ABSTRACT. Today, melting ice sheets, rising sea levels, and other climate hazards are of particular scientific interest. Satellite observations are vital for monitoring and tracking long-term environmental changes in polar and sub-polar regions - the Arctic, Antarctica, Iceland, Greenland, Faroe Islands, sub-Arctic and sub-Antarctic areas, where climate change is particularly prominent. Thanks to the satellite images accumulated over several decades, the study and tracking of the interrelationships of the modern complex processes taking place in the environment for these areas are becoming more and more thorough, and their explanation - is scientifically based. The increasing availability of satellite imagery due to rapid advances in remote sensing technology is expanding the horizon for selecting image sources. For polar regions, satellite images in different spectral ranges from low to high-orbit satellites date back to the early 1990s. A wide range of scientific communities uses them for various studies and applications. This paper analyzes the availability of space-based imagery for Livingstone Island, located in the South Shetland Islands of the Antarctic Peninsula. The images have different spectral, spatial, radiometric, and temporal characteristics, making it possible to use them for various studies of these regions. Data from the ESA Climate Office providing access to various datasets published by the Antarctic Ice Sheet ECV project, NASA Image Galleries, and other sources are used in the analysis. Satellite images with high and very high spatial resolution are summarized for the studied region, and the available open and commercial datasets are provided and compared. A systematic review of specific characteristics and time ranges of satellite imagery provided by space agencies and private companies was performed to provide insight into their practical use for change detection studies. Based on the analysis and the cost-based ranking, some recommendations are given for using satellite imagery in future studies of this area.

ABSTRACT. An improved concept design for the PRISMA Second Generation (PSG) mission for hyperspectral remote sensing is proposed, focusing on enhancing the satellite's decision autonomy through on-board edge-computing with machine learning. PSG, the future hyperspectral Italian constellation, leverages the first PRISMA mission for data continuity and spatial resolution enhancement. Financed by the project “ComeOnBoardPSG!” supported by the Italian Space Agency, this research aids in developing new Earth observation missions.

The project aims to enhance platform autonomy for an optimized acquisition plan of the primary hyperspectral payload. One avenue involves potential integration of an additional forward-looking camera with RGB bands positioned between nadir and satellite velocity. The presence of the secondary camera is under feasibility study; it could enable Real-Time Cloud Detection, facilitating cloud coverage assessment for informed decisions on future acquisitions.

Furthermore, the project explores the feasibility of equipping the main payload with Machine Learning-based real-time Fire Detection capabilities. This would enable the satellite to detect wildfires and high-temperature phenomena, providing timely warnings to support ground operations during natural disasters.

Data from additional and main cameras are processed by a dedicated edge-computing devices such as graphic or vision processing unit, or system-on-chip with field programmable gate array. Preliminary results show promising fire detection with PRISMA hyperspectral imagery, using convolutional neural networks trained on datasets from Australia and Oregon. Edge-implementation of these networks using an Nvidia Jetson TX2 demonstrates the suitability for space missions providing accuracies up to 98%, inference times of about 3.0 ms, and average computing power 4.8 W. The cloud detection component is still pending implementation but leverages the experience from previous successful missions such as the Phisat-1.

This work describes the concept design studied with the ComeOnBoardPSG! Project for an improved PSG platform supported by on-board edge-computing, showing the achievable improvements in terms of optimized data management and responsiveness in disaster management.

