Slovakian amateur football club TJ Tatran Cierny Balog prides itself on having one of the most unique stadiums in the world complete with a railway track and a steam engine running straight through it a small Slovakian town of about 5,100 people has become somewhat of a tourist spot in the last seven years or so and it was all thanks to its football stadium a video of a steam engine passing through the stadium on tracks positioned right between the field and the only existing grandstand went viral online leaving a lot of people scratching their heads or was there actually a train regularly passing right through the stadium The Čiernohronska Railway goes right through Cierny Balog stadium and a steam-powered tourist train passes through it all summer long Cierny Balog stadium was built sometime during the 1980s at a time when the old railway was not in use anymore It’s unclear why the municipality chose that exact location to build a football stadium the train tracks passing through it were just part of the decor when a group of Slovakian enthusiasts started repairing the train tracks it was officially reopened as a heritage railway for tourists tourists wanting to admire the beauty of the Lower Tatras mountains from aboard a train can buy tickets for a vintage steam engine-powered train that runs on a 17-km-long line between Chvatimech and Vydrovo Passing through the Cierny Balog stadium is one of the highlights of the trip Seeing the steam train rolling through the stadium during a football match is quite a surreal experience The players try to ignore the train and its powerful whistling signal while others try to keep their focus on the game Cierny Balog is believed to be the only stadium in the world with an active railway track passing straight through it And here we thought having a tree growing in the stands of a football stadium was weird… Metrics details Natura 2000 is a network of protected areas covering Europe's most valuable and threatened species and habitats biota belonging to these networks have been threatened by both climate change and various human impacts Regular monitoring is needed to ensure effective protection and proper management measures in these sites and habitats but conventional field approaches are often time-consuming and inaccurate New approaches and studies with different focuses and results are being developed Our approach includes point data from field research and phytosociological databases as starting points for automatic segmentation which has been developed just recently as a novel method that could help to connect ground-based and remote sensing data Our case study is located in Central Slovakia in the mountains around the village of Čierny Balog The main aim of our case study is to apply advanced remote sensing techniques to map the area and condition of vegetation units We focus on forest habitats belonging mainly to the Natura 2000 network We concentrated on the verification of the possibilities of differentiation of various habitats using only multispectral Sentinel-2 satellite data Our software NaturaSat created by our team was used to reach our objectives After collecting data in the field using phytosociological approach and segmenting the explored areas in the program NaturaSat spectral characteristics were calculated within identified habitats using software tools which were subsequently processed and tested statistically We obtained significant differences between forest habitat types segmentation accuracy was tested by comparing closed planar curves of ground based filed data and software results This provided promising results and validation of the methods used The results of this study have the potential to be used in a wider area to map the occurrence and quality of Natura 2000 habitats The recognition of single habitats using Sentinel-2 data could give us a great opportunity to monitor Natura 2000 habitats on the European level and describe their spatiotemporal dynamics maps of Natura 2000 habitat occurrence are presented as a square grid with information about habitat presence/absence derived from phytosociological relevés coordinates these phytosociological relevés have a large potential to serve as a base for exact habitat mapping if they are transformed from point to spatial data using They work based on evolving planar curves and are efficient and robust segmentation tools when an “initial estimate” of the desired area is available the pointwise estimate of habitat occurrence is available from vegetation databases or field surveys The aims of the present study are (i) to show the feasibility of recognizing Natura 2000 forest habitats based on optical information from Sentinel-2 satellite data; (ii) to find a suitable period of the year for habitat recognition based on phenological aspects; and (iii) to test the feasibility of using phytosociological relevés with coordinates as starting points for semiautomatic and automatic segmentation to find the exact habitat area The case study area is situated in the vicinity of the village of Čierny Balog in central Slovakia at an altitude of 559–1338 m a.s.l. (Fig. 1). wood processing and pastoralism were the main activities that contributed to the gradual transformation and deforestation of the study area the study area is characterized mainly by forest habitats: 9130 Asperulo-Fagetum beech forests; 9180 Tilio-Acerion forests of slopes and ravines; and 9410 Acidophilous Picea forests of the montane to alpine levels (Vaccinio-Piceetea) (spruce forests) and planted spruce forest (secondary spruce forests) During the 2020 and 2021 vegetation seasons the boundaries of 24 selected habitats (8 segments of 9130 Asperulo-Fagetum beech forests; 9180 Tilio-Acerion forests of slopes forest segments with an easier definition of boundaries (which are surrounded by a different type of habitat) and areas with more complex boundaries (which border on the same or similar type of forest) were chosen The compact stands of the native spruce forests (habitat 9410) were concentrated only in one locality which did not allow us to sample an appropriate amount of data the occurrence of 9410 habitat in the study area was on its southern border of distribution and habitat did not represent typical characteristics there In the case of problems in the field (loss of GPS signal or overly steep impassable rocky slopes) GPS tracks were additionally corrected in Google Earth Pro software The overall segmentation results are given by an interconnection of several open curve segments The quality of these segments is controlled by the user and governed by the mathematical model and its parameters The automatic segmentation could start from the user-defined initial closed curve or from the phytosociological relevé representing the current habitat or plant community A total of 107 areas were segmented (with an area of 586.51 ha of polygons obtained semiautomatically) which were divided into four groups according to habitat type resulting from the species composition of phytosociological relevés and expert knowledge The automatic segmentation is influenced by the expanding term including the homogeneity detector function The values of these parameters could give different segmentation results in the case of different habitats monodominant forests require a smaller min–max range of the homogeneity function in comparison with mixed forests Tuning these parameters for the studied habitats was one of the aims of the present research NaturaSat tools allow us to compute spectral characteristics inside segmented regions and maximal intensity values and standard deviations of all Sentinel-2 optical bands inside each forest segment The set of computed values in all optical bands creates a habitat spectral characteristic that could be used for its identification The Sentinel-2 satellite records 17 spectral bands from which 14 were used in our analyses for vegetation monitoring (AOT-Aerosol Optical Thickness and WVP-Scene-average Water Vapour map and Vegetation classification band values B05 SCL-Scene classification and SNW-Snow map were omitted Each habitat is thus characterized by 56 spectral values for each date The spatial resolution of Sentinel-2 images is up to 10 m (spatial resolution of bands B02 This RLH classifier was developed and implemented in NaturaSat software to detect structural differences and capture the heterogeneity (diversity) of data The RHL value is given by the finite difference numerical approximation of the Laplace operator applied to the intensity of the smoothed image in each pixel and then by calculating the ratio of high Laplacian pixels and all pixels within the segmented region The high Laplacian pixels are defined relative to the maximum MAX of the Laplacian in the segmented region We denoted by RHL N% the results of the RHL classifier for the segmented region when the pixel was considered to have the high Laplacian if its Laplacian was greater than N% of MAX Because the high Laplace operator is achieved at the points with the highest curvature of the graph of image intensity the RHL represents the diversity of the forest observed in the optical bands In addition to testing the distinguishability of all four habitat groups we were interested in which period of the year differences in spectral characteristics of habitats were the most significant The characteristics were computed for datasets recorded in different parts of the year: spring—May 13 The relevance of division was tested using PERMANOVA to confirm the significant differences in the species composition of the identified habitats All four forest habitats classified by Twinspan were first analysed together we also analysed when was the best season to distinguish the planted spruce forest from habitat 9410 (native spruce forest) we analysed which band values and combinations were most appropriate and which classification results were most significant we examined the distinguishability of the two mixed habitats 9130 Asperulo-Fagetum beech forests and 9180 Tilio-Acerion forests of slopes These distances were computed using the NaturaSat software tools for sets of semiautomatic segmentations versus GPS tracks and automatic segmentations versus GPS tracks The process of our research approach is shown in