The world of fragile underground ice caves offers its beauty in many countries worldwide. Their uniqueness attracts the attention of many researchers as well as the general public and they can be considered as one of the least well-known parts of the global cryosphere (Kern and Perșoiu, 2013; Perşoiu and Lauritzen, 2018). According to Ford and Williams (1989), ice caves can be defined as rock caves containing seasonal and/or perennial ice. Another definition by Perşoiu and Onac (2012) states that an ice cave is a cave formed in bedrock that contains perennial accumulations of water in its solid phase (i.e., ice and snow). Ice caves have been described in many parts of the world at altitudes where the mean annual air temperature is several degrees above 0°C (e.g., Harris, 1982; Holmlund et al., 2005). Depending on the cave morphology, this is due to the accumulation of cold air or unidirectional advection of cold air into the cave during winter (Luetscher and Jeannin, 2004). Cave ice is a special class of natural ice having various geneses, but it exists in a cavity and is located below the Earth’s surface. It is connected with the open air. Cave ice belongs to the category of subaerial - underground ice (Perşoiu and Lauritzen, 2018). The interplay of cave climate, i.e., the values, distribution, and dynamics of temperature (air and substrata), humidity, and air movement in caves, on a background of peculiar cave morphology, is responsible for the genesis, accumulation, and ice dynamics in caves. The composition and structure of cave ice provides the basis for their separation into an independent class of natural ice, intermediate between subaerial and underground (Perşoiu and Lauritzen, 2018).

Cave glaciation (perennial or temporary) is a widespread phenomenon on Earth. For example, approximately 10% of all karst cavities known in the former USSR have been subjected to cave glaciation (Perşoiu and Lauritzen, 2018). According to Mavlyudov (2018), it appears that the main areas where cave glaciation can exist are in North America and Eurasia. Due to its climate, Australia has the potential for only seasonal cave glaciation in its high mountain regions. In contrast, cave glaciation may occur in the mountains of New Zealand’s South Island. In Africa, most areas are not adapted to cave glaciation, with the exceptions being Mount Kilimanjaro and the Atlas Mountains. Cave glaciation in South America has been poorly studied, but there are suggestions of cave glaciation presence in the upper regions of the Andes Mountains. Also, there have been confirmations of ice-filled caves at high altitudes in South America. In addition to Europe being the most active region for ice caves research, many ice caves can be found in Canada (Yonge et al., 2018), Iran (Soleymani et al., 2018), Asia (Mavlyudov and Kadebskaya, 2018), Russia (Mavlyudov and Kadebskaya, 2018), and United States (Higham and Palmer, 2018).

The study of the underground cryosphere is, at present, considered to be extremely important and urgent because ice degradation processes are widely observed all over the world (Luetscher et al., 2005) and because little is known about the processes controlling the mass balance of perennial subsurface ice deposits and its evolution under changing climate conditions (Luetscher et al., 2008; Morard et al., 2010). However, cave ice serves as a local resource (for example, for tourism), its degradation may destabilize underground, and it preserves environmental information that has remained largely unexplored so far (Obleitner and Spötl, 2011). In caves, the volume and surface morphology of ice bodies result from the extent of glaciation and interannual or seasonal changes of perennial ice fill (Perşoiu and Lauritzen, 2018). Since seasonal changes cause changes in ice morphology, it is important to study not only the dynamics of spatial changes, such as the loss and gain of ice fill, but also changes in ice structure, such as changes in icing and the formation of sedimentary, metamorphosed, or sublimation ice.

Since the first discoveries of these unique natural beauties, researchers have noticed that the ice fill is subject to change over time (Șerban et al., 1948; Racoviță et al., 1987; Ohata et al., 1994; Racoviță, 1994; Fuhrmann, 2007; Tulis and Novotný, 2003; and others). Their research focuses primarily on the morphology and the dynamics of the ice body’s shape changes depending on the climatic conditions prevailing underground, but also on their age dating, genesis, and geographic location. Awareness of the world’s ice caves was relatively small until the 19th century and narrowly geographically confined to Central Europe and Russia. Ice caves in Jura and the Western Alps (now France and Switzerland) have been known for a long time, but Dobšiná Ice Cave, together with Demänovská Cave (Slovakia) and Pestera Scărisoara Cave (Romania), are among the first caves discovered in the 19th century in Central Europe outside the Alpine regions (Meyer, 2018). As stated by Meyer (2018), the history of ice cave exploration in the distant past (19th-20th century) was devoted to the unification of heterogeneous nomenclature, to the description of phenomena influencing the formation of ice in caves - depending on the airflow pattern, climatic conditions (defining the relationship between internal and external climatic conditions (Korzystka et al., 2011), microclimate, climate in front of the cave, as well as regional macroclimate), cave shapes. Many of these theories have been contradicted and disproved over time.

