Advances in geomorphometry and landform mapping: possibilities, challenges and perspectives
Advances in geomorphometry and landform mapping: possibilities, challenges and perspectives
Geomorphometry and landform mapping are important tools used for understanding landscape processes and dynamics on Earth and other planetary bodies. The recent rapid advances in technology and data collection methods have made available vast quantities of geospatial data offering unprecedented spatio-temporal range, density, and resolution, but it also created new challenges in terms of data processing and analysis.
This inter-disciplinary session on geomorphometry and landform mapping aims to bridge the gap between process-focused research fields and the technical domain where geospatial products and analytical methods are developed. The increasing availability of a wide range of geospatial datasets requires the continued development of new tools and analytical approaches as well as landform/landscape classifications. However, a potential lack of communication across disciplines results in efforts to be mainly focused on problems within individual fields. We aim to foster collaboration and the sharing of ideas across subject-boundaries, between technique developers and users, enabling us as a community to fully exploit the wealth of geospatial data that is now available.
We welcome perspectives on geomorphometry and landform mapping from ANY discipline (e.g. geomorphology, planetary science, natural hazard assessment, computer science, remote sensing). This session aims to showcase both technical and applied studies, and we welcome contributions that present (a) new techniques for collecting or deriving geospatial data products, (b) novel tools for analysing geospatial data and extracting innovative geomorphometric variables, (c) mapping and/or morphometric analysis of specific landforms as well as whole landscapes, and (d) mapping and/or morphometric analysis of newly available geospatial datasets. Contributions that demonstrate multi-method or inter-disciplinary approaches are particularly encouraged. We also actively encourage contributors to present tools/methods that are “in development”.
Lorena Abad, Daniel Hölbling, Zahra Dabiri, and Benjamin Robson
Landslide assessments require timely, accurate and comprehensive information, where Earth observation (EO) data such as optical and radar satellite imagery has played an important role. Volume estimates are important to understand landslide characteristics and (post-failure) behaviour. Pre- and post-event digital elevation model (DEM) differencing is a suitable method to estimate landslide volumes remotely, leveraging EO techniques. However, high costs for commercial DEM products, limited temporal and spatial coverage and resolution, or insufficient accuracy hamper the potential of this method. Sentinel-1 synthetic aperture radar (SAR) data from the European Union's Earth observation programme Copernicus opens the opportunity to leverage free EO data to generate multi-temporal topographic datasets.
With the project SliDEM (Assessing the suitability of DEMs derived from Sentinel-1 for landslide volume estimation) we explore the potential of Sentinel-1 for the generation of DEMs for landslide assessment. Therefore, we develop a semi-automated and transferable workflow available through an open-source Python package. The package consists of different modules to 1) query Sentinel-1 image pairs that match a given geographical and temporal extent, and based on perpendicular and temporal baseline thresholds; 2) download and archive only suitable Sentinel-1 image pairs; 3) produce DEMs using interferometric SAR (InSAR) techniques available in the open-source Sentinel Application Platform (SNAP), as well as performing necessary post-processing such as terrain correction and co-registration; 4) perform DEM differencing of pre- and post-event DEMs to quantify landslide volumes; and 5) assess the accuracy and validate the DEMs and volume estimates against reference data.
We evaluate and validate our workflow in terms of reliability, performance, reproducibility, and transferability over several major landslides in Austria and Norway. We distribute our work within a Docker container, which allows the usage of the SliDEM python package along with all its software dependencies in a structured and convenient way, reducing usability problems related to software versioning. The SliDEM workflow represents an important contribution to the field of natural hazard research by developing an open-source, low-cost, transferable, and semi-automated method for DEM generation and landslide volume estimation.
How to cite:
Abad, L., Hölbling, D., Dabiri, Z., and Robson, B.: An open-source Python package for DEM generation and landslide volume estimation based on Sentinel-1 imagery, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-693, https://doi.org/10.5194/egusphere-egu22-693, 2022.
Spaceborne digital elevation models (DEMs) are fundamental data for mapping and analyzing geomorphic features at regional and continental scale, but are limited by both their spatial resolution and accuracy. Typically, accuracy is measured using point- or profile-based geodetic measurements (e.g., sparse GNSS). We develop new methods to quantify the vertical uncertainty in spaceborne DEMs relevant to geomorphic analysis, focusing on the pixel-to-pixel variability internal to a given DEM, which we term the inter-pixel consistency. Importantly, the methods we develop are not based on external, geodetic measurements. Our codes are published open-source (https://github.com/UP-RS-ESP/DEM-Consistency-Metrics), and we particularly highlight a novel sun-angle rotation and hillshade-filtering approach that is based on the visual, qualitative assessment of DEM hillshades. Since our study area is in the arid Central Andes and contains diverse steep (volcano) and flat (salar) features, the environment is ideal for vegetation-free assessments of DEM quality across a range of topographic settings. We compare global 1 arcsec (~30 m) resolution DEMs (SRTM, ASTER, ALOS, TanDEM-X, Copernicus), and find high quality (high inter-pixel consistency) of the newest Copernicus DEM. At higher spatial resolution, we also seek to improve the stereo-processing of 3 m SPOT6 optical DEMs using the open-source AMES Stereo-Pipeline. This includes optimizing key parameters and processing steps, as well as developing metrics for DEM uncertainty masks based on the underlying image texture of the optical satellite scenes used to triangulate elevations. Although higher resolution spaceborne DEMs like SPOT6 are only available for limited spatial areas (depending on funds and processing power), the improvement in geomorphic feature identification and quantification at the hillslope scale is significant compared to 30 m datasets. Improved DEM quality metrics provide useful constraints on hazard assessment and geomorphic analysis for the Earth and other planetary bodies.
How to cite:
Purinton, B. and Bookhagen, B.: DEM quality assessment and improvement in noise quantification for geomorphic application in steep mountainous terrain, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-1191, https://doi.org/10.5194/egusphere-egu22-1191, 2022.
Vincent Regard, Mélody Prémaillon, Thomas Dewez, Sébastien Carretier, Catherine Jeandel, Yves Godderis, Stéphane Bonnet, Jacques Schott, Kevin Pedoja, Joseph Martinod, Jérôme Viers, and Sébastien Fabre
The eroding rocky coasts export sediment to the ocean, the amount of which is poorly known. At the global scale it could amounts 0.15-0.4 Gt/a (1). Recent evaluations of large retreat rates on monitored sections of sea cliffs indicate it can be comparable to the sediment input from medium to large rivers. We quantify rocky coast input to the ocean sediment budget at the European scale, the continent characterized by the best dataset.
The sediment budget from European rocky coasts has been computed from cliff lengths, heights and retreat rates. For that, we first compiled a large number of well-documented retreat rates; the analysis of whom showed that the retreat rates are at first order explained by cliff lithology (GlobR2C2, 2). Median erosion rates are 2.9 cm/a for hard rocks, 10 cm/a for medium rocks and 23 cm/a for weak rocks. These retreat rates were then applied to the European coast classification (EMODnet), giving the relative coast length for cliffs of various lithology types. Finally the cliff height comes from the EU-DEM (https://ec.europa.eu/eurostat/web/gisco/geodata/reference-data/elevation).
Due to data availability, we only worked on ~70% of the whole Europe, corresponding to a 127,000 km-long coastline (65,000 km of rocky coast). We calculated it originates 111±65 Mt/a, corresponding to 0.38 times the sediment input from rivers from the equivalent area (3.56 106 km2), calculated after Milliman and Farnsworth (3)’s database (290 Gt/a). A crude extrapolation to the 1.5 106 km-long Earth’s coastline reaches an amount of 0.6-2.4 Gt/a, an order of magnitude less that the sediment discharge from rivers (11-21 Gt/a, e.g., 3).
This up-to-now overlooked sedimentary source must further be explored for: (i) its effects on the geochemical ocean budget; (ii) the rising sea level control on the cliff retreat rates; and (iii) the characteristics and location of sediment deposition on ocean margins.
(1) Mahowald NM, Baker AR, Bergametti G, Brooks N, Duce RA, Jickells TD, Kubilay N, Prospero JM, Tegen I (2005). Atmospheric global dust cycle and iron inputs to the ocean: ATMOSPHERIC IRON DEPOSITION. Global Biogeochemical Cycles 19. DOI: 10.1029/2004GB002402
(2) Prémaillon M, Regard V, Dewez TJB, Auda Y (2018). How to explain variations in sea cliff erosion rates? Insights from a literature synthesis. Earth Surface Dynamics Discussions:1–29. DOI: https://doi.org/10.5194/esurf-2018-12
(3) Milliman J, Farnsworth K (2011). River Discharge to the Coastal Ocean: A Global Synthesis. Cambridge University Press
How to cite:
Regard, V., Prémaillon, M., Dewez, T., Carretier, S., Jeandel, C., Godderis, Y., Bonnet, S., Schott, J., Pedoja, K., Martinod, J., Viers, J., and Fabre, S.: Coastal erosion: an overlooked source of sediments to the ocean. Europe as an example, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-2877, https://doi.org/10.5194/egusphere-egu22-2877, 2022.
Linear dunes show a wide variety of morphometrical patterns; their sizes, spacing, defect density, and orientations differ not only between but also within dunefields (Thomas 1986; Bullard et al. 1995; Hesse 2011; Hugenholtz et al. 2012). The first step towards characterising dune patterning is to accurately and precisely map dunefields, which is challenging, especially when dunefields are too large to be mapped manually. Thus, (semi-)automatic approaches have been brought forward (Telfer et al. 2015; Shumack et al. 2020; Bryant & Baddock 2021). Here, we are presenting the prototype of a deep learning workflow that allows for the automated mapping of large linear dunefields through semantic segmentation.
The algorithm includes the following components: 1) the download of satellite imagery; 2) pre-processing of training and prediction data; 3) training of a Neural Network; and 4) applying the trained Neural Network to classify satellite imagery into dune and non-dune pixels. The workflow is python-based and uses the deep learning API keras as well as a variety of spatial analysis libraries such as earthengine and rasterio.
A case study to apply and test the algorithm’s performance was conducted on Sentinel-2 satellite imagery (10 m spatial resolution) of the southwest Kalahari Desert. The resulting predictions are promising, despite the small amount of data the model was trained on.
The presented prototype is work in progress. Further developments will include parameter optimisation, exploring ways to improve the objectiveness of training data, and the conduction of case studies applying the algorithm to digital elevation rasters.
How to cite:
Nowatzki, M., Bailey, R., and Thomas, D.: Prototype of a deep learning workflow to map dunes in the Kalahari, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-3002, https://doi.org/10.5194/egusphere-egu22-3002, 2022.
Pit-and-mound (treethrow, windthrow) topography is a result of tree uprooting caused by the impact of hurricane-speed wind events. Analyzing its location and morphometric features can improve our knowledge about the influence of winds on forest ecosystem dynamics and on changes in the forest floor microrelief. This is important in terms of hillslope denudation and soil evolution.
The occurrence and evolution of pit-mound topography can be studied with the use of high-resolution elevation data. Such data can be obtained from LiDAR (Light Detection and Ranging) surveys. Polish Institute of Geodesy and Cartography carried the LiDAR survey in the years 2010-2015. Point cloud data for the entire area of Poland with the minimal density of 4 points per m2 is currently available on the Internet.
Under the present project, we have analyzed Digital Elevation Models (DEMs) produced from the above-mentioned LiDAR data in order to develop and test a new method for automatic detection of pit-mound topography. As far as we know, no such method exists at the moment. We generated DEMs with 0.5 m spatial resolution for three study sites with the confirmed occurrence of pit-mound topography, located in Southern Poland. A script with the method was written in the R programming language.
The proposed method is based on contour lines. We found that the detection of pit and mound topography formed on gentle hillslopes is possible when closed contours are delineated. Detected forms can be classified into “pits” and “mounds” by investigating point positions with the highest and the lowest elevation within the closed contour. On the other hand, for steep surfaces pit-mound topography can be detected by calculating distances between contours and selecting slope segments with between-contours distances above a certain threshold value. This leads to the identification of gently-sloped areas within the study site. With a high probability, such areas indicate places, where pit-mound topography was formed. To validate our methods, we performed the on-screen assessment of DEMs for the presence of forms that could be interpreted as pit-mound topography.
The study has been supported by the Polish National Science Centre (project no 2019/35/O/ST10/00032).
How to cite:
Godziek, J. and Pawlik, Ł.: Automatic detection of pit-mound topography from LiDAR based DEMs, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-3781, https://doi.org/10.5194/egusphere-egu22-3781, 2022.
Frances E. G. Butcher, Anna L. C. Hughes, Jeremy C. Ely, Christopher D. Clark, Emma L. M. Lewington, Benjamin M. Boyes, Alex C. Scoffield, Stephen Howcutt, and Thomas P. F. Dowling
Data-driven reconstructions of palaeo-ice sheets based on their landform records are required for validation and improvement of numerical ice sheet models. In turn, such models can be used to better predict the future responses of the Antarctic and Greenland ice sheets to climate change. We are exploiting the recent expansion in availability and coverage of very-high-resolution (1–2 m) digital elevation models (DEMs) within the domain of the former Fennoscandian Ice Sheet to reconstruct its flow pattern evolution from the glacial landform record.
