Summary Results of Radon-222 Activity Monitoring in Karst Caves in Bulgaria

Abstract

Cave systems are a kind of natural laboratory for interdisciplinary research on karstogenesis in the context of global changes. In this study, we investigate the concentration of ²²²Rn at 65 points in 37 representative caves of Bulgarian karst through continuous monitoring with passive and active detectors with a duration of 1 to 13 years. The concentration changes strongly both in the long term and seasonally, with values from 0.1 to 13 kBq m⁻³. These variations are analyzed from different perspectives (location and morphological features of the cave system, cave climate, ventilation regime, etc.). The seasonal change in the direction and intensity of ventilation is a leading factor determining the gas composition of the cave atmosphere during the year. Parallel measurements of ²²²Rn and CO₂ concentrations in the cave air show that both gases have a similar seasonal fluctuation. Cases of coincidences of an anomalous increase in the concentration of ²²²Rn with manifestations of seismic activity and micro-displacements along tectonic cracks in the caves have also been registered. The dependencies between the ²²²Rn concentration in the caves and in the soil above them are also discussed, as well as the possible connections between global trends in climate change and trends in ²²²Rn emissions. Special attention is paid to the risks of radiation exposure in show caves. A calculation procedure has been developed to achieve the realistic assessment of the effective dose of cave guides. It is based on information about the annual course of the ²²²Rn concentration in the respective cave and the time schedule of the guides’ stay in it. The calculation showed that the effective dose may exceed the permitted limits, and it is thus necessary to control it.

1. Introduction

Karst is spread over one-quarter of the territory of Bulgaria in the altitude range from 2914 m a.s.l. (Pirin Mountain) to −70 m below the actual sea level of the Black Sea (on the northern Bulgarian coast). It is formed in carbonate rocks (limestones, dolomites, marbles and their transitional types) with an age from the Proterozoic to the Neogene. They participate in the creation of the main morphological regions in the country, in which natural factors, during the evolution of the relief, have developed different types of karst. Over 6500 caves have been studied and mapped from the underground karst in Bulgaria [1], among which the longest (24,044 m) and deepest (−561 m) is the Kolkina Dupka Cave in the Balkan Mountains.

The classical karst, which is most widespread in Bulgaria, is formed by the dissolution of carbonate rocks by aggressive waters containing carbonic acid. Therefore, the main system-forming role in karstogenesis is played by the water cycle and the carbon cycle and the karst processes associated with them. They reach great depths and form specific geosystems with a well-defined vertical structure in the rock massifs. It includes two subsystems: surface (with a typical karst relief) and underground (with karst cavities and caves). There is high structural permeability between them, and paradynamic and paragenetic material–energy interactions take place, which are the basis of the functioning and dynamics of the karst geosystem [2,3,4]. It is typical for it to be delayed in the reaction of the underground subsystem to the processes and events on the surface and vice versa (relative dynamic autonomy). Another of its features is the frequent mismatch of the boundaries between the surface and underground subsystems. These features define karst geosystems as some of the most complex and sensitive ones. The high structural permeability between subsystems makes them highly vulnerable, so they are exposed to an increased risk of impacts, especially against the backdrop of increasing global changes, some specific global and local risk factors being temperature changes, water regime, pollution penetration and anthropogenic influences [3,4]. In order to ensure sustainable development of karst geosystems, interdisciplinary research is needed within the framework of the geosystem approach described in previous publications of the authors, e.g., [3]. In this approach, all elements of the karst geosystem are equivalent, and the relationships between them are studied. Based on the geosystem approach and many years of research experience, the Experimental Laboratory of Karstology (ELK) at the National Institute of Geophysics, Geodesy and Geography of the Bulgarian Academy of Sciences (NIGGG-BAS) developed an original methodological platform called ProKARSTerra [5]. It is applied in karst research in Bulgaria and includes integrated monitoring of karst geosystems (MIKS). It provides continuous objective information on the state and reactions of geosystems to impacts—both anthropogenic and various extreme natural phenomena and changes of a global nature. MIKS is also a basis for studying the feedback loop—the role of karst in global changes [3,4]. Due to the specificity of the underground part of karst geosystems, Speleo-MIKS integrated monitoring of cave karst systems has also been developed within the framework of MIKS.

MIKS is carried out in three variants [3]:

• Periodic measurements with portable instruments at fixed points, selected according to the structure of the studied karst geosystems;
• Installation of stationary instrumental equipment for continuous measurements in cave systems and in their adjacent territories;
• Remote uses of the available results of measurements, including from instruments mounted on satellites in orbit around the Earth.