In the past, before electronic rangefinders become more widespread, ice level was typically measured as the distance from a fixed point (monitoring datum) on the cave wall or ceiling above the ice floor (marked by a screw or nail permanently inserted into the rock) to the surface of the ice floor (Smith, 2014). When multiple observations from a cave were available for a given year, the data offered the opportunity to model inter-annual variation in ice height. Modern geodetic (Gómez-Lende et al., 2014; Securo et al., 2022) and geophysical (Behm and Hausmann, 2008; Podsushin and Stepanov, 2008; Hausmann and Behm, 2011; Gómez-Lende et al., 2016; Securo et al., 2022) tools provide quality results for surveying ice-fillings dynamics. Behm and Hausmann (2008) examined several caves in the Northern Calcareous Alps of Austria, where ground-penetrating radar (GPR) was used to determine the thickness, volume, and internal structure of ice fillings, revealing that cave ice is heterogeneous. Hausmann and Behm (2011) examined the internal structure and formation of ice deposits in four caves in the Northern Calcareous Alps using GPR, revealing non-uniform ice with varying basal reflection signatures and banded internal reflections, likely caused by sediment layers, and suggesting temperate ice with a homogenous liquid water content of approximately 2%. Gómez-Lende et al. (2014) focused on the ice caves of the Picos de Europa (Spain), where terrestrial laser scanning (TLS) was used to monitor ice cave evolution and morphology, revealing dynamic temperature zones within the Peña Castil ice cave, distinct seasonal ice accumulation linked to snowmelt, and the identification of unique ice structures and cryospeleothems, such as radicular crystallization hoarfrost, observed for the first time in the region. In another study Gómez-Lende et al. (2016) used GPR to investigate the internal structure, stratification, and ice volume of the Peña Castil ice cave in northern Spain, revealing a minimum ice thickness of 54 m, potential vertical displacement within the ice, and banded reflections from interbedded clasts and sediment layers, demonstrating georadar’s effectiveness in providing insights into ice cave dynamics and internal features where direct visual inspection is challenging. Securo et al. (2022) explored the use of a terrestrial structure-from-motion multi-view stereo approach combined with GPR to monitor topographic changes and estimate ice mass balance in caves in the Julian Alps, offering a low-cost, efficient, and accurate method for long-term ice monitoring.