The Fennoscandian Ice Sheet reached its maximum extent at 21–20 ka. Previous data-driven reconstructions over the whole ice sheet domain (encompassing Fennoscandia, northern continental Europe and western Russia) have necessarily relied upon landform mapping from relatively coarse-resolution (decametre-scale) data, predominantly from satellite images and aerial photographs. However, high-resolution (1–2 m/pixel resolution) LiDAR DEMs have recently become available over a large portion of the ice sheet domain above contemporary sea level. This reveals previously unobserved assemblages of landforms which record past ice sheet flow, including fine-scale cross-cutting and superposition relationships between landforms. These observations are likely to reveal previously unidentified complexity in the flow evolution of the ice sheet. However, the richness of the data available over such a large area amplifies labour-intensity challenges of data-driven whole-ice-sheet reconstructions; it is not possible to map every flow-related landform (or even a majority of the landforms) manually in a timely manner. We therefore present a new multi-scale sampling approach for systematic and comprehensive ice-sheet-scale mapping, which aims to overcome the data-richness challenge while maintaining rigor and providing informative data products for model-data comparisons.
We present in-progress mapping products covering Finland, Norway and Sweden produced using our new multi-scale sampling approach. The products include mapping of >200 000 subglacial bedforms and bedform fields, and a summary map of ‘landform linkages’. Landform linkages summarise the detailed landform mapping but do not extrapolate over large distances between observed landforms. Thus, they provide a reduced data product that is useful for regional-scale flow reconstruction and model-data comparisons and remains closely tied to landform observations. The landform linkages will be reduced further into longer interpretative flowlines, which we will then use to generate ‘flowsets’ describing discrete ice flow patterns within the ice sheet. We will use cross-cutting relationships observed in the detailed landform mapping to ascribe a relative chronology to overlapping flowsets where relevant. We will then combine the flowsets into a new reconstruction of the flow pattern evolution of the ice sheet.
How to cite:
Butcher, F. E. G., Hughes, A. L. C., Ely, J. C., Clark, C. D., Lewington, E. L. M., Boyes, B. M., Scoffield, A. C., Howcutt, S., and Dowling, T. P. F.: A new, multi-scale mapping approach for reconstructing the flow evolution of the Fennoscandian Ice Sheet using high-resolution digital elevation models., EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-4765, https://doi.org/10.5194/egusphere-egu22-4765, 2022.
Giuseppe Spilotro, Gioacchino Francesco Andriani, Giuseppe Di Prizio, Katia Decaro, Alessandro Parisi, and Maria Dolores Fidelibus
The Bradanic Trough (Southern Italy) is the Pliocene-present-day south Apennines foredeep. It is filled by a thick Pliocene to Pleistocene sedimentary succession constituted by hemipelagites (Blue Clay Fm.) in the lower part, and coarse grained deposits (sands and conglomerates) in the upper part, shaped in marine or continental terraced environment.
On the eastern border of the Bradanic Trough along the Murgian Plateau (Apulia, Italy) numerous morphological lineaments are associated with sequential lowering and rotation of the surface, aligned with the carbonate substrate dip direction.
These morphologies have been interpreted so far as erosion products; their association with medium-deep water circulations and surface phenomena, like mud volcanoes, now allows their interpretation as a lumped mass, detached and tilted along shear surfaces.
The surface patterns of such surfaces may be easily detected for the presence, at some distance, of a quite similar twin track, which overlaps with good agreement.
The numerical analysis of the tracks extracted from accurate DTMs allows us to reconstruct the kinematic patterns of the tectonic displacement (distance of the detachment; rotation; angle of the shear plane). This type of analysis might reveal very useful in some fields of engineering geology, such as underground works, and for interpreting many hydrogeological phenomena within the study area. Finally, the correct 3D representation of the detached masses helps to identify the true causes of the direct faulting, which is not always linked to the tectonics, not active in the concerned regions.
How to cite:
Spilotro, G., Andriani, G. F., Di Prizio, G., Decaro, K., Parisi, A., and Fidelibus, M. D.: Kinematic patterns of tectonic displacements in the Blue Clay outcrops along the eastern border of the Bradanic Trough (Southern Italy) from DTM data processing, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5872, https://doi.org/10.5194/egusphere-egu22-5872, 2022.
The Gulf of Mexico is characterized by a high geodiversity that influences hydrodynamics patterns and drives biological and human uses of the seafloor. In 2017, the United States Bureau of Ocean Energy Management released a 1.4-billion-pixel bathymetric dataset of the deep northern Gulf of Mexico, with a pixel size of about 12m. The computational power required to analyze this dataset has limited its use so far. Here, geomorphometry was used to characterize the seafloor of the deep northern Gulf of Mexico at multiple spatial resolutions. Flat areas and slopes cover more than 70% of the studied area, yet thousands of smaller morphological features like peaks and pits were identified. Spatial comparisons confirmed that analyses at different spatial scales capture different features. A composite product combining seafloor classification at multiple scales helped highlight the dominant seafloor features and the scale at which they are best captured.
How to cite:
Lecours, V.: Geomorphometry of the deep Gulf of Mexico, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5990, https://doi.org/10.5194/egusphere-egu22-5990, 2022.
Roos M. J. van Wees, Daniel O'Hara, Pablo Grosse, Gabor Kereszturi, Pierre Lahitte, and Matthieu Kervyn
The long-term (ka to ma) degradation of a volcanic edifice is controlled by both regional (e.g., climate, tectonics) and local factors (e.g., original morphology, lithology), resulting in both long-lasting weathering and river incision and short-term hazardous events, such as flank collapses and lahars. Trends among the morphometry of stratovolcanoes, their drainage network, denudation, and regional factors were recently characterised for composite volcanoes along the Indonesian arc. Denudation was shown to be negatively correlated with drainage density; the across-arc variations expose a tectonic control on the level of denudation and volcanoes’ irregularity. This study applies the same method on age-constrained volcanoes in Japan to find coherent trends between arcs despite the different local and regional factors. We aim to better understand the factors that control erosion rates and patterns, and the evolutionary phases of volcano degradation.
We first compile a dataset of 35 singular, non-complex composite volcanoes with known eruption ages and spatially spread throughout the Japanese Island arc system. Using 30m TanDEM-X Digital Elevation Models, morphologies, and drainage metrics (e.g., volume, height, slopes, irregularity index, Hack’s Law exponent, and drainage density) are extracted for each volcano, using the MORVOLC algorithm adapted in MATLAB as well as the newly developed DrainageVolc algorithm. Correlations between the morphometric parameters and potential controlling factors (e.g., age, climate, lithology, and tectonics) are analysed to determine quantitative relationships of edifice degradation throughout the arc. Finally, we compare relationships and correlation values of the Japanese Arc system to those from the Indonesian Arc.
The analysis shows that volcano age is positively correlated with irregularity and negatively correlated with height and volume. From the drainage parameters, we find that basins become wider and merge, resulting in lower drainage densities. The variation in erosion rates along the Japanese arc provides evidence for the degree of climatic control on the volcano degradation. The between-arc comparison shows which trends are susceptible to arc-scale variations and highlights consistent trends that have the potential to be extrapolated to other volcanic arcs and be used as a relative age determination tool for composite volcanoes.
How to cite:
van Wees, R. M. J., O'Hara, D., Grosse, P., Kereszturi, G., Lahitte, P., and Kervyn, M.: Quantifying the morphometry and drainage patterns of composite volcanoes: A comparison of the Japanese and Indonesian volcanic arcs , EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-6152, https://doi.org/10.5194/egusphere-egu22-6152, 2022.
The Augšdaugava spillway valley located in SE Latvia has a system of river terraces formed by both glacio-fluvial and fluvial processes. The flight of terraces forms a staircase-like relief in the riverine landscape and represents the evidence of valley evolution during the transition from glacial to post-glacial conditions in this region. Hence terraces are substantial ‘archives’ of paleoenvironmental data and their geomorphometry could provide key information for untangling geomorphological history of the spillway valley. Hence the need for precise identification and mapping of terraces is obvious. However, these landforms, particularly upper terraces commonly are poorly preserved. It is a result of the interplay of many geological processes – channel incision, lateral erosion in the course of the river Daugava meandering, mass wasting etc., leaving discontinuous remnants of terraces along to the present-day long profile of the river. Previously, mapping of these features was performed via extensive field surveys and to some extent by interpretation of aerial images or topographic maps, because the presence of tree cover hinders the identification of terraces by conventional geomorphological techniques. Thereby due to the poor preservation of fluvial landforms and the abundant vegetation cover, the previously mapped terrace surfaces and inferred levels may be questionable.
Yet the now available high-resolution LiDAR data in Latvia and application of modern GIS-based techniques offer an opportunity to resolve these problems. Hence the main goal of the study was to apply a methodology based on using LiDAR-derived DEM and combining different semi-automated GIS analysis tools for the identification, mapping and morphometric analysis of fluvial terraces in the valley. In this study, LiDAR data coverage (courtesy of the Latvian Geospatial Information Agency) was used to generate a DEM. LiDAR coverage consists of 317 data folders in *.LAS format, each one of 1 km2 extent. DEM with 0.5 x 0.5 m pixel resolution and <15 cm vertical accuracy was created by ArcGIS PRO tool ‘LAS Dataset to Raster’ following the standard procedure of the IDW interpolation. After the construction of DEM, the TerEx toolbox integrated into the ArcGIS environment was used for the extraction and delineation of terrace surfaces. After the completion of GIS works, the ground-truthing of the obtained data on the location of fluvial terraces was performed during field geomorphological reconnaissance.
DEM analysis allowed to identify the terrace sequence in the Augšdaugava spillway valley consisting of eight different terrace levels – T1 to T8. From the applied methodology, authors were able to delineate surfaces of river terraces in those parts of the valley, where in the course of previous research terraces were interpreted incorrectly or even not identified at all. However, only terraces T1 and T2 can only be unambiguously identified by GIS-based extraction. Upper terraces with smoothened edges due to mass wasting and surfaces dissected by gullies are not easily recognizable. Hence, the presence of minor landforms which increase the topographical roughness of the surface directly influences the quality of extracted data, thus leading to the necessity of an extensive amount of manual editing.
How to cite:
Soms, J. and Vorslavs, V.: Identification, GIS-based mapping and morphometric analysis of river terraces from airborne LiDAR data in the Augšdaugava spillway valley, South-eastern Latvia, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-6177, https://doi.org/10.5194/egusphere-egu22-6177, 2022.
Kathryn Russell, Jonathan Garber, Karen Thompson, Jasper Kunapo, Matthew Burns, and Geordie Zhang
Bankfull channel dimensions are of fundamental importance in fluvial geomorphology, to describe the geomorphic character of a river, as inputs to models which explain variations in morphology through time and space, and as initial processing steps in more detailed morphometric techniques. With ever-increasing availability of high-resolution elevation data (e.g. LiDAR), manual delineation of channel extents is a bottleneck which limits the geomorphic insights that can be gained from that data.
We developed and tested two automated channel delineation methods that define bankfull according to different criteria and thus reflect different conceptualisations of bankfull extent: (1) a cross-sectional method (termed HydXS) that identified the elevation which maximises hydraulic depth (cross-section area/wetted width); and (2) a neural network image segmentation model trained on images derived from a LiDAR digital elevation model.
HydXS outperformed the neural network method overall, but the two methods were comparable in larger streams (> 20 m bankfull width; Dice coefficient ~0.85). Prediction accuracy of HydXS was generally high (overall precision 89%; recall 81%), performing well even in small streams (bankfull width ~ 10 m). HydXS performed worst in incised and recovering stream sections (precision 93%; recall 64%) where the choice between macro-channel and inset channel was somewhat arbitrary (both for the algorithm and manual delineation). The neural network outperformed HydXS where an inset channel was present. The neural network method performed worst in small streams and where other features (e.g. road embankments, small ditches) were misclassified as channels. Neural network performance was improved markedly by trimming the area of interest to a 100-m wide buffer along the stream, eliminating many areas prone to misclassification.
The two methods provide different ways to effectively leverage high-resolution LiDAR datasets to gain information about channel morphology. These methods are a significant step forward as they can delineate bankfull elevation, as well as bankfull width, and operate using morphology alone. HydXS is an objective method that doesn’t require training, can be run on consumer-level hardware, and can perform well in small streams, but requires manual work to develop the necessary spatial framework of an accurate channel centerline. The neural network model is a promising method to delineate larger channels (>20 m wide) without requiring detailed centerline or cross-section data, given adequate training data for the stream type of interest (i.e. expert-delineated bankfull channel extents). We envisage that further improvement of the neural network method is possible by scaling the input image extents to catchment area, and training on a larger dataset from multiple regions to increase generalizability.
How to cite:
Russell, K., Garber, J., Thompson, K., Kunapo, J., Burns, M., and Zhang, G.: Automated tools for identifying bankfull river channel extents: developing and comparing objective and machine-learning methods, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-6681, https://doi.org/10.5194/egusphere-egu22-6681, 2022.