the Fig. 2. Flow chart of our research solution process Four different forest vegetation types (habitats) were distinguished in the case study area. From these, three of them represent Natura 2000 habitats (9130, 9180, and 9410), and one denotes planted spruce forests (secondary spruce forests) (Table 1) A significant differentiation of grouped relevés based on species composition showing relationships between target habitats. Abbreviations: s – 9410 – Acidophilous Picea forests of the montane to alpine levels, n – planted spruce forest, m – 9130 – Asperulo-Fagetum beech forests, and r – 9180 – Tilio-Acerion forests of slopes, screes, and ravines. spatial resolution of Sentinel-2 satellite image is 10 m 1 pixel on the satellite image represents 10 × 10 m on the Earth surface) Coniferous forests are dominated by Picea abies in 9410 with an admixture of Sorbus aucuparia Mixed forests on steeper slopes and screes (9180) are species rich They are floristically well differentiated from the other forests by the tree (sub)dominant species Acer platanoides and Ulmus glabra and numerous herbs of the order Fagetalia together with the nitrophilous and nutrient-demanding species Aegopodium podagraria In mixed Fagus sylvatica-Abies alba forests (9130) all these species are absent (they are negatively differentiated) A higher frequency and dominance of Abies alba is a typical feature of these forests Spruce forests differ from mixed ones by several acidophilous species: Avenella flexuosa Native spruce forests (9410) are typical of numerous mountain to (sub)alpine plants (e.g. some of which are diagnostic species of the tall herb vegetation of nutrient rich and moistened habitats of the class Mulgedio-Aconitetea (Adenostyles alliariae the stands are enriched by Asarum europaeum and Viola reichenbachiana—typical species of the order Fagetalia surviving from previous native broad-leaved or mixed forests Forest clearing species of the class Epilobietea angustifolii represent succession residuals after cutting original forests Picea abies frequently dominates in the herb layer Within the NaturaSat software environment, we were able to identify 107 segmented areas using the Sentinel-2 satellite images according to the procedure described in the methodology section (Fig. 3) By analysing the target habitats, all selected optical values described in the methods were used. The results of additional PERMANOVAs (the P values were less than 0.001, unless otherwise stated) confirmed the assumption that target forest habitats can be recognized remotely (Fig. 5). A significant differentiation of target forest types based on various combinations of optical bands extracted from Sentinel-2 data (obtained on May 13 Abbreviations: s – 9410 – Acidophilous Picea forests of the montane to alpine levels and r – 9180 – Tilio-Acerion forests of slopes The distinguishability of coniferous habitats and those in which deciduous trees were more numerous was confirmed at a high significance level these habitats were analysed separately to provide better insight into specific differences To test the usage of RHL values to distinguish between 9410 spruce forest and planted spruce forest (forest areas older than 80 years), RHL values were calculated for 48 segments of the tested habitats. Segments of spruce forests (s) always had higher RHL values, while segments of planted spruce forests (n) had lower and more variable values of RHL (Fig. 6). Differentiation of RHL 10% of selected forest types in the B04-Red optical band intensity value in August (dataset from August 31 Abbreviations: n – planted spruce forest and s – 9410 – Acidophilous Picea forests of the montane to alpine levels In the analyses of coniferous forests, we also added the values of the RHL 10% parameter to the values of mean, max, min, and Std, as it turned out that the results of this combination of analysed values were the most significant. The period of early spring (end of April, the beginning of May) and late autumn (November) seemed to be most suitable (Fig. 7). Differentiation of two coniferous forest types using mean Abbreviations: s – 9410 – Acidophilous Picea forests of the montane to alpine levels and n – planted spruce forest The most significant combination of mean and max values distinguished deciduous habitats. The summer aspect (end of August) seemed to be the most suitable for this purpose (Fig. 8). Differentiation of two mixed habitats using mean and max optical values (data from August 31 Abbreviations: m – 9130 – Asperulo-Fagetum beech forests and r – 9180 – Tilio-Acerion forests of slopes The accuracy of semiautomatic segmentation, i.e., comparison of semiautomatic segmentation and GPS tracks by means of the Hausdorff distance, is presented in Table 2 The mean Hausdorff distance was on average 8.86 m which is smaller than the spatial resolution (10 m) of Sentinel-2 data This indicates that by using semiautomatic segmentation we were able to detect habitat borders more accurately as the image resolution allowed this The maximal Hausdorff distance was on average approximately 28.67 m (less than 3 pixels) In the next part of our research, we