Digital tacheometry and, more recently, terrestrial laser scanners or digital photogrammetry tools allow data collection with high temporal and spatial resolution (Pukanská et al., 2023; Urban et al., 2024). These technologies have been highly proven in many research works (for instance, Hofierka et al., 2016; Pukanská et al., 2018; Šupinský et al., 2019; Pukanská et al., 2020; Bella et al., 2021). Pukanská et al. (2018) explored the modern surveying techniques for mapping underground morphological structures, emphasizing the use of TLS and digital close-range photogrammetry method of Structure-from-Motion (SfM), for accurate mapping of complex underground spaces. The study finds that while TLS provides the most accurate data, SfM offers comparable results with lower financial and time demands and better photo-texture quality, though it is limited by underground lighting conditions. In contrast, optical scanning produces less accurate results with more severe limitations, making TLS the most suitable technology for mapping, with SfM being viable under optimal lighting. Šupinský et al. (2019) used laser scanning technology to monitor the inter-seasonal dynamics of ice volume in the Silická ľadnica cave, revealing continuous changes with significant ice loss over time, and highlights the need for long-term observations to better understand the underlying mechanisms. In another work by Pukanská et al. (2020), authors mapped Ochtiná Aragonite Cave in Slovakia, known for its unique geological and mineralogical features, using TLS and digital photogrammetry, providing detailed 3D models and morphometric data that offer insights into the cave’s complex morphology and contribute to its environmental protection and tourism management. For the same cave, Bella et al. (2021), based on the same 3D data, revealed detailed depositional histories and slow sedimentation rates, with magnetostratigraphy and U-series dating tracing its development from the Early Pleistocene to the Late Pleistocene. Regarding modern surveying technologies, also mobile mapping systems can be utilized. However, they are more suitable for obtaining a general 3D structure of cave spaces and not for detailed ice surface mapping (Dušeková et al., 2024). The measurements have been used to detect bulk masses or to trace some ice masses but mainly to survey the geomorphological features of cave spaces - cartographic and topographic representation or the creation of geological maps (Petters et al., 2011; Milius and Petters, 2012; Berenguer-Sempere et al., 2014). Milius and Petters (2012) focused on the three-dimensional surveying and visualization of the Eisriesenwelt, one of the world’s largest ice caves, using 158 terrestrial laser scans and over 2,000 texture images, producing a realistic 3D model and virtual flythrough, with applications for scientific research, tourism, and cultural heritage documentation. On the other hand, Berenguer-Sempere et al. (2014) demonstrated the potential of combining TLS and thermographic techniques to create 3D thermographic models for monitoring climatological parameters in ice caves, providing a versatile and accurate method for studying ice evolution in challenging environments like the Peña Castil Ice Cave in Northern Spain.

Over time, as the technology and methodologies of topographic research have evolved, so have the knowledge and refinement of scientific results. From the early historical geodetic measurements with optical instruments and analog photogrammetry, we are now moving to technologies such as digital tacheometry and photogrammetry, TLS, or modern geophysical methods such as microgravimetry, GPR, and seismometry.

Given that Dobšiná Ice Cave is currently one of the most glaciated ice caves in the world (Perşoiu and Lauritzen, 2018; Hruby, 2022), it provides a wide range of different types of ice structures, with large areas of vertical (Ruffinyi’s Corridor, Great Curtain) and horizontal glaciation (Small and Great Hall), as well as areas with significant ice loss (Collapsed Dome), is significantly affected by climate change and precipitation activity, thus providing scope for the design of an accurate and comprehensive methodology for geodetic and geophysical measurement of morphological changes in different types of ice filling, taking into account current technological possibilities for spatial measurement, visualization, analysis and publication of data.

The Dobšiná Ice Cave (“Dobšinská ľadová jaskyňa” in Slovak) belongs to the most well-known ice cave in the world. It is located in the southern part of Slovenský raj (Slovak Paradise) National Park in the Spiš-Gemer Karst (north of the town of Dobšiná) (Figures 1A–C). Its entrance lies at 969 m a.s.l., on the right side of the Hnilec River valley, on the west slope of the Duča Plateau, 130 m above the valley bottom. The length of the cave is 1,483 m with a vertical span of 112 m. It is an inflow part of the larger genetic system of the Stratenská Cave, which is longer than 23.6 km. The whole system was formed by two underground streams - the Hnilec River and the Tiesňava Stream in the Late Pliocene (Tulis and Novotný, 1989; Novotný, 1993; Novotný and Tulis, 1996; 2005; Bella et al., 2014). The conditions for the formation of the ice filling were probably created in the Middle Pleistocene after the collapse of the ceilings (Duča sinkhole) and the breaking of the passage between Dobšinská Ice Cave and Stratenská Cave (Novotný and Tulis, 1996). At present, it is mostly filled with ice that reaches the ceiling in some places, and that divides the cavity into several parts: Small Hall, Great Hall, Collapsed Chamber, Ruffinyi’s Corridor, Ice tunnel, Ground Floor, and Hell. The ice filling of the cave is mainly made of floor and wall ice, ice stalagmites, and icefalls (Figures 1D, E). In 1995, a joint geodetic and geophysical survey of the ice volume and area was carried out, which revealed the ice volume was 110,132 m³, the surface of ice filling was 9,772 m², the maximal thickness of floor ice reached 26.5 m, and the average thickness of floor ice was 13 m (Novotný and Tulis, 1996; see also Bella, 2006; 2018; Bella and Zelinka, 2018; Bella et al., 2020). The surface area (in the vicinity of the cave) is in moderately cool (mean annual air temperature 4, 7°C; mean air temperature in January −5,4°C, in July 14, 2°C) and very humid subregion (mean annual precipitation total 900–1,000 mm) (according to Lapin et al., 2002; Faško and Šťastný, 2002; Šťastný et al., 2002).

