-glacial landforms, as well as their response to a warming climate.

The variations in soil erosion significantly impact regional ecological security. Under rapid urbanisation, extensive ecological restoration and climate change, soil erosion development in the red soil region of southern China is ambiguous. Therefore, this study investigated the current (1980s–2020) and future (2050) erosion characteristics in a typical soil erosion control catchment (Changting section catchment) in this region by using the Cellular Automata Markov model and CMIP6 data to predict future scenarios and the Revised Universal Soil Loss Equation to estimate soil erosion. The results showed significant changes in the vegetation coverage of major land uses from 1980s to 2020, which was mainly caused by continuous soil and water conservation (SWC). The land use subtypes that were obtained by reclassifying land use based on the threshold of vegetation cover on soil erosion control, reflect a continuous transformation from those with poor SWC effectiveness to those with great SWC effectiveness. Therefore, the estimated soil erosion intensity continued to decrease from 1980s to 2020, and the contribution of land use/land cover (LULC) impacts ranged from 74%–195%. However, predictions of land use subtypes indicated that LULC may be stable after 2020; thus, soil erosion changed little when the climate was almost unchanged in 2050. Under climate change scenarios, soil erosion may increase by 111%–121%, and the contribution of precipitation impacts was 63%–66%. The major driving factor of soil erosion changes may shift from LULC to precipitation after 2020. Therefore, in the future, the potential for reducing soil erosion by vegetation restoration may be limited, and more engineering measures should be applied to address the erosion risk caused by climate changes. This study provides prospects for land use/land cover and soil erosion in the red soil region of southern China.

Splays—fan-shaped depositional landforms produced by overbank deposition by unconfined flows—can damage structures, degrade arable land and incur substantial mitigation costs. Splay-related hazards along many rivers are likely to worsen with the increasing magnitude and frequency of major floods. The highly incomplete understanding of splays on braided streams is a conspicuous knowledge gap in a changing world with more frequent and intense floods. The largest recorded flood on the braided, sand-dominated lower Platte River (eastern Nebraska, USA) in March 2019 resulted from the rapid melting of a deep, moist snowpack during an extreme rain-on-snow, bomb-cyclone event. This flood produced 32 large (as much as 234 ha) splays that buried structures and cropland under sand. A total of 1,438 ha of row crop was buried, equating to 1.2 million dollars in lost revenue. These splays diverged from the channel by 14° to 104° along a 122 km reach. The topography of preexisting abandoned channels strongly controlled the shape and orientation of most splays, although forested areas tended to trap or divert sediment. The flood eroded 2.2 to 202 m² m⁻¹ of the streambank at 11 of the splays. The five largest splays (>100 ha) deposited as much as 2.4 m of sand. Ground-penetrating radar profiles of the largest splay indicate that it consisted almost entirely of overbank deposits exhibiting simple downstream accretion that buried the pre-flood soil under ≤ 1 m or less of sand. Locally, however, this soil was eroded during the flood. Climate models predict increasing winter precipitation in the Platte River basin; therefore, the frequency of major floods should increase, making splays recurrent hazards. Our geomorphic assessment of the splays on the lower Platte River illustrates the need for future hazard and mitigation planning.

The advent of machine learning techniques has led to a proliferation of landscape classification products. These approaches can fill gaps in wetland inventories across the United States (U.S.) provided that large reference datasets are available to develop accurate models. In this study, we tested the feasibility of expediting the classification process by sourcing requisite training and testing data from existing national-scale land cover maps instead of customized sample sets. We created a single map of water and wetland presence by intersecting water and wetland classes from available land cover products (National Wetland Inventory, Gap Analysis Project, National Land Cover Database and Dynamic Surface Water Extent) across the U.S. state of Arizona, which has fewer wetland-specific mapping products than other parts of the U.S. We derived classified samples for four wetland classes from the combined map: open water, herbaceous wetlands, wooded wetlands and non-wetland cover. In Google Earth Engine, we developed a random forest model that combined the training data with spatial predictor variables, including vegetation greenness indices, wetness indices, seasonal index variation, topographic parameters and vegetation height metrics. Results show that the final model separates the four classes with an overall accuracy of 86.2%. The accuracy suggests that existing datasets can be effectively used to compile machine learning training samples to map wetlands in arid landscapes in the U.S. These methods hold promise for the generation of wetland inventories at more frequent intervals, which could allow more nuanced investigations of wetland change over time in response to anthropogenic and climatic drivers.

Clustering and machine learning-based predictions are increasingly used for environmental data analysis and management. In fluvial geomorphology, examples include predicting channel types throughout a river network and segmenting river networks into a series of channel types, or groups of channel forms. However, when relevant information is unevenly distributed throughout a river network, the discrepancy between data-rich and data-poor locations creates an information gap. Combining clustering and predictions addresses this information gap, but challenges and limitations remain poorly documented. This is especially true when considering that predictions are often achieved with two approaches that are meaningfully different in terms of information processing: decision trees (e.g., RF: random forest) and deep learning (e.g., DNNs: deep neural networks). This presents challenges for downstream management decisions and when comparing clusters and predictions within or across study areas. To address this, we investigate the performance of RF and DNN with respect to the information gap between clustering data and prediction data. We use nine regional examples of clustering and predicting river channel types, stemming from a single clustering methodology applied in California, USA. Our results show that prediction performance decreases when the information gap between field-measured data and geospatial predictors increases. Furthermore, RF outperforms DNN, and their difference in performance decreases when the information gap between field-measured and geospatial data decreases. This suggests that mismatched scales between field-derived channel types and geospatial predictors hinder sequential information processing in DNN. Finally, our results highlight a sampling trade-off between uniformly capturing geomorphic variability and ensuring robust generalisation.

Grass-planting measure is a crucial vegetation approach to mitigate understory soil erosion and improve ecological environment in the red soil region of southern China. This study aimed to quantify the effects of grass (Paspalum wettsteinii Hackel.)-planting measures on runoff and sediment reduction on slopes of Masson pine plantations under rainstorm conditions. We conducted a rainfall simulation experiment at a rainfall intensity of 2.0 mm/min, comparing single strip (MT1, strip spacing: 145 cm), double strips (MT2, strip spacing: 70 cm), and triple strips (MT3, strip spacing: 45 cm) grass-planting measures on slope surface runoff generation and soil erosion processes of the young Masson pine (MT0, no grass strip) plantation, and the bare slope (CK) was selected as the control. Results revealed that grass-planting measures significantly decreased slope erosion parameters compared to CK and MT0. As the average grass coverage increased (MT1 from 10% to 25%, MT2 from 7.5% to 22.5%, MT3 from 7.3% to 25%), the slope surface erosion parameters under the grass-planting measures decreased, resulting in significantly improved runoff and sediment reduction benefits. The runoff reduction effect could reach 32%, while the sediment reduction effect could reach 88%. Moreover, MT3 demonstrated superior performance over MT2 and MT1, with minimal runoff and sediment reduction effects observed for the MT0. Overall, this study suggests that grass-planting measures, coupled with the increasing of grass coverage rates, significantly improve runoff and sediment reduction benefits on slopes in regions experiencing heavy rainfall. Among the tested configurations, MT3 emerged as most effective measure for controlling understory soil erosion in Masson pine plantations, especially when its average grass coverage rate reached 25%. These findings provide valuable insights for selecting appropriate grass-planting strategies, as well as for understanding the underlying mechanisms of how these measures mitigate soil erosion. This scientific reference will aid in the design and implementation of soil and water conservation measures in the region.