Transition from manual (visual) interpretation to fully automated gully detection is an important task for quantitative assessment of modern erosion, especially when it comes large mapping areas. Existing approaches semi-automated are based on either object-oriented selection multispectral images or a probabilistic model obtained using digital elevation models (DEMs). These cannot be used the erosion territory European part Russia most affected by due lack national large-scale DEM and...

This study addresses the challenge of accurate crop detection using satellite data, focusing on application Long Short-Term Memory (LSTM) networks. The research employs a “spatial generalization” approach, where model trained one geographic area is applied to another with similar vegetation conditions during growing season. LSTM networks, which are capable learning long-term temporal dependencies, used overcome limitations traditional machine techniques. results indicate that although more...

This study focuses on the Kuibyshev reservoir (Volga River basin, Russia)—the largest in Eurasia and third world by area (6150 km2). The objective of this paper is to quantitatively assess dynamics bank landslides shoreline abrasion at active zones based integrated use modern instrumental methods (i.e., terrestrial laser scanning—TLS, unmanned aerial vehicle—UAV, a global navigation satellite system—GNSS) GIS analysis historical imagery. A methodology for application different assessment...

Evaluation of the vegetation and agricultural-management factor (C-factor) is an important task, solution which affects correct assessment intensity soil erosion. For vast area European part Russia (EPR), this task particularly relevant since no products allow taking into account C-factor. An approach based on automated interpretation main crop groups MODIS satellite imaging data from Terra Aqua satellites with LSTM machine-learning method was used to achieve goal. The accuracy group...

Unmanned aerial vehicles (UAV) have long been well established as a reliable way to construct highly accurate, up-to-date digital elevation models (DEM). However, the territories which were modeled by results of UAV surveys can be characterized very local. This paper presents surveying Sarycum area Dagestan Nature Reserve Russia with an 15 sq. km using DJI Phantom 4 UAV, methodological recommendations for conducting work on such large territory. As result this work, DEM 0.5 m resolution...

This study examines the application of unsupervised classification techniques in mapping forest vegetation, aiming to align vegetation cover with Braun-Blanquet system through remote sensing. By leveraging Landsat 8 and 9 satellite imagery advanced clustering algorithms, specifically Weka X-Means, this research addresses challenge minimizing researcher subjectivity mapping. The methodology incorporates a two-step approach accurately classify communities, utilizing comprehensive set indices...

In a changing climate, forest ecosystems have become increasingly vulnerable to continuously exacerbating heat and associated drought conditions. Climate stress resilience is governed by complex interplay of global, regional, local factors, with hydrological conditions being among the key players. We studied Scots pine (Pinus sylvestris L.) ecosystem located near southern edge boreal ecotone, which particularly subjected frequent prolonged droughts. By comparing dendrochronological series...

Increased anthropogenic pressure including intensification of agricultural activities leads to long-term decline natural biotopes, with planted forests often considered as promising compensatory response, although reduced biodiversity and ecosystem stability represent their common drawbacks. Here we present a complex investigation the impact large Grey Heron ( Ardea cinerea L.) colony on soil biogeochemistry vegetation in Scots pine forest representing situ experiment an engineered...

Abstract. Despite the large variety of methods for estimating slope erosion intensity, it is still difficult to obtain accurate rates. Therefore, our goal was develop a method provide estimates sheet and rill intensities, evaluate denudation quantities due abrasion, landslides talus processes using high-precision laser scanning system (Trimble® GX). Differential maps between all stages surveying TIN-models were built directly on point clouds in "Trimble® RealWorks" software. Inspection...

Digital elevation models (DEM) are an important source of information on the terrain, allowing researchers to evaluate various exogenous processes. The higher accuracy DEM better level work possible. An data for construction DEMs point clouds obtained with terrestrial laser scanning (TLS) and unmanned aerial vehicles (UAV). In this paper, we present results constructing a small catchments using UAVs. Estimation UAV showed comparable TLS if real time kinematic Global Positioning System...