focused on the automatic segmentation of selected habitats. The results of automatic segmentation and GPS tracks compared by the Hausdorff distance are presented in Table 3 which was only slightly more than the spatial resolution of the satellite data The highest differences could be found in the areas with ecotone zones where forest patches were connected to surroundings by shrub zones or in segments that were connected to a similar type of vegetation separated in the field only by a natural barrier (narrow forest road which occurs in this type of habitat and is typical for this area but faintly recognizable on satellite images with a resolution of 10 m An example of a visual comparison of semiautomatic, automatic, and GPS curves is presented for the segment of habitat 9180 Tilio-Acerion forests of slopes, screes, and ravines, relevé IJ4299 (Fig. 9). The figure was obtained using NaturaSat software (a dataset from October 17, 2019). The mean Hausdorff distances of segment of planted spruce forest relevé 769,205 were 11.3 m (comparison of semiautomatic segmentation and GPS track) and 11.9 m (comparison of automatic segmentation and GPS track) and 4.8 m (result of comparison between automatic and semiautomatic segmentation) which is very close and even less than the pixel resolution of the Sentinel-2 data the present study makes an important contribution to the application of RS in nature conservation the pixel-based approach was replaced by a new evolving curve approach that allows obtaining the exact area by automatic segmentation with phytosociological relevés as a starting point This unique interconnection between traditional ground-based information and remote sensing gives us the advantage of exact affiliation of the segmented area into Natura 2000 habitat The approach was tested in a case study area in Čierny Balog but the same methods could be used in different parts of Slovakia or Europe where phytosociological relevés of studied habitats occur which suggests strong potential for the use of this method our results show that freely available multispectral remote sensing data can be used to identify some types of forests at the scale of single Natura 2000 habitats that until now were not determined based on satellite images The results of automatic segmentation in our dataset have an accuracy of approximately 15 m and the results of the semiautomatic segmentation have an accuracy of only approximately 9 m which is less than the pixel resolution of Sentinel-2 data (1 px = 10 m) Mapping Natura 2000 forest habitats is a major challenge Biotopes belonging to the Natura 2000 network are often more complex have a transitional ecotone area with smaller trees and shrubs or are adjacent to forests with a similar type of vegetation and are separated by obstacles visible only in the terrain (e.g. we also noticed these limitations in semiautomatic segmentation but they may not be easily recognizable on satellite images and are not considered in automatic segmentation There is also a difference between the tree trunk and tree crown/branches The tracking of boundaries in the field may be different especially when the trees are tall and/or the trees of different habitats are touching The combination of these segmentation methods allows the necessary simplification and acceleration which could be part of habitat mapping and monitoring for both scientists and conservation needs The methods used were able to reach the accuracy of the pixel resolution of Sentinel-2 data which are suitable for forest habitat mapping and identification; however the methods have large potential when using data with a ALS or UAV resolution our methods have the potential to achieve pixel resolution in these cases The overall accuracy of different methods at diverse scales of target phenomena is very variable it is complicated to compare the evolving curve approach to pixel-based approaches Our accuracy is based on the comparison of whole areas identified by GPS tracks and segmentation curves situated in forested areas with exactly defined Natura 2000 units Commonly used approaches focus on the selection of a network of points in the whole landscape—from afforested sites to abandoned land the cover of a single forest habitat type was identified against other forests and the area that NaturaSat software could find by automatic segmentation using only optical multispectral data differed from the forest patch area by an average of 15%; the minimum value was only 0.44% (i.e. The borders of forest differed by 1–5 pixels where the mean Hausdorff distance was 14.6 m which was only 1.5 pixels of Sentinel-2 resolution between GPS-tracked and automatically segmented areas the area of the target types of forest habitats was identical to the area obtained by GPS this approach is particularly applicable in European countries Alpine and Pannonian biogeographical regions with a long tradition of phytosociological approaches although the methods could be transferred anywhere The interconnection between the phytosociological approach and remote sensing techniques brings new possibilities for the exact mapping of forest