ABSTRACT. Mie lidar has been profoundly applied in the retrieval of aerosol optical coefficients in vertical distribution. However, few studies further explore the strategies for the retrieval of aerosol mass profiles quantitively from lidar observation. To meet the rising demand of aerosol mass concentration in spatial and temporal distribution, an iteration algorithm for profiling aerosol mass composition as well as extinction coefficient based on spaceborne dual-wavelength lidar data is proposed. By constructing the relationship between mixed aerosol mass profiles and optical properties at different wavelengths on the basis of Mie theory, new constraints are induced to im-prove the accuracy of lidar ratio, which is essential for the retrieval of aerosol extinction coefficients by solving the lidar equation. Meanwhile, aerosol composition profiles can also be deduced based on the a prior estimation of aerosol compositions and intrinsic optical features of the aerosols. This method is first applied in simulated data with wavelengths at 532 nm and 1064 nm based on the reanalysis data of aerosol mass concentration profiles in a study region of Inner Mongolia, China. Comparing with the retrieval results from the Fernald method with empirical estimated lidar ratio, the results of the proposed method yield mean accuracy improvement of integrated aerosol extinction coefficients in 532 nm and 1064 nm channels by 19.70% and 3.41%, and column dust and sulfate aerosol by 12.90% and 20.73%, respectively. The retrieval model is further validated by CALIOP data, suggesting modified extinction coefficient and informative aerosol composition results.

ABSTRACT. When classifying hyperspectral images, ensemble learning techniques are frequently employed to increase the reliability and accuracy of the classification models. Data is captured over numerous narrow and contiguous spectral bands in hyperspectral imaging, resulting in a high-dimensional dataset. The difficulties presented by this kind of data, like as noise, excessive dimensionality, and class imbalance, can be surmountable with the aid of ensemble methods. Ensemble learning is the process of combining numerous base classifiers to generate a more powerful and accurate model. The basic idea underlying ensembles is to take advantage of the variety among base models, which can lead to improved generalization and reduced overfitting. In this study Bagging type of ensemble is the process of separately training numerous base classifiers using various, arbitrary subsets of the training data. The boosting type sequentially focuses on training base classifiers while increasing the weights assigned to the incorrectly classified data points by earlier classifiers. Salinas dataset is used for hyperspectral image classification through bagging and boosting ensemble classifiers such as Random Forests, Support Vector Machines, XGBoost and Light GBM. The best F1-score is yielded using the XGBoost algorithm with approximately %3 than the second and the worst results are obtained using the Random Forest algorithm. The results are discussed under the f1-score, overall accuracy and precision recall metrics on different classes. At the end of the experiments classification results are visualized as thematic maps comparatively.

ABSTRACT. Public space lighting is an important part of urban life, which directly affects citizen's perception and activities of urban space. However, excessive and misguided spatial lighting can affect human health, astronomical observations, and the ecological environment, leading to energy waste. With the development of human-centered thinking and sustainable lifestyles, people are increasingly concerned about the comfort of their living environment. As one of the most intuitive living environments, the urban light environment has put forward higher requirements for the comfort of the light environment. How much night lighting is appropriate? How to evaluate public space lighting? Based on the above two questions, this paper presents a method for assessing the quality of urban public space lighting at night based on nighttime light (NTL) remote sensing and citizen participation. Field measurements and questionnaire surveys of night light environment cost a lot of time and manpower. Therefore, we explore the possibility of using NTL remote sensing data to retrieve ground light properties and further retrieve citizens’ perception of lighting, and then estimate potential light pollution. This study combines the most recent SDGSAT-1 glimmer imagery of the Beijing metropolitan areas with ground-measured light attributes and citizens’ perception data of nighttime light, a coupling model of citizens’ perception, measured light attributes and night-light remote sensing was established. Based on the NLT remote sensing inversion model and data, the map of citizens' feeling of safety and comfort of urban public space lighting in Beijing was drawn. Spatial differences exist in the citizens’ perception score of Beijing from the pixel-level citizens’ perception map. There are some areas, especially commercial and landscaped, in which light intensity is too high and leads to energy waste and unnecessary light pollution.

ABSTRACT. Archiving or digital preservation is a multidisciplinary concept that combines policies, strategies, and actions to ensure access to digital content over time. It is a critical aspect of data management, although it is often misunderstood or overlooked by data producers. This is because the benefits of digital preservation are usually realized in the long term, while the costs and efforts are usually counted only in the short term. However, as our reliance on digital data grows, so does the importance of digital preservation. At its core, digital preservation is about ensuring that data remains accessible and usable. This usually involves extracting data from various types of databases and converting it into an interoperable format that can be accessed and understood without the need for the original software or hardware. Such approach should also be applied to maps and GIS products to ensure availability of this type of data over long time. However, it´s not just about preserving the data itself, but about preserving the context and meaning of that data too. This is where metadata comes into play. It provides crucial information about the content, context, and structure of the data, making it a key component of digital preservation. All this generates high demands on both sides, archives and producers especially in the area of technical knowledge, skills, software and hardware, data throughput, data management and strategies in digital archives. The presentation presents a possible solution and best practices implemented in Czechia. The key part of the research is a reflection on the anticipated future use of geodata in relation to its long-term readability, trustworthiness, integrity and accessibility in various ways. In fact, it can be described as a craft rather than a science although the constant technological developments and the fragility of geodata in its complex form are increasingly an interdisciplinary scientific challenge.

ABSTRACT. Geotourism is one of today's most dynamically developing tourism sectors. What distinguishes it from other tourism branches is its ability to draw visitors' attention to the scientific, rather than merely aesthetic, value of spectacular geoscientific sites. It does so while emphasising their cultural and local importance. Hungary's most popular long-distance hiking trail is the National Blue Trail. As it mainly passes through hilly and mid-mountain areas, it invites hikers to take a journey through the Earth’s history, in addition to the abiotic and cultural values. Until now, the route’s geo-values have not been presented comprehensively and coherently, but the author of this year's GeoBlueTrail book has undertaken this task. However, the idea is also interesting from a cartographic point of view, as maps are an essential part of a book on tourism and geoheritage. In Hungary, only a few geotourism maps (according to the definition, i.e. they should include geological information) have been published so far. The GeoBlueTrail volume presents the geo-values of Hungary for 11 landscape units – 11 maps have been produced accordingly. The map editing was a difficult task, for lack of a domestic example. The generalisation of the large-scale geological content, the presentation of the cultural and geoheritage sites in the vicinity of the trail and reaching balance with other map elements were all the work of the authors. The result is 11 maps whose primary function is to put the written text into a geospatial context. They provide the non-scientist reader with an overview of the geological outlines of an area and an idea of the important cultural and geological sites within a section. In this manuscript, we aim to provide the professional community with the process and principles of geotourism map editing so that more and more of these science-promoting materials can be produced.

ABSTRACT. There are different conceptions, understandings and definitions of a model in the different scientific knowledge. One of them is related to a generalized representation of real objects and process for a certain purpose or study. Another understanding presents the model as an artificially created object in the form of a map, scheme, diagram, mathematical formulas, physical structures and algorithms for their processing, reproducing in a special way the construction and properties of the studied objects. The important element of the scientific models and model systems are opportunities for development of the projections and prognosis. These are real constructions for integration between science and processes in business and society activities.

ABSTRACT. The cadastral map preserves the dimensions and shape of an owned land parcel, but also the spatial relationship of all individual parcels to each other. Hundreds of thousands of cadastral maps all over the world are aided land assignment and taxation, presenting statistics essential for territorial administration, or as a symbol of state control over land. The historical land cadastre introduced in in the Austrian Empire, was the foundation of many countries' land register systems, including southern part of Poland annexed by Austria in the eighteenth century during the Partitions.

The study evaluates the use of old Austrian cadastral maps in southern Poland, mainly in matters related to property boundaries. After the Second World War cadastral maps lost its importance in Poland due to sociopolitical changes. After the 1989 transformation it became a source of data on real property boundaries once again.

The article presents numerous examples of modern uses of historical land cadastre maps including land tax registers in land surveying and legal procedures in Poland. Modern court proceedings on boundary disputes, easement appurtenant decisions, release of property, grant of property title, or usucaption often need to refer to historical legal sources, including the Austrian Cadastre to determine ownership boundaries.

The results show that data from the historical land cadastre remain a source of information when determining ownership boundaries despite the time passed. They should still be a subject of education for students as well as an ispiration for GIS specialists to combine cadastral maps with currently valid maps not only as historical scientific information but also as an information for determining property rights.

ABSTRACT. A major environmental and ecological concern in worldwide remains the changes of Land Use and Land Cover (LULC) due the increasing of urbanization. A crucial parameter ranging from climate change to urban and agriculture management and monitoring is Land Surface Temperature (LST). In this context, the present paper provides and explore the potential of Geoinformation Science and Systems (GISS) for calculating and analyzing the Land Surface Temperature (LST) using a well-known automated algorithm and Remote Sensing (RS) datasets. In order to analyze and evaluate the changes of land temperature in Mitrovica were used the open access datasets such as Landsat 5 (5 August 2000) and Landsat 8 (5 August 2020) with 30m spatial resolution satellite datasets provided by the US Geological Survey (USGS) through earthexplorer geoportal, incuding here the temporal scale (two last decades, 2000-2020). The input data processing and harmonization were done in QGIS software environment. For evaluating and validating the performed results of the present study, the comparison of Land Use/Land Cover (LULC) and the air temperature data were considered, as well. Overall in the present study, the algorithm for automated mapping and Geoinformation Science and Systems (GISS), proved effective approaches for mapping the Land Surface Temperature (LST) in Mitrovica on temporal scale.

ABSTRACT. For several years, one can observe an increase in the availability of modern spatial information technologies (SIT) and geographic information systems (GIS), global positioning systems (GPS) and software for remote sensing image analysis, and with them also the development of mapping techniques. Based on increasingly newer mapping technologies, large-scale mapping studies are being created that contain up-to-date information on the spatial distribution of general geographic objects, elements of land and building records, including property boundaries, as well as utility networks. The starting point for all maps is the base map, on the basis of which maps are created for legal, cadastral, investment, inventory purposes, for spatial planning, for economic purposes, etc. Prepared according to strictly defined standards, large-scale maps called maps for legal purposes allow residents to determine the extent of individual property rights for the purpose of establishing the land register, redeeming real estate, expropriation, sale or transfer of real estate, abolition of co-ownership, division of inheritance, establishment of easements and others. Determination of the legal status is also necessary in the preparation of a map for design purposes, which is a surveying and cartographic study prepared for the purpose of making a construction project necessary to obtain a construction permit decision, allowing the implementation of the investment. The purpose of the article is to analyze the technological path of preparation of maps for legal purposes and maps for investment purposes using examples from Poland, including the possibility of using modern technologies for their preparation.

ABSTRACT. The technology of unmanned aerial vehicles (UAVs) is increasingly penetrating agriculture, finding various applications from weed monitoring, addressing fertilisation and irrigation needs, to determining yields. In the organic cultivation of cereal crops, weed control is identified as a primary issue. Another significant concern is achieving sustainable yields. In this study, we examined the potential use of images obtained from UAVs for monitoring weeds, einkorn biomass, and yields in organic einkorn fields. The BBCH (Biologische Bundesanstalt, Bundessortenamt and Chemical) industry scale for cereals was employed to determine the phenological phases of the crop. Images were captured during two main phenological phases of the crop, BBCH – 45 and BBCH – 75. Weed and einkorn plant biomass were collected during UAV imaging, and at BBCH – 99, plants were harvested to determine yield. The obtained images were analysed at a pixel level of 7 × 7 cm and aggregated at a pixel level of 1 × 1 m. Regarding the crop-characterising indicators, it became evident that the data from UAVs, in both types of image processing, cannot sufficiently characterise the crop. It was established that aggregating pixels to a size of 1 × 1 m can better predict yield and productivity elements, and that BBCH – 75 phase is more suitable for this purpose.

ABSTRACT. The aim of the paper is to compare the most frequently used methods for surveying historical objects. These methods are: terrestrial confusion, laser scanning, photogrammetry and GNSS satellite methods. The article discusses the advantages and disadvantages of the mentioned methods, their laboriousness, time-consuming nature and problems during final processing in suitable software. Experimental results were obtained during surveying of the heritage protected chapel "Christ on the Mount of Olives", which is part of the Ursula convent on Josefská street in Brno (Czech Republic).

ABSTRACT. All activities carried out during the preparation, implementation, operation and maintenance of construction objects are and will be supported to the maximum extent by digital technologies. This is related to the training of those who are already or will be involved in the process to master new software, new procedures and applications in their own construction activity. The specialization of Geodesy and Cartography contributes to the electronicization of the entire construction life cycle process. The aim of the paper is to present the results of the previous analyzes of the needs of the construction industry in the area of digitization in the context of the innovation of the Geodesy and Cartography study program and the innovation of construction study programs at the Faculty of Civil Engineering of the Brno University of Technology.

ABSTRACT. In Europe, semi-natural grasslands are among the most species-rich habitats and are essential for their ecosystem services in the form of climate change mitigation and food security. In recent decades, there have been significant changes in grassland use in the form of intensification and abandonment, which have significantly impacted grassland quality. Therefore, determining grassland use intensity (GUI) is necessary to understand spatiotemporal variation in their functionality, biodiversity conservation, and sustainable decision-making in landscape policy and planning. The GUI is often expressed in mowing frequency and grazing intensity. However, full-scale information on these operations is unavailable in most agricultural regions. The aim of this contribution is to map the components of GUI using freely available optical data from the Sentinel-2 satellite mission along with ground-truth information from the Agricultural Paying Agency across six municipalities in Slovakia with diverse environmental conditions. The mapping approach consisted of two steps. In the first step, we processed the satellite imagery into a set of spatiotemporal metrics that statistically and thematically summarize the spectral information into consistent datasets. Using the Random Forest model, we identified whether an area was mowed, grazed, or used a combination of these management practices. In the second step, we then identified the time and number of mowing events and duration of grazing using a time series analysis of the NDVI index. The findings from the mapping of components of GUI indicate that the grassland plots under analysis typically underwent either a single mowing or less intense grazing. Additionally, we observed a higher accuracy in identifying mowed grasslands compared to grazed ones. This was likely due to the extensive grazing activities occurring throughout the growing season, resulting in low intra-annual variation within the time series data.

ABSTRACT. Nowadays with huge amount of Earth remote sensing data, emergence of new types of data, growth of computer capacities for their processing, development of artificial intelligence systems, expert interpretation of satellite images acquires new significance. Interpretation of urbanized territories, buildings, industrial facilities, agricultural fields and crops, etc. from space images is already largely can be carried out by computer methods. However, natural landscapes of the Earth have a huge diversity and high rate of variation of states, especially in the conditions of increased anthropogenic pressure on nature and climate change. The development of artificial intelligence systems in solving the problems of recognizing and assessing the state of landscapes requires appropriate training of specialists with a good understanding of the natural landscapes individualities in their representation by means of remote sensing. In this regard, the Geography Department of Moscow University pays special attention to the training of students in methods of expert interpretation, recognition of different types and states of natural landscapes. An important place in this process is occupied by summer field practices of students, which are held in geographically different regions of the country. Students in the field study landscape features and their image features on images of different spatial resolution and spectral ranges. The paper intends to consider methodological issues of teaching thematic expert interpretation both in the classroom and in situ

ABSTRACT. Based on Sentinel 2a/b satelite data and characterisation of the morphological features of the floodplain terraces along the lower reaches of the Veleka river. The study traces the development of riverine forests along the river in each of the three years. The Sentinel - 2 a/b data will be used to generate index images (NDVI) for the forests along the lower river current, for the months of January, March, May, July, September and November(it is possible to change the months in case of lack of cloud-free satellite scenes in any of the selected months), for each of the years of the study. The index images will be used to produce thematic maps in ArcGIS Pro. Maps, alongside with climate data (derived from meteoblue.com) and information on groundwater levels for the period (obtained via Black Sea Region Basin Directorate) are going to be used in order to interpret the condition of the forest.

ABSTRACT. Light detection and ranging (LiDAR) are originally designed to measure the range from the sensor to the target, and has been widely used in topographic mapping, building extraction, vertical structure information of vegetation, 3D/4D modelling, etc. In recent years, with the development of multi-spectral and hyper-spectral LiDAR system, many LiDAR sensors can also record the echo amplitude which is commonly referred as “intensity”. The multi-spectral LiDAR intensity data provides important information about the physical backscattering properties of the observed targets and can be useful for identify target classes (different vegetation, asphalt, or gravel), target properties (reflectivity, orientation of scatters) and so on. However, without a proper radiometric calibration, multi-spectral LiDAR intensity data acquired from different sensor or during different flight campaigns cannot be directly compared. Then to fully utilize the potential of multi-spectral LiDAR intensity data in applications, it is necessary to perform a radiometric calibration to convert the sensor raw data into physical parameters. And also, because in some cases, we cannot acquire the system performance parameters, it is even not quite clear what these system record. So, there are not yet calibration standards of theoretical framework model and practical procedures. As a contribution to support the radiometric calibration of intensity data, this paper systematically summarize the radiometric calibration and radiative transfer theories of LiDAR and built a theoretical framework model describing the BRF（bidirectional reflectance factor）of target to the received intensity data. The BRF is the reflectance measured from SVC spectrometer and had been widely used in remote sensing application. Then analysis the radiometric characteristics of intensity based on the theoretical framework model, to study all the parameters relevant to describe the effects on detected intensity signal. Finally, use experiment data to validate and improve the analysis results in real-world application, and find some useful results to advance the performance of multi-spectral intensity data in applications.

ABSTRACT. Assessment of land use land cover changes is important for policy makers to understand their effect on soil, waterflow, and forests. Rwanda has undergone large land use changes between 1990 and 2016 as 32.1% (1715.26 km2) of grassland and 64.5% (7090.02 km2) of forests were lost during this period. Built-up areas and cropland increased by 304.3% (355.02 km2) and 135.3% (8503.75 km2) respectively. Rwandan land use and land cover change has greatly been influenced by economic developments and population growth.

Rwanda is one of the landlocked tropical countries located in East Africa bordering Uganda, Tanzania, Burundi, and the Democratic Republic of Congo. It has a land surface area of approximately 25,364.52 km2 and it lies between 1 to 3 South Latitude and 28 to 31 East Longitude. Rwanda’s topography is steep with an altitude ranging between 915m to 4486m above sea level. It experiences tropical temperate climate, with a mean temperature of between 16C to 20C.

The purpose of this conference paper is to determine the land use land cover types in Rwanda for a specific year using satellite images. Five land use classes will be used Urban/Built-Up Areas, Water, Forest/Vegetation, Crop/Agricultural areas, and Bare/Barren Land. Various Machine Learning Algorithms will be used including Random Forest [RF], Support Vector Machine [SVM], Classification and Regression Trees [CART] to see if there are any differences between these algorithms. A map showing the various land use and landcover types will be derived and a comparison of the land use land cover will be made for different years.

ABSTRACT. Forest fires pose significant threats to both natural ecosystems and human settlements, with far-reaching impacts on land use and property ownership. Understanding the spatial distribution and severity of forest fire impacts on cadastral properties is crucial for effective land management and disaster mitigation strategies. This study presents a comprehensive spatial mapping and analysis of forest fire impact on different cadastral property types utilizing remote sensing and Geographic Information Systems (GIS) tools. Through the integration of high-resolution satellite imagery and GIS techniques, the extent and severity of forest fire damage in different cadastral property categories has been assessed. Our methodology involved the classification of remotely sensed data to delineate burned areas and the identification of affected cadastral boundaries. The main techniques used to find the damage from forest fires are related to the analysis of multispectral satellite images, the differences in terrain reflectivity and the use of indices to highlight the differences between areas with healthy vegetation and burned areas. The data of the individual property types are extracted from the official cadastral-administrative information system for Bulgaria. Subsequently, spatial analysis techniques were employed to quantify the degree of impact on different property types. Open source software and freely available data were mainly used for the overall research. The results revealed spatial variations of forest fire impact on cadastral properties, influenced by different factors. Furthermore, the study elucidated the differential vulnerability of various property types to forest fires, providing valuable insights for land use planning, disaster preparedness, and post-fire recovery efforts. This research underscores the significance of integrating remote sensing and GIS tools for spatially mapping and analyzing forest fire impacts on cadastral properties. The findings contribute to advancing our understanding of landscape-scale dynamics and support evidence-based decision-making for sustainable land management and resilience-building initiatives in fire-prone regions.

ABSTRACT. Farmland shelterbelts are being developed in many countries to prevent from wind erosion and provide appropriate microclimate conditions for agricultural production. However, detecting the distribution of farmland shelterbelts remains challenging due to linear features. This study used a convolutional neural network to detecting farmland shelterbelts in Northeast China Plain (NCP), utilizing Sentinel-2 data from September 1 to November 1, 2022. 4665 images (256*256 pixels in size) were visually interpreted and manually labeled to train/validate a U-Net model. We predicted farmland shelterbelts in the study area and conducted zone-based area statistics. This study shows that the total area of farmland shelterbelts in the northeastern provinces is approximately 1.5793 million hectares (Heilongjiang 1.0372 million hectares, Jilin 0.3436 million hectares, Liaoning 0.1985 million hectares) and are primarily distributed in the Sanjiang Plain, Songnen Plain, and Liaohe Plain. There is also a small amount of distribution in the valleys of mountainous areas. The overall level of gridification in farmland shelterbelts within the study area is relatively high, with Heilongjiang having the highest degree, followed by Jilin, and Liaoning ranking last. The average orientation angle of the main belts and secondary belts is approximately 170°|80° in Heilongjiang, around 140°|56° in Jilin, and about 113°|22° in Liaoning. Estimates of accuracy were performed by randomly selecting 2-3 validation regions (each approximately 40 square kilometers) in each province, with the estimated F1_Scores consistently surpassing 0.9. This study outcome can assist the government in the construction and management of farmland shelterbelts, while also providing foundational data for wind erosion evaluation.

ABSTRACT. Significant changes were made in the Czech Republic in the area of medium-scale civil fundamental State Map Series in 2023. the existing set of the Base Maps of the Czech Republic (ZM CR) and the State Map 1 : 5000 (SM 5) have been replaced by a new form of topographic maps produced by the Land Survey Office - the Base Topographic Maps of the Czech Republic (ZTM CR). ZTM CR are topographic maps prepared on the basis of data from the digital vector geographic model of the territory of the Czech Republic ZABAGED® (the Fundamental Base of Geographic Data of the Czech Republic) and data from the database of geographic names of the Czech Republic (Geonames). Compared to ZM CR, ZTM CR have expanded content and a richer list of symbols, which enables a more detailed differentiation of the elements and attributes of the underlying data. Another change is the extension of the scale series by ZTM 5, i.e. a topographic map on the border between large and medium scale. The scale of the least detailed map has also changed from 1 : 200 000 to 1 : 250 000. New maps are processed for two coordinate reference systems - the national Datum of Uniform Trigonometric Cadastral Network and then also for the European Terrestrial Reference System ETRS89 in the Universal Transverse Mercator (UTM) projection. This corresponds to two map sheet indexes and their markings. Both forms of ZTM CR are published in several forms and formats - PDF print files with complete map sheet content, including coordinate grids, frame and marginal information; georeferenced files with a composite raster image of map content; vector data of ZTM CR cartographic models in DGN and SHP format. All these data are provided as open data under the CC BY 4.0 license.

ABSTRACT. Floods are characterized by the massive flow of large amounts of water from water bodies such as lakes, rivers, and oceans to dry lands. Floods are among the most devastating natural disasters; can have devastating effects on human life, infrastructure, and mobility in urban environments. In recent years, flood events have become more frequent in Türkiye with the significant effects of climate change. Geographic Information Systems (GIS) play a major role in flood calculations and water management. Multi-criteria decision-making methods come to the fore, especially by using data obtained from different sources together. Analytical Hierarchy Process (AHP) is one of the widely used multi-criteria decision-making methods. AHP is a mathematical method used to analyze and prioritize complex decision-making processes. AHP makes decision-making processes in flood risk management scientific and systematic by ranking and prioritizing different factors. Bozkurt district of Kastomunu province (Türkiye), which experienced a major flood disaster in 2021, was determined as the study area. In this study, flood risk analysis will be carried out in Kastamonu Bozkurt district, Türkiye with AHP method. The parameters to be used in the analysis were selected as rainfall amount, temperature, digital elevation model, land cover land use map, soil properties, humidity, aspect, and slope. The weights of these parameters will be determined through an appropriate survey conducted by experts. Thus, it is aimed to create a valid flood risk map for the study area. It is expected that the study will be a very important basis for the measures to be taken against flood events in the study area and for future projects to be developed.

ABSTRACT. People with visual impairments (PVI) have perception limitations that require the maps content to be highly simplified, with limited number of symbols. Therefore, such maps are complex to develop. Moreover, methods used for tactile maps reproduction are very expensive and also have limitation in printing tactile symbols. All this means that few tactile maps are developed, mainly as products for schools or navigation. However, PVI exist within the same space as sighted individuals and cannot be excluded from the discourse on visual phenomena. Maps are vital for their autonomy, and serve as a source of knowledge, preventing their informational and social exclusion. These heavily relies on PVI's access to accurate, clear, and affordable tactile maps. One particular area of exclusion for PVI is their insufficient access to cultural goods experiences - there are almost no maps supporting it. In our research we try to fulfil this gap and propose the method of design tactile maps of historic gardens, as an important part of cultural heritage in each country. In the presentation, a comprehensive technology for developing tactile maps of historic gardens in five garden design styles (Baroque, Renaissance, Romantic, English and Japanese) will be presented. Proposed technology includes: the scope of content for garden tactile maps at different level of map details, the design of tactile signs, the principles of map generalization and editing, and selected a low-cost printing techniques (eg. 3D, UV). All our stages are based on methodological solutions, which were verified during 6 test sessions with PVI. Each group includes 15-20 PVI with diversified sociodemographic characteristics. During presentation we will show main achievements of the technology, and experiences from testing sessions, as well as technical and practical solutions useful in low-cost methodological tactile mapping. Finally, tactile map prototypes for 5 various gardens will be also presented.

ABSTRACT. The automatic realization of cartographic generalization is one of the most important recent goals of thematic cartography. There are some preliminary studies to make this process, which has been carried out manually and on-screen digitization for many years, semi- or fully automatic. In this context, the literature has been scanned in detail, and the initiatives and studies carried out using computer vision, deep learning, and CNN have been examined. The results are presented chronologically with the help of tables in this study. The results of using deep learning networks in cartographic generalization are promising in some classes. However, the classification of different classes with deep learning requires high-level expert knowledge in this field, and advanced computers are needed. In this study, the generalization of land cover and land use (LULC) maps was carried out automatically with the help of digital image processing and classification algorithms using object-based image classification (OBIA) with the help of optimally determined parameters. The generalization achieved using this method is quite practical and produces good results. Different segmentation algorithms were used as the OBIA method, and the results obtained in the residential area, green area, impervious surface, and wetland classes were evaluated comparatively with the general accuracy metric. Accordingly, higher accuracy results were obtained with the multi-resolution segmentation (MRS) algorithm. and the LULC maps obtained at different scales are presented as a result of the study as a visualization of the generalization.

ABSTRACT. Probe data is defined as data that is generated by monitoring the position of individual vehicles (i.e., probes) over space and time rather than measuring characteristics of vehicles or groups of vehicles at a specific place and time. HERE Probe Data is a location enrichment service for understanding traffic patterns and mobility trends. It consists of anonymous GPS traces from third party data providers and collected from vehicles and mobile devices that use the HERE services. The present article’s aim is to build a spatial model in ArcGIS Pro. To perform probe data analysis for the city of Sofia and to visualize data. Real time spatial traffic model for Sofia city would determine specific trends and patterns. Achieved results would serve for further multidisciplinary areas of research and development.