Unmanned Aerial Vehicles (UAVs) are increasingly involved in surveying work, becoming a reliable basis for information on three-dimensional terrain features. Until now, ground reference points have been used to provide planimetric evidence measurements from drones. Their placement and coordinates measurement takes quite long time, which increases proportionally the area under study. In addition, use of these marks produces model distortions (especially over large areas 1 sq. km or more),...

Abstract. This work is continuation of the research aimed at developing a system for analysing risks destruction archaeological objects Volga-Bulgaria period (X–XIII centuries AD) on territory Republic Tatarstan. Most settlements located small rivers banks. fact determining risk their by natural processes. In order to identify transformation sites fluvial processes bank erosion dynamics Lukovskoe fortified settlement evaluated. Modern field survey with use UAV and GNSS methods applied study...

Soil erosion worldwide is an intense, poorly controlled process. In many respects, this a consequence of the lack up-to-date high-resolution maps. All over world, problem insufficient information solved in different ways, mainly on point-by-point basis, within local areas. Extrapolation results obtained locally to more extensive territory produces inevitable uncertainties and errors. For anthropogenic-developed part Russia, especially urgent because assessment intensity processes, even with...

АннотацияВ статье представлены результаты оценки точности четырех глобальных цифровых моделей рельефа

Introduction . Nonparametric Bayesian networks are a promising tool for analyzing, visualizing, interpreting and predicting the structural dynamic characteristics of complex systems. Modern interdisciplinary research involves processing heterogeneous data obtained using sensors various physical nature. In study forest fund, both methods direct dendrological measurements remote observation unmanned aerial vehicles widely used. Information these must be analyzed in conjunction with...

Accurately identifying the boundaries of agricultural land is critical to effective management its resources. This includes determination property and rights, prevention non-agricultural activities on land, natural There are various methods for accurate boundary detection, including traditional measurement remote sensing, choice best method depends specific objectives conditions. paper proposes use convolutional neural networks (CNNs) as an efficient tool automatic recognition boundaries....

This study uses modern and historic spatial data to analyze land use around 13th Century AD Bulgarian fortified settlements examine the current state of these features how changes in over past two centuries have affected features. Historic maps from late 18th early 19th were used alongside Landsat 8 images 2015–2018 as a source information about use. Based on results archival map analysis classification satellite imagery, interpretation was possible. method generated new dynamics near...

The paper presents the results of studies medieval archaeological sites, Laishevsky and Ostolopovsky settlements, hillfort, using multitemporal remote sensing data modern field research methods. studied sites are located in zone active bank transformation have been destroyed since creation Kuibyshev reservoir. To assess dynamics coastline, were used. Shoreline positioning for historical period 1950s was interpreted from archival aerial images (Kazan University Library). 1975 satellite...

Abstract In a changing climate, forest ecosystems become increasingly vulnerable to the continuously exacerbating heat and drought stress conditions. Climate resilience is governed by complex interplay of global, regional local factors, with hydrological conditions among key roles. Using modified detrended partial cross-correlation analysis (DPCCA), we analyse interconnections between long-term tree-ring width (TRW) data climate variations at various scales time lags. By comparing...

For the territory of Republic Tatarstan (Russia), a spatial model forest communities was built as basis for assessing sequestration potential ecosystems. Combination multi-temporal and multi-spectral satellite imagery from Landsat 8 9 platform input data Google Earth Engine cloud were used. The set 292 vegetation indices metrics computed pre-processed imagery, combined into dataset. Weka X-Means clustering algorithm trained applied to study area. unsupervised classification carried out by...

To assess the carbon sequestration potential of terrestrial ecosystems Republic Tatarstan, vegetation classification was carried out at class level using Braun-Blanquet system. 17,000 relevés from "Flora" database were used as source materials. The EuroVeg Checklist expert system in Juice 7.0 software. A total 34 classes identified. 14 largest have been selected for cover modelling.