habitats and can contribute to enhancing many future studies in vegetation and landscape ecology The datasets used and/or analysed during the current study available from the corresponding author on reasonable request Measuring β-diversity by remote sensing: A challenge for biodiversity monitoring Satellite remote sensing of ecosystem functions: Opportunities Remote Sensing in Ecology and Conservation 4(2) Remote sensing for mapping natural habitats and their conservation status–New opportunities and challenges Identification of areas of very high biodiversity value to achieve the EU Biodiversity Strategy for 2030 key commitments Probabilistic description of vegetation ecotones using remote sensing Digital mapping of peatlands–A critical review Advances in microclimate ecology arising from remote sensing New research methods for vegetation information extraction based on visible light remote sensing images from an unmanned aerial vehicle (UAV) Consistency in land-cover mapping: Influence of field workers Effects of data temporal resolution on phenology extractions from the alpine grasslands of the Tibetan Plateau Using remote sensing to characterize riparian vegetation: A review of available tools and perspectives for managers Mapping species composition of forests and tree plantations in Northeastern Costa Rica with an integration of hyperspectral and multitemporal Landsat imagery Assessment of Sentinel-2 satellite images and random forest classifier for rainforest mapping in Gabon Sentinel-2: ESA’s optical high-resolution mission for GMES operational services European Vegetation Archive (EVA): An integrated database of European vegetation plots EUNIS Habitat Classification: Expert system characteristic species combinations and distribution maps of European habitats constant and dominant species of the higher vegetation units of Slovakia Katalóg biotopov Slovenska [A catalogue of biotopes of Slovakia] (DAPHNE-Institút aplikovanej ekológie Európsky významné biotopy na Slovensku (Štátna ochrana prírody SR v spolupráci s DAPHNE-Inštitút aplikovanej ekológie Phenological differences in Tasseled Cap indices improve deciduous forest classification Delineating forest canopy species in the northeastern United States using multi-temporal TM imagery Temporal optimisation of image acquisition for land cover classification with Random Forest and MODIS time-series The effect of Landsat ETM/ETM+ image acquisition dates on the detection of agricultural land abandonment in Eastern Europe Hyperspectral data dimensionality reduction and the impact of multi-seasonal Hyperion EO-1 imagery on classification accuracies of tropical forest species Neural network classification of mangrove species from multi-seasonal Ikonos imagery Mapping tropical forests and rubber plantations in complex landscapes by integrating PALSAR and MODIS imagery Phenology-based vegetation index differencing for mapping of rubber plantations using Landsat OLI data Mapping rubber tree growth in mainland Southeast Asia using time-series MODIS 250 m NDVI and statistical data Mapping rubber plantations and natural forests in Xishuangbanna (Southwest China) using multi-spectral phenological metrics from MODIS time series NaturaSat—A software tool for identification monitoring and evaluation of habitats by remote sensing techniques Semi-automatic segmentation of NATURA 2000 habitats in Sentinel-2 satellite images by evolving open curves An automated segmentation of NATURA 2000 habitats from Sentinel-2 optical data Geologická mapa Slovenského rudohoria a Nízkych Tatier (1: 50 000) Historical terraces—Current situation and future perspectives for optimal land use management: The case study of Čierny Balog Geoportal, 2021. https://www.geoportal.sk/sk/geoportal.html Council Directive 92/43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora Kritische bemerkungen und vorschläge zur quantitativen vegetationsanalyse In Classification of Plant Communities (ed a comprehensive data base management system for vegetation data A list of lower and higher plants of Slovakia Semi-implicit scheme for semi-automatic segmentation in Naturasat software LGIS, 2020. LGIS Lesnícky geografický informačný system (2020). http://gis.nlcsk.org/lgis/ TWINSPAN-a FORTRAN program for multivariate data in an ordered two-way table by classification of the individuals and attributes Team, R.C., A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2014. www.r-project.org Extracting agricultural fields from remote sensing imagery using graph-based growing contours Remote sensing of grassland production and management—A review Forest types classification based on multi-source data fusion A review of spectral properties of plants and their potential use for crop/weed discrimination in row-crops Pa—precision agriculture: Remote-sensing and mapping of weeds in crops Opportunities and limitations for image-based remote sensing in precision crop management Remote sensing of forest insect disturbances: Current state and future directions Assessing sustainable development prospects through remote sensing: A review Integrating remote sensing in Natura 2000 habitat