Despite the long record of applications and the well-known theoretical framework, geostatistical based image/surface texture tools have still not gained a wide diffusion in the context of geomorphometric analysis, even for the evaluation of surface roughness. Many geomorphometric studies dealing with various aspect of surface roughness use well-known approaches based on vector dispersion of normals to surface or even the popular Topographic Ruggedness Index. In many comparative studies on roughness metrics, geostatistical approaches are cited but not tested; in other studies, geostatistical approaches are tested using algorithms not adapted to the analysis of morphometric data. In remote sensing, geostatistical approaches are more popular, even if there is not a consensus on which are the most suited metrics for computing image texture indices. In metrology of manufactured surfaces, equipped by various industrial standards for surface texture measurements, approaches based on autocorrelation are widely adopted. However, “natural” surfaces and related morphogenetic factors are much more complex than manufactured surfaces and ad-hoc concepts and algorithms should be devised. This presentation is mainly focused on topographic surface analysis, but the considerations and results are applicable also in the context of image analysis. This presentation aims to clarify some aspects of the geostatistical methodologies, highlighting the effectiveness and flexibility in the context of multiscale and directional evaluation of surface texture. In doing this, the connections with other methodologies and concepts related to spatial data analysis are highlighted. Finally, it is introduced a simplified algorithm for computing surface roughness indices, which does not require the preliminary detrending of the input DEM.
ATKINSON, P.M. and LEWIS, P., 2000. Geostatistical classification for remote sensing: An introduction. Computers and Geosciences, 26(4), pp. 361-371.
BALAGUER, A., RUIZ, L.A., HERMOSILLA, T. and RECIO, J.A., 2010. Definition of a comprehensive set of texture semivariogram features and their evaluation for object-oriented image classification. Computers and Geosciences, 36(2), pp. 231-240.
GUTH, P.L., 2001. Quantifying terrain fabric in digital elevation models. GSA Reviews in Engineering Geology, 14, pp. 13-25.
HERZFELD, U.C. and HIGGINSON, C.A., 1996. Automated geostatistical seafloor classification - Principles, parameters, feature vectors, and discrimination criteria. Computers and Geosciences, 22(1), pp. 35-41.
TREVISANI, S., CAVALLI, M. and MARCHI, L., 2009. Variogram maps from LiDAR data as fingerprints of surface morphology on scree slopes. Natural Hazards and Earth System Science, 9(1), pp. 129-133.
TREVISANI, S., CAVALLI, M. and MARCHI, L., 2012. Surface texture analysis of a high-resolution DTM: Interpreting an alpine basin. Geomorphology, 161-162, pp. 26-39.
TREVISANI, S. and ROCCA, M., 2015. MAD: Robust image texture analysis for applications in high resolution geomorphometry. Computers and Geosciences, 81, pp. 78-92.
TREVISANI, S. and CAVALLI, M., 2016. Topography-based flow-directional roughness: Potential and challenges. Earth Surface Dynamics, 4(2), pp. 343-358.
How to cite:
Trevisani, S.: Returning to geostatistical-based analysis of image/surface texture: from generalization to a basic one-click short-range surface roughness algorithm, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-6924, https://doi.org/10.5194/egusphere-egu22-6924, 2022.
Linda Grinberga, Armands Celms, Krisjanis Sietins, Toms Lidumnieks, Miks Brinkmanis-Brimanis, and Jolanta Luksa
With the development of remote sensing technologies the application of different geospatial models in research has become increasingly important. Terrain relief is the difference in elevation between the high and low points of a land surface, that is, the change in the height of the ground over the area. Terrain relative relief (or elevation) is the relative difference in elevation between a morphological feature and those features surrounding it (e.g. height difference between a peak and surrounding peaks, a depression and surrounding depressions etc.). Together with terrain morphology, ppland other terrain attributes, it is useful for describing how the terrain affects intertidal and subtidal processes.
Appropriate decision-making tools are required for urban and rural planning, design and management. The usage of DEM (Digital Elevation Model), DSM (Digital Surface Model) and DTM (Digital Terrain Model) helps researchers and designers to analyse issues connected with drainage, geology, earth crust movements, sound and radio-wave distribution, wind effects, exposure to sun, etc. Analysis of the future scenarios of geospatial models has an essential role in the field of water management and various environmental topics. This research aims to focus on the environmental issues in a context of water quality and hydrology.
How to cite:
Grinberga, L., Celms, A., Sietins, K., Lidumnieks, T., Brinkmanis-Brimanis, M., and Luksa, J.: The Application of Relief Models for Environmental Solutions: Review, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-7860, https://doi.org/10.5194/egusphere-egu22-7860, 2022.
Siddhartho Shekhar Paul, Eliza M. Hasselquist, William Lidberg, and Anneli M. Ågren
High-resolution Light Detection and Ranging (LiDAR) data provide unique opportunities for landscape-scale mapping of hydrological features. LiDAR-derived digital elevation models are particularly valuable for identifying channel networks in densely forested landscapes, where satellite imagery-based mapping approaches are challenged by forest canopies. Artificial drainage practices have caused widespread alteration of northern landscapes of Europe and North America which likely have had significant impacts on hydrological connectivity and ecosystem functioning. However, these artificial channels are rarely considered in ecosystem management and poorly represented in existing geomorphological datasets. In this study, we conducted a landscape-scale analysis across 11 selected study regions in Sweden using LiDAR data for the virtual reconstruction of artificial drainage ditches to understand the extent of their ecological impacts.
We utilized a 0.5 m resolution digital elevation model for mapping natural channel heads and artificial ditches across the study regions. We also implemented a unique approach by back-filling ditches in the current digital elevation model to recreate the prehistoric landscape. This enabled us to map and model the channel networks of prehistoric (natural) and current (drained) landscapes. We found that 58% of the prehistoric natural channels had been converted to ditches. Moreover, the average channel density increased from 1.33 km km‑2 in the prehistoric landscape to 4.66 km km-2 in the current landscape, indicating substantial ditching activities in the study regions.
Our study highlights the need for accurate delineation of natural and artificial channel networks in northern landscapes for effective ecosystem restoration and management. We presented an innovative technique for comparing the channel networks between the prehistoric natural landscape and current modified landscape by integrating advanced LiDAR data, extensive manual digitization, and modeling; a highly suitable combination for channel network mapping in dense forest landscapes. The developed methodology can be implemented in any landscape for understanding the extent of human modification of natural channel networks to guide future environmental management activities and policy formulation.
How to cite:
Paul, S. S., Hasselquist, E. M., Lidberg, W., and Ågren, A. M.: Mapping of natural and artificial channel networks in forested landscapes using LiDAR data to guide effective ecosystem management, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-8728, https://doi.org/10.5194/egusphere-egu22-8728, 2022.