Although the “ice hole” under the Duča had been known to shepherds and loggers for a long time, it was discovered only on 15 June 1870 by Eugene Ruffinyi and Co., (Lalkovič, 2000; Kudla, 2020). The cave was opened to the public on 8 March 1871. The first map of the Dobšiná Ice Cave was made by its discoverer Eugen Ruffínyi in 1871; later, new drawings were added. The first stereophotogrammetric surveys of the cave spaces were carried out in 1936. Another measurement was the research of spatial changes of the ice fill by geodetic and photogrammetric methods in the years 1976–1990, in which the Department of Mine Surveying and Geophysics of the Technical University of Košice (VŠT Košice) was also involved.

Different surveys of the cryosphere, and especially of ice caves, have been part of their documentation for many years. There are measurements of ice extent, ice surface characteristics, the volume of ice filling in caves, spatio-temporal dynamics of the ice bodies, and others. Analyses of the dynamics of the ice in caves have been presented in several papers (Strug and Zelinka, 2008; Perşoiu and Pazdur, 2011; Gómez-Lende and Sánchez-Fernández, 2018; Perşoiu et al., 2021). Also, the ice measurement itself has been widely reviewed and presented (Gindraux et al., 2017; Alfredsen et al., 2018; Pfeiffer et al., 2022; Hruby, 2022). Morard et al. (2010) recorded rapid changes in the Diablotins ice cave in Switzerland by monitoring the cave climate and suggested an hourly or daily monitoring interval based on a laser scanner or automatic cameras. The dynamics of ice bodies and their margins can also be monitored by data from a time-lapse camera and SfM photogrammetric processing (Mallalieu et al., 2017) or by UAS images and photogrammetric processing (Li et al., 2019). Over time, various technologies have been implemented and have been used with more or less success.

In this research, data from TLS proved to be a good and quick technique to get a general spatial overview of the cave and the extent of the glaciated part with sufficient accuracy (Figures 1, 8, 9). However, determining changes in the ice volume by TLS can be problematic due to the properties of the laser beam incident on the ice surface. For this purpose, it is more appropriate to use digital tacheometry, which can provide a spatial position of individual points on the ice surface with high accuracy and, therefore, create a planimetric and altimetric map of the ice surface for individual epochs. The comparison of tacheometric data from years 2018 and 2023 revealed a significant loss of ice area, especially in the Small Hall and Collapsed Dome parts (Figure 8). The vertical loss of the ice level over the 5 years amounts to −0.3 to −0.9 m. By comparing the amount of glaciation over this five-year period, it can be clearly stated that the ice volume has increased by 99.5 m³ in some places but decreased by up to 667 m³ in others. In 2018, the area covered by ice was 2,986 m². Detailed inter-annual and inter-semi-annual (seasonal) measurements of the change in floor ice height will be analyzed in a separate paper.

However, the TLS results show that using a laser scanner with different wavelengths, for instance, Leica RTC360 with infrared 1,550 nm, can provide more accurate results with less penetration into the ice (Figures 4A, B).

On the other hand, digital tacheometry is suitable for use only in parts with horizontal (floor) ice, where it is possible to measure the spatial position of points using a common surveying prism. For specific parts in terms of their geometry, spatial orientation, and texture/structure of the ice, digital close-range photogrammetry would be the best option. In Dobšiná Ice Cave, we identified several specific glaciated parts suitable for the photogrammetric measurements. Due to the extensiveness of the research, we presented results for two of them. However, digital photogrammetry (the SfM method) has proven to be very suitable for parts that are hard to survey using traditional surveying techniques (and/or TLS). Moreover, we applied a new photogrammetric imaging technique using a cross-polarized illumination to properly image and model highly reflective ice surfaces (see also Bartoš et al., 2023).

In terms of geophysical methods, both of them proved to be suitable and, what is more important, complementary to TLS, tacheometry, and photogrammetry since they can provide data which are not obtainable by geodetic methods (mainly the depth of the ice body and its structure under the surface).

This means, from the point of view of speleologists, it is also important to map the total volume of ice in the cave, which is, however, a methodologically more complex and demanding task. At this stage of the study of the Dobšiná Ice Cave, we are able, on the basis of the measurements so far, to at least tentatively assess the possibility of determining the total thickness of the ice filling based on the interpretation of the CBA.

In terms of GPR, it has also been used in ice surveys for a long time. Dallimore and Davis (1992) used GPR to detect massive ground ice and map near-surface geology. GPR was also used in combination with SfM photogrammetry to describe the topography and ice thickness and determine the volume changes of ice volumes in ice cave environments (for instance, Securo et al., 2022). The usability of GPR technology to study permafrost rocks and the emergence of cryogenic processes was evaluated by Sokolov et al. (2020).

In Dobšiná Ice Cave, from repeated measurements along identical lines (acquired in both directions), the accuracy of the ice thickness could be estimated as the approximate value ± 0.2 m. This error is relatively high when compared with other geodetic methods, but the transition between ice and basement is not a clear boundary but a zone where the basement rocks are broken by the influence of ice. For this reason, the repetition of measurements along each line is important. This kind of interpretation (estimation of ice thickness or the depth of ice basement) can be performed along a system of parallel and crossing lines, and an area map of these values can be plotted–contributing to a 3D model of ice cover in selected parts of the cave.

This research demonstrates that various geodetic (digital tacheometry, laser scanning, digital photogrammetry) and geophysical methods (microgravimetry, GPR) are necessary for a comprehensive study of ice filling dynamics in an ice cave, especially in terms of long-time monitoring. Ice bodies in caves can have different characteristics for individual parts of the cave, with changes in the structure and/or texture, even in a small area. Therefore, combining these methods can give us a comprehensive view of ice filling dynamics and a suitable tool for long-term monitoring. However, we see a potential for using modern laser scanners that are available to filter reflected laser beams according to the number of the returned signal. Using such, it is possible to measure these surfaces in a comparable density and accuracy as with digital photogrammetry.

In this study, we reviewed the current methodologies for surveying the ice filling and the thickness of the underground ice, and the dynamics of its evolution in the Dobšiná Ice Cave. We evaluated the suitability of using different geodetic and geophysical methods and technologies depending on the expected results. Due to the physical properties of the laser beam and its penetration/reflection from the ice surface, as well as limiting factors in photogrammetric imaging, the use of different methodologies should be considered.

We proposed the use of a digital tacheometric method to measure 3D data on the floor ice surfaces and a photogrammetric SfM method using cross-polarized light to measure the volumetric changes of ice in specific areas of the cave. The tacheometric data showed significant ice volume changes between 2018 and 2023, with vertical decrease of ice surface ranging from 0.3 to 0.9 m in areas of the Small Hall and Collapsed Dome. The total ice volume of the floor ice decreased by up to 667 m³ in some areas, while increasing by 99.5 m³ in others. The combination of surveying technologies also allowed us to detect dynamic changes in ice structures, such as a widening of the ice tunnel by 20 cm in the southeast direction and a volume loss of 9 m³ in the vertical ice wall of Ruffinyi’s Corridor.

We found the GPR, geophysical method, and partial microgravimetry to be clearly successful in measuring the thickness of the ice. GPR measurements revealed that the ice thickness varies from 10 to 25 m, with an estimated error of ±0.2 m, providing critical data about the subsurface ice structure.

All of the technologies utilized have their accuracy limitations depending on the instruments, methods of measurement, and characteristics of the environment. However, the presented methodology can combine all of their advantages with comparable accuracy to give a general view of the behavior of this underground ice body. Nonetheless, modern ice cave research has not reached its zenith, and there is still room for improvement - depending on geodetic and geophysical measurement technology development. Since the underground ice body is characterized by its dynamic inter-annual changes, dependent on climatic developments in both the exterior and interior of the cave, as well as anthropogenic factors, there is a constant need for regular monitoring in order to preserve this unique natural phenomenon as much as possible. This research shows the high applicability and suitability of the combination of these methods and technologies to monitor spatio-temporal changes in ice caves over a long period.