monitoring: Prospects on the way forward Transitioning from change detection to monitoring with remote sensing: A paradigm shift Fuzzy modelling of vegetation from remotely sensed imagery Using remote sensing to assess biodiversity Mapping Natura 2000 habitat conservation status in a pannonic salt steppe with airborne laser scanning Mapping the local variability of Natura 2000 habitats with remote sensing and support vector machine classifiers for land cover classification using Sentinel-2 imagery Assessing forest cover dynamics and forest perception in the Atlantic Forest of Paraguay combining remote sensing and household level data Forest type classification based on integrated spectral-spatial-temporal features and random forest algorithm—A case study in the qinling mountains and mapping of functional guilds by hyperspectral and simulated multispectral Sentinel-2 data Mapping and assessment of vegetation types in the tropical rainforests of the Western Ghats using multispectral Sentinel-2 and SAR Sentinel-1 satellite imagery and Sentinel-2 data for mangrove plantation aboveground biomass monitoring in Senegal Evaluating Sentinel-2 and Landsat-8 data to map sucessional forest stages in a subtropical forest in Southern Brazil Mikula, K., et al., Natural Numerical Networks for Natura 2000 habitats classification by satellite images. arXiv preprint arXiv:2108.04327 sPlot—A new tool for global vegetation analyses Download references This research was financially supported by the projects APVV16-0431 APVV-19-0460 and VEGA 2/0119/19 and European Space Agency We would like to thank Michaela Michalková and local forester Peter Kovalčík for their help during field research All authors have read and agreed to the published version of the manuscript The authors declare no competing interests Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations Download citation DOI: https://doi.org/10.1038/s41598-022-23066-3 Anyone you share the following link with will be able to read this content: a shareable link is not currently available for this article Sign up for the Nature Briefing newsletter — what matters in science Trains and sports are almost never things that pop up on my radar at the same time if I’m watching a movie like A League Of Their Own For the people of the municipality of Čierny Balog in Slovakia they get to watch steam trains roll right through a soccer field Even more interesting is how it all happened One of my dreams is to experience some of the best rail experiences the world has to offer and one day take a romantic ride aboard an overnight train through the mountains I’d also love to check out the weird railways out there Tim is presenting perhaps the only place in the world where a stream train will chug its way through an active league soccer (football for the rest of you) match Tim’s video is actually a follow-up to a previous video about the bizarre setup but this one has the most current information Tim asked the football club to explain the one question we all have: Why he was told that the football pitch was created after the railway closed in 1982 you would have realized it couldn’t have happened like that Tim’s video opens up with clips showing a train steaming through a soccer field before showing another clip of a train chugging through an active game The Čierny Hron Railway is a narrow gauge railway that runs along the Čierny Hron River in Central Slovakia the railway runs preserved equipment and when you start in Čierny Balog you’ll want to choose the route to Dobroč stadium the railway is closer to that of a museum in motion and provides tourists with a scenic ride through the valley it served a vital role in the region’s industry As the railway’s official website notes logging and wood processing have been one of the main sources of income for people living in the Slovak mountains throughout history it was common for timber to be transported on rafts these rafts became insufficient to support the region’s growing wood industries The area’s geography presented a better option: The forest railway the region would see the construction of narrow gauge railways that wrapped their way around rivers and tributaries while connecting sawmills and large railroads to industries all over what we call Slovakia today the railways could better adapt to the mountainous region while saving on construction costs these trains would be hauled by horses before steam locomotives became the preferred method Later came diesel and electric locomotives the region was home to about 40 forest railways that ranged from several kilometers long to full-blown public transportation networks the Ministry of Agriculture in Budapest ordered the director of state forests in Banská Bystrica to develop an economic justification for the construction of a forest railway in the Čierne Hron valley A proposal with an appropriate economic rationale was submitted in 1901 The construction of the main section from Hronec to Čierny Blh (as it was then called Čierny Balog) an official traffic-legal commission was carried out Other branches of the ČŽŽ were gradually built with the help of prisoners