The widespread distribution of karst in Bulgaria and its unique diversity make it a natural laboratory for interdisciplinary and monitoring studies of karstogenesis. These are concentrated in model karst areas, representative of the main types of karst in Bulgaria, and they include the activity of radon-222, especially given its high concentration in underground cavities [6].

Radon-222 (hereafter referred to as radon) is a radioactive noble gas which occurs commonly in the environment and makes up about 70% of natural background radiation or 50% from all sources of irradiation [7]. It is a member of the uranium–radium family—a product of the alpha decay of the radium isotope ²²⁶Ra, which is present in almost all rocks that make up the Earth’s crust. The half-life of radon is T_1/2 = 3.8 days (E_α = 5.6 MeV), in which the short-lived products ²¹⁸Po (E_α = 6 MeV) and ²¹⁴Po (E_α = 7.7 MeV) are formed, with half-lives of less than 30 min. They are very active radioactive sources and are distributed attached to aerosol particles or independently as an unbound fraction [8]. Their prolonged inhalation poses a serious health risk of radiation exposure, causing lung cancer [9,10].

The typical average concentration of radon in ground atmospheric open air is 10 Bq.m⁻³. Its main source is soil gas. Radon, like carbon dioxide (CO₂), is a heavier gas than air. This suggests that in underground spaces, such as mines and karst cavities (natural reservoirs), especially if they are poorly ventilated, significant amounts of radon can accumulate. It enters cave systems through two processes: emanation and exhalation. Their intensity depends on the structure and cracking of the rocks, structure and humidity of the soil, atmospheric and hydrological conditions, ventilation in the caves, etc.

The concentration of radon in karst caves depends on their morphology, the location and altitude of the entrance(s), the cave climatic zones and the cave ventilation regime, which also determines the seasonal dynamics in the volumetric activity of radon [3,11]. As a substance potentially hazardous to human health, radon is included in international documents on radiation protection [12], including also the Bulgarian Regulation on radiation protection [13]. According to the published summarized data of A.A.Cigna [6], in 90% of the 303 tourist caves cited by the author around the world, the radon activity concentration is between 3 and 10 kBq.m⁻³. High concentrations of radon have also been found in Bulgarian show caves [11,14].

The important scientific and applied importance of radon is the reason why it is also the subject of speleological research and monitoring. The investigations are mainly focused on two areas:

• Research of radon as a tracer for air movement inside caves (e.g., [15,16,17,18,19,20,21,22,23], etc.);
• Radiation risks and protection of people working or visiting in caves (e.g., [8,24,25,26,27,28,29,30,31,32,33,34], etc.)

Radon studies in Bulgarian caves have been conducted since 1974. The first study was carried out by a team from the Institute of Radiobiology and Radiation Hygiene, Sofia [35]. It includes nine caves, of which five are show caves (Bacho Kiro, Orlova Chuka, Snezhanka, Magura and Ledenika). The concentration of radon was measured by the ionization method with an SG-11 emanometer, and the level of potential alpha energy of the Rn progeny was determined by a PB-4 aerosol radiometer, a portable instrument based on pumping a fixed amount of air through an aerosol filter and a subsequent measurement of alpha activity of the collected aerosols by a ZnS(Ag) scintillator; their beta activity is measured by a thin-window GM counter. The authors of the study found that in 97% of the measurement points in the caves, the results obtained exceeded the maximum limits for polymetallic and uranium mines accepted at that time in Bulgaria (370 Bq.m⁻³ and 1110 Bq.m⁻³, respectively). Despite the published alarming findings, no new studies were undertaken in the tourist caves in the following years, and no control of radon exposure in caves as workplaces was introduced. Additional studies of the radiation background and air ionization were conducted only in the Magura Cave, where a cave sanatorium for the treatment of bronchial asthma operated in 1974–1975.

A new method for determination of the radon concentration in air using CD/DVD plates as passive track alpha detectors was developed by D.Pressyanov and his group at the Faculty of Physics of the Sofia University in the first decade of the 21st century [36]. They are exposed to direct radiation, and no cameras are used. This method has also been experimented in two Bulgarian caves in the Vitosha mountain.

Within the framework of bilateral academic cooperation, since 2007, ELK and with the Department of Radiation Dosimetry, Nuclear Physics Institute, Czech Academy of Science (DRD NPhI—AS CR) have been developing and have successfully implementing a methodology for cave monitoring of radon. During the first 2 years, diffusion chambers with track detectors for passive radon measurements were designed and tested. The exposