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