of war from the First World War and led to most of the valleys of the Čierne Hron basin Their total length reached 131.98 km and they climbed up to the mysterious Dobročská forest under the mighty Klenovský Vepro The smallest radius of the curve of the track was 60m the biggest slope of the track was 70 per mile A transfer point from ČŽŽ to the state railway was established at the Hronec station Steam sawmills supplied by the forest railway worked in Čierno Balog and Štiavnička At the time of the largest hauling of wood ie the period of calamities 1927 – 1929 about 260,000 m3 of wood was transported annually and in the years 1953 – 1955 even up to 300,000 m3 of wood Up to seven steam locomotives of various designs used to run here daily which were later joined by 3 RÁBA diesel-hydraulic locomotives from Hungary The Čierny Hron Railway also explains that in 1944 the region’s rail network was vast and it became used as a supply network for Slovak resistance fighters: The Forest Railway also made its mark in the history of the Slovak National Uprising in 1944 when it ensured the importation of provisions and ammunition into the mountains for the partisans It significantly contributed to the fact that the Germans never conquered Čierna Balog While it’s true Germany never took over Čierna Balog the railways were still made a military target and during the uprising Tim believes that of all of the preserved rail equipment on the Čierny Hron Railway today just one of the locomotives is a survivor of wartime The others were either destroyed by war or scrapped later on that’s not stopping the railway’s volunteers from saving as much rail equipment as they can The volunteers will save equipment from other railways and restore them back to their former glory I told you this place is like a living museum the region started seeing rapid improvements in road infrastructure along with the adoption of cars and trucks started bringing an end to the rail networks and advanced machinery made trucking logs down cheaper than getting them loaded onto steam trains tourists gravitated toward luxury coaches as steam trains were seen as dirty and out-of-date at the time trains started falling by the wayside as people hopped into cars and there was a dearth of others who wanted to modernize rail technology Railway after railway closed until only Čierny Hron Railway and its 36 km (22.3 miles) of track were left The government ordered the rails to be ripped up and the equipment either sold off or scrapped by 1985 but then railfans fought back by putting the tracks back and successfully getting the railway listed as a state cultural monument The railfans were faced with an uphill battle of essentially rebuilding the entirety of the Čierny Hron Railway from scratch the first steam locomotive rode down the railway The amateur Tatran Čierny Balog league football club at Dobroč initially told Tim that the football pitch (that’s a soccer field for us Americans) was built after the railway closed in 1982 But that cannot be correct because you can see the grandstands and the rails in the background of the team’s historical photos predating 1982 Even the clubhouse was there long before 1982 the football pitch has been alongside the rails since the team formed in 1933 Here’s another video of trains going through the stadium: This still doesn’t really answer the question of why there are not just one but two football pitches along this rail line but the region’s geography gives some clues There’s not a lot of flat land out there in Čierny Balog and Dobroč What is pretty flat is the land near the rails Tim thinks it’s possible that at the time there just wasn’t anywhere better to put large sports fields it needed to build facilities including seating and a clubhouse so perhaps the easiest decision was to build the seats on the other side of the rails rather than move the field or the rails If you want to ride on the Čierny Hron Railway it runs daily from April until mid-September Tickets are 12 euros each and it’s half-price for kids Today the Čierny Hron Railway not only remains in operation but its volunteers even have ambitious plans such as restoring and introducing electric trains onto the line and even a temporary blockade from a property developer but it stands triumphant today as a piece of local and national history If it weren’t for railfans wanting to preserve history If you’re interested in obscure transportation history presented by a charismatic tourist, I highly recommend giving The Tim Traveller a watch (Screenshots: The Tim Traveller on YouTube) Google News MSN News Apple News+ I remember seeing videos of the line passing a game a few years ago but I didn’t know it’s history so I don’t know if this is common knowledge there but I used to live in Winston-Salem where the Winston-Salem State football team (DII) plays on a field inside the Bowman Gray Race Track I’ve never heard of another stadium that combines motor sports and field sports Where else in the world can someone roll coal through a soccer game This seems like something you’d see in a Johnny English movie He kicks the “football” and it