Martina Burnelli, Massimiliano Alvioli, Laura Melelli, and Alessia Pica
Ecodiversity stems from the interaction between the biosphere and the geosphere, and it is one of the necessary conditions for achieving a sustainable planet. Thus, the relationship between geodiversity and biodiversity should be clearly defined. The relationship between climate and topography in roughened mountain areas at low-latitudes, as constrains for the high values of biodiversity, has already been established. As a consequence, topography is the first and most important input parameter for investigating the connections between abiotic and biotic variety. Spatial analysis in a GIS framework is the key approach to better understand the role of topographic and hydrographic variables in evaluating geodiversity (geomorphodiversity) .
In this paper we focused on analyzing urban areas, where in 2030 60% of the world's population is expected to live. A science of cities is the future challenge for Earth Sciences: urban geomorphology could be the key to have a complete overview on the abiotic and biotic parameters in sustainable cities. To achieve this aim, the conservation of urban biodiversity is fundamental. Analysing the correlation between substantial geodiversity and biodiversity may be a guideline for science of cities and for designing and managing sustainable urban areas.
These ideas, if transposed in an urban context, should go beyond morphometric analysis of topography and take into account anthropogenic features and natural landforms modified by humans in time. To this end, geomorphological mapping is fundamental to calibrate the quantitative models in a truly multidisciplinary approach to a science of cities and urban biodiversity. We consider our contribution as a new model for the analysis of geodiversity in urban areas.
How to cite:
Burnelli, M., Alvioli, M., Melelli, L., and Pica, A.: Geodiversity as a key component for the evaluation of urban biodiversity, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-9650, https://doi.org/10.5194/egusphere-egu22-9650, 2022.
Sara Cucchiaro, Laura Carretta, Paolo Nasta, Federico Cazorzi, Roberta Masin, Nunzio Romano, and Paolo Tarolli
One of the main environmental threats to sustainability and crop productivity in the agricultural sector is soil erosion. For the mitigation of this problem in agricultural fields, no-till management is considered a key approach. The measurement of soil erosion is particularly challenging, especially when surficial morphological changes are relatively small. Conventional experiments are commonly time-consuming and labour-intensive in terms of both field surveys and laboratory methods. However, the Structure from Motion (SfM) photogrammetry technique has enhanced the experimental activities by enabling the temporal evolution of soil erosion to be assessed through detailed micro-topography. This work presents a multitemporal quantification of soil erosion, using SfM through Uncrewed Aerial Vehicles (UAV) survey for understanding the evolution of no-till (NT) and conventional tillage (CT) in experimental plots. Considering that plot-scale soil surface (mm grid size) by several orders of magnitude, it was necessary to minimise SfM errors (e.g., co-registration and interpolation) in volumetric estimates to reduce noise as much as possible. Therefore, a methodological workflow was developed to analyse and identify the effectiveness of multi-temporal SfM-derived products, e.g. the conventional Difference of Digital Terrain Models (DoDs) and the less used Differences of Meshes (DoMs), for soil volume computations. To recognise the most suitable estimation method, the research validated the erosion volumetric changes calculated from the SfM outputs with the amount of soil directly collected through conventional runoff and sediment measurements in the field. This study presents a novel approach for using DoMs instead of DoDs to accurately describe the micro-topography changes and sediment dynamics. Another key and innovative aspect of this research, often overlooked in soil erosion studies, was to identify the contributing sediment surface, by delineating the channels potentially routing runoff directly to water collectors. The sediment paths and connected areas inside the plots were identified using a multi-temporal analysis of the sediment connectivity index for achieving the volumetric estimates. The DoM volume estimates showed better results with respect to DoDs and a mild overestimation compared to in-situ measurements. This difference was attributable to other factors (e.g., the soil compaction processes) or variables rather than to photogrammetric or geometric ones. The developed workflow enabled a very detailed quantification of soil erosion dynamics for assessing the mitigation effects of no-till management that can also be extended in the future to different scales with low-costs, based on SfM and UAV technologies.
How to cite:
Cucchiaro, S., Carretta, L., Nasta, P., Cazorzi, F., Masin, R., Romano, N., and Tarolli, P.: Assessment of soil erosion induced by different tillage practices through multi-temporal geomorphometric analyses, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-2000, https://doi.org/10.5194/egusphere-egu22-2000, 2022.
State-of-the-art applications in various earth science domains shows that different classification methods are playing an increasingly important role in mapping due to their improving accuracy. However, in the field of geological mapping, the exclusive use of morphometric and spectral indices in classification models are still often considered as subsidiary mapping tools. This is particularly true in areas where the surface is covered by vegetation and the soil layer is relatively thick, since in such places geological structures can only be observed at first hand at rock outcrops. The aim of our research is to investigate the automatic mapping of rock outcrops in the Dorog Basin in Hungary, where outdated geological maps are currently being updated. In this research, we applied the random forest classification combined with a wider range of input data including satellite imagery and ecosystem information.
The Dorog Basin, located in northern central Hungary, has a medium-density settlement network, with built-up and cultivated areas alternating with areas of wooded or scrub-covered terrain with rugged topography. The region is tectonically fragmented, where former fluvial erosion is of great importance. In several cases the Mesozoic carbonates, Paleogene limestones or limnic coal sequences outcrop the Quaternary sediments resulting a diverse, although a well identifiable surface. In the 86.86 km2 study area, the input of the model included 14 morphometrical raster layers derived from SRTM-1, six raster layers with mineral indices derived from Sentinel II, and one ecosystem layer , all set to a uniform ~25m resolution. To test the performance of random forest classification in modelling pre-Quaternary formations, we applied two different approaches. In the first one, we used conventional training areas to model pre-Quaternary outcrops, as well as we modelled the physical characteristics of the surface formations. Whereas in the second one, we modelled the pre-Quaternary outcrops and physical characteristics of the surface formations by using randomly selected zones on the study area with around 6000-10000 random training polygons. The randomly generated training polygons were circles of about 1-2 pixels in size around points. The training areas were derived from the former geological map of the Dorog Basin . The importance of input parameters were also observed for further use. A six-fold cross-validation of the selected training areas showed that the two methods were equally accurate, but the automatic processing of randomly selected training areas was faster.
Based on the modelling results, the pre-Quaternary rock outcrops of the area can be determined with at least 80% confidence using random forest classification. These results will be used in future field mapping, which will also provide a field validation of the method.
From the part of G.A. financial support was provided from the NRDI Fund of Hungary, Thematic Excellence Programme no. TKP2020-NKA-06 (National Challenges Subprogramme) funding scheme.
 Ecosystem Map of Hungary. DOI: 10.34811/osz.alapterkep
 Gidai, L., Nagy, G., & Sipass, S. (1981). Geological map of the Dorog Basin 1: 25 000. [in Hungarian] Geological Institute of Hungary, Budapest.
How to cite:
Pogácsás, R. and Albert, G.: Automatic detection of rock outcrops on vegetated and moderately cultivated areas, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10469, https://doi.org/10.5194/egusphere-egu22-10469, 2022.