unluckily gets stuck in the exhaust stack of the steam locomotive causing the boiler to blow up… I believe Midway stadium in St Paul MN had a train running next to the stadium There are some noticeable differences between the Shinkansen and the high-speed maglev train that connects the Shanghai airport to the Shanghai train station The Shinkansen has an elongated nose (inspired by the kingfisher’s beak) so they don’t push as much air ahead of them; the Shanghai trains had flatter noses and when two trains passed in opposite directions there was a loud bang On a business trip ~15 years ago I had a free afternoon in Shanghai so I rode the maglev from the station to the airport and back again 🙂 They may have improved the aero since then He’s got some great transportation videos Reminds me of the train going through the baseball field on Brewster’s Millions That was hilarious (Also- “Vote None Of The Above” ha ha) This is awesome…well done on another great train article Mercedes I haven’t seen it in thirty years; here’s hoping it holds up I probably watch it once every year or 2…Richard Pryor is a legend John Candy is also one of my all time faves Read More » Amateur team TJ Tatran Čierny Balog plays on through greater distractions than most football teams I would like to be emailed about offers, events and updates from The Independent. Read our Privacy notice Every football stadium takes pride in its odd yet characterful eccentricites, however Slovakian team TJ Tatran Čierny Balog have a strong claim to the strangest. The amateur team's pitch lies to the side of a running railway track, with a stand on the other side of it. Trains continue to use the line, even as a match is being played. A fan of the team uploaded a video of one such timely service, which the players seem largely undistracted by, even if the same can't be said for the fans. The team's support enjoy the obstruction to their view greatly, taking photos and cheering the locomotive's passage through the ground. Perhaps it is a testament to the frequency of such an event, as play continues despite thick steam and a loud whistle coming from the train. It remains unknown how the ball is retrieved when settling on the line, or how strong the train's windows are. Join thought-provoking conversations, follow other Independent readers and see their replies Two pedal draisines on the railway between Hronec and Čierny Balog Tourists can now “cycle” on railways between the Hronec and Čierny Balog stations in the Horehronie region The railway cannot be used properly because of its bad condition so there was the idea of launching the cycling draisine there are two paddling draisines accessible on the railways borrowed from Poland,” said Aleš Bílek Two people can paddle at the same time and two or three other people can sit on the draisine “We are proud to bring the attraction to the region as the first of its kind in Slovakia,” he added The rest of this article is premium content at Spectator.skSubscribe now for full access Immediate access to all locked articles (premium content) on Spectator.sk Special weekly news summary + an audio recording with a weekly news summary to listen to at your convenience (received on a weekly basis directly to your e-mail) PDF version of the latest issue of our newspaper Access to all premium content on Sme.sk and Korzar.sk Sport | Football There are a number of things that could potentially distract football players during matches - but a train hurtling its way past a pitch during a game has got to be perhaps the most unusual - and distracting - of them all that is the reality for Slovakian amateur team TJ Tatran Čierny Balog who play their games next to a railway track that regularly has trains making its way past the ground - even during matches A video uploaded onto social media by a fan attending a game involving the Slovakian team shows just how focused players must be as a steam train travels along the length of the pitch during a match while spectators sitting in stands just in front of the track watch on Aside from the fact that a train goes right past the ground the noise made by the engine coupled with the steam emitted from the train prove to be the ultimate distraction for players - although supporters certainly seem to enjoy the show with many taking pictures and indeed videos of the unusual scene Watch a clip of TJ Tatran Čierny Balog's stadium playing host to a train during a game above Prince Louis steals the show at VE Day parade as he keeps dad William looking sharp and mimics brother George Prince Louis steals show with sweet antics at VE parade Ukraine 'launches stunning Kursk offensive' in major blow for Putin ahead of Victory Day celebrations Ukraine 'launches stunning Kursk offensive' in blow for Putin VE Day 2025 fashion: best looks from the day VE Day 2025 fashion: Princess of Wales to Lady Victoria Starmer New visa crackdown as Home Office plans to restrict applications from nationalities most likely to overstay New visa crackdown as Home Office plans to restrict applications worst and wackiest looks on the red carpet — Rihanna worst and wackiest looks on the red carpet