Conor McDowell, Helm Carina, Reid David A., and Hassan Marwan
Remotely piloted aircrafts (UAVs) and Structure-from-Motion photogrammetry (SfM) have become a widely used approach for producing high-resolution topographical measurements of river systems. This approach has the benefit of capturing data over large spatial scales while requiring little time in the field. In small, forested rivers, the dense canopy has hindered the use of remote sensing techniques, limiting topographic data collection to more time-consuming and lower-resolution methods. This complicates monitoring the response of these systems to individual floods, as in many situations there is not enough time to complete more time-consuming surveys between events.
In this study, we pilot the use of sub-canopy UAV surveys (flown at 1-3 m altitude) to monitor the response of a small mountain stream (1-3 m wide) in British Columbia to a sediment pulse generated by the removal of an upstream culvert. Using eleven surveys flown over a three-year period, we track the downstream propagation of the pulse and the subsequent responses in bed topography and roughness along the 240 m reach. We observe a “build-and-carve” response of the channel, where some channel segments aggrade during the first floods after pulse generation, whereas others undergo little morphologic activity. In subsequent floods, these aggradational segments rework through the carving of well-defined channels that release this aggraded sediment downstream. These “build-and-carve” segments serve as temporary storage reservoirs that caused the pulse to fragment as it progressed downstream. The locations of these storage reservoirs were set by the initial channel morphology and the movement of in-stream wood and debris. This study highlights the importance of temporary sediment storage reservoirs for fluvial morphodynamics and provides some insights and suggestions for the future monitoring of forested river systems using sub-canopy drone surveys.
How to cite:
McDowell, C., Carina, H., David A., R., and Marwan, H.: Response of a small mountain river to a sediment pulse tracked using sub-canopy UAV surveys, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10675, https://doi.org/10.5194/egusphere-egu22-10675, 2022.
Anshita Srivastava, Ashutosh Tiwari, Avadh Bihari Narayan, and Onkar Dikshit
Advancements in processing strategies of time series interferometric synthetic aperture radar (InSAR) has resulted in improved deformation monitoring and DEM generation. Both of the applications use phase unwrapping, which involves finding and adding the unknown correct number of phase cycles to the wrapped phase. It is an inverse process of recovering the absolute phase from the wrapped phase, and the objective is to remove the 2π-multiple ambiguity. Ideally, it could be achieved by addition or subtraction of 2π at each pixel depending on the phase difference between the neighboring pixels. The problem appears effortless but brings challenges due to noise and inconsistencies. The conventional methods require improvements in terms of accurately estimating the unknown number of phase cycles and dealing with phase jumps. Recently, deep learning methods have been used extensively in the domain of remote sensing to solve complex image processing problems such as object detection and localization, image classification, etc. Since all the pixels in a stack of interferograms are not used in unwrapping, and the pixels used are scattered irregularly, modeling the unwrapping problem as an image classification problem is infeasible. In this work, we deploy Graph Neural Networks (GNNs), a class of deep learning methods designed to infer information from input graphs to solve the unwrapping problem. Phase unwrapping can be posed as a node classification problem using GNN, where each pixel is treated as a node. The method is aimed to exploit the capability of GNNs in correctly predicting the phase count of each pixel. The proposed work aims to improve the computational efficiency and accuracy of the unwrapping process, resulting in reliable estimation of displacement.
How to cite:
Srivastava, A., Tiwari, A., Narayan, A. B., and Dikshit, O.: InSAR phase unwrapping using Graph neural networks , EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-11010, https://doi.org/10.5194/egusphere-egu22-11010, 2022.
Understanding the mechanism of fault rupture is important to minimize earthquake damage and to estimate the impacts of future earthquakes. In this study, we observed surface displacements caused by the Hovsgol earthquake (Mw 6.7) in January 2021 using three Differential Interferometric SAR (DInSAR) pairs of Sentinel-1B at descending node and ALOS-2 at ascending and descending nodes, and then estimated the source parameters of the earthquake by the inversion of the observed displacement fields. The maximum surface displacement in the radar look direction was 21 cm at the Sentinel-1 descending node, and 32 cm and 26 cm at the ALOS-2 ascending and descending node, respectively. All differential interferograms showed three fringe patterns near the epicenter, which suggests that there were three rupture planes with different slips. We performed the inversion modeling of the DInSAR-observed surface displacements assuming three rupture planes with different slip magnitudes and directions. The values of normalized root mean square error (NRMSE) between the modelled and observed displacements were smaller than 4% for all DInSAR observations. The spatial distribution of modelled displacements was matched to the observed one. The source parameters of fault estimated by the inversion were closely consistent with the measurements by United States Geological Survey and Global Centroid Moment Tensor. The inversion results demonstrated that the assumption of our inversion modeling (three rupture planes) is reasonable.
How to cite:
Kim, T. and Han, H.: Source parameters of the 2021 Hovsgol earthquake (Mw 6.7) in Mongolia estimated by using Sentinel-1 and ALOS-2 DInSAR, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-11152, https://doi.org/10.5194/egusphere-egu22-11152, 2022.
Monitoring land use and land cover change (LULCC) is one of the best methods to understand the interactive changes of agriculture, climate change, and ecological dynamics. In eastern Asia, Taiwan is characterized by high population density, rich biodiversity, and complex terrain. However, recent climate change has impacted the people and ecosystems in Taiwan. Therefore, we applied landscape metrics and the deep learning U-net semantic segmentation model to enhance the remote sensing images based LULCC monitoring efficiency and take a case study in suburban areas of central Taiwan, a place that plays an important economic role in Taiwan occupied with intensive agricultural activities.
This study focuses on six townships in Nantou County in Central Taiwan, where the major agricultural products are rice, tea, and fruit. We obtained four dates of Sentinel-2 images in February for 2018 and 2021 and classified the landscape into five classes: agricultural, forest, built-up, free water bodies, and bare land. The spectral bands information (Blue, Green, Red, NIR), the normalized difference vegetation index (NDVI), and soil-adjusted vegetation index (SAVI) were obtained for establishing the deep learning U-net semantic segmentation model. The accuracy and the loss function of the training model results are 0.89 and 0.02, respectively. In addition, the ground truth data was consulted with the official land-use classification information and the high spatial resolution imagery in Google Earth Pro. Finally, we analysed the classified images' results to detail the study area's changing trajectory to explore the complex spatiotemporal landscape patterns.
3. RESULTS AND DISCUSSIONS
According to the result, the forest area on the eastern side accounts for more than 70% of the study area. The construction area and the agricultural area have an upward trend during the research period (16% and 5%); in addition, except for the number of patches in free water bodies decreased, all other categories had an upward trend, especially the construction and agricultural area are the largest. The Shannon's Evenness Index reflects that all patches are evenly distributed in space and the area-weighted average fractal dimension index decreases reflecting possible influences of anthropogenic activities. Thus, the results indicate an increasing level of fragmentation, supported by the decrease of the area-weighted average fractal dimension index. In conclusion, using satellite imagery with the deep learning U-net semantic segmentation model can sufficiently discern a detailed LULCC. Furthermore, with the combination of landscape matrix information, the interactions between humans and the environment can be understood better quantitatively.
Huete, A. R., Hua, G., Qi, J., Chehbouni, A., & Van Leeuwen, W. J. D., 1992: Normalization of multidirectional red and NIR reflectances with the SAVI. Remote Sensing of Environment, 41(2-3), 143-154.
Ronneberger, O., Fischer, P., & Brox, T., 2015: U-net: Convolutional networks for biomedical image segmentation. In International Conference on Medical image computing and computer-assisted intervention (pp. 234-241). Springer, Cham.
Rouse, J. W., Haas, R. H., Schell, J. A., Deering, D., Deering, W. 1973: Monitoring vegetation systems in the Great Plains with ERTS, ERTS Third Symposium, NASA SP-351 I, pp. 309-317.
How to cite:
Zhuang, Z.-H. and Tsai, H. P.: Application of Deep Learning Model to LULCC Monitoring using Remote Sensing Images-A case study in suburban areas of central Taiwan, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-11764, https://doi.org/10.5194/egusphere-egu22-11764, 2022.
Hesam Ahmady-Birgani, Parisa Ravan, Zhengyi Yao, and Gabriela Mihaela Afrasinei
To enhance the understanding of aeolian landforms and their processes, the assessment of origin, migration and evolution of newly-born sand dunes is vital. In this regard, Lake Urmia, in NW Iran, was considered as a representative case study, given that it has lost approximately two-thirds of its water volume in the past two decades and, consequently, the newly-born sand ridges and sand dunes on its western shores were formed. The emerging sand dunes are located close to the villages, adjacent to the agricultural and farmlands, international transit road, and industrial zone, encompassing the whole area. The present study aims to assess the sand dunes’ origin and their migration both in speed and direction in the past decade.
To understand the questions above, remote sensing techniques and in-field studies were coupled. Therefore, wind data from the closest meteorological station were employed to calculate the wind rose, drift potential (DP), the resultant drift potential (RDP), and the resultant drift direction (RDD) across the region. Change detection techniques using high-resolution satellite images were chosen to detect the migration rate and morpho-dynamic changes of Lake Urmia sand dunes. To classify the geomorphological features and land uses in the region, a hybrid supervised classification approach including a customised decision tree classifier was used to distinguish sand dune units from other signatures. Using the minimum bounding geometry method, feature classes were created. These feature classes represent the length, width, and orientation of sand dunes, retrieved after the image classification process. Also, fieldwork surveying was carried out on the sixteen sand dunes in different periods to measure the morphological and evolutionary changes.
As the wind results show, the trend of DP parameters between the years 2006-2009 and the years 2015-2020, the percentage of wind speeds above the threshold velocity (V>Vt%) to DP has significant gaps, suggestive of weaker winds in those periods. However, between the years 2009-2015, the V>Vt% and DP values are corresponding and coequal. This indicates that the most erosive and shifting winds are between 2009-2015, with the weakest wind power in tails. Moreover, the annual variability of DPt is well correlated with Lake Urmia water level changes; but there is no correlation between the DPt and precipitation amount. The evaluation of image processing results depicted that after 2003, the area of sand dunes had dramatically increased. On average, the smallest area belongs to 2010 (287.3 m2), and the largest area is for years 2019 (775.96 m2), 2018 (739.08 m2), and 2017 (739.74 m2). In addition, between the years 2010 and 2014, a significant increase in area of the sand dunes from 287.25 to 662.8 m2 was observed. The migration rate is the highest between 2010 and 2015, with the lowest values before 2010 and after 2015.
The results of this study have broad implications in the context of sustainable development and climate-related challenges, ecosystem management and policy-making for regions with sand dune challenges, hence crucial insights can be gained by coupling remote sensing techniques and in-situ studies.
How to cite:
Ahmady-Birgani, H., Ravan, P., Yao, Z., and Afrasinei, G. M.: Newly-Born Sand Dunes of Lake Urmia: Assessing Migration Rate and Morphodynamic Changes Using Remote Sensing Techniques and Field Studies, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-12041, https://doi.org/10.5194/egusphere-egu22-12041, 2022.
The supervised mapping of landforms last years got high levels based on classic classification methods and new artificial intelligence techniques. However, it is often difficult to create train data for large and diverse areas, and we can face up with differences between expert-to-expert landforms interpretation. It can be solve using unsupervised classification - a less effective in general case, but more objective. The way to make more effective classification - to create special input variables (to account local specificity of landforms) aimed to show real terrain structure. Study region - Yamal Peninsula (Arctic coast of Russia), covered sea accumulative and erosional plains, reshaped by some cryogenic processes, especially thermokarst, with many lake hollows. We used ArcticDEM 32m and decomposition of DEM with 2D FFT by moving windows with sequence of sizes from 1.5 to 3 km (by the interval of 0.3 km) and with lag around 150 m (overlapping - 90-95 %). The 9 variables were computed: 1) magnitude of the main wave in the height field, 2) wavelength of the main wave, 3) importance (share of the height variation) of the fix pool of biggest harmonic waves, 4-6) orthogonal (N-S and W-E) components of the general direction of the height fluctuations (and the significance of the direction), 7-9) coefficients of the exponential trend equation for approximation wave's frequencies/magnitudes distribution. We then trained the model of landforms clustering for the study area using Kohonen network and the hierarchic clustering was used for additional generalization. The medium-scale (750 m / pix, it is matched to maps at the scale 1:500 000 - 1:1 000 000) map of Yamal Peninsula landforms was created. Seven classes of landforms were recognized. The study was supported by Russian Science Foundation (project no. 19-77-10036).
How to cite:
Kharchenko, S.: Medium-scale unsupervised landform mapping of the Yamal Peninsula (Russia) using 2D Fourier decomposition of the ArcticDEM, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-12383, https://doi.org/10.5194/egusphere-egu22-12383, 2022.
Please use the buttons below to download the presentation materials or to visit the external website where the presentation is linked. Regarding the external link, please note that Copernicus Meetings cannot accept any liability for the content and the website you will visit.
We are sorry, but presentations are only available for users who registered for the conference. Thank you.
Please decide on your access
Please use the buttons below to download the presentation materials or to visit the external website where the presentation is linked. Regarding the external link, please note that Copernicus Meetings cannot accept any liability for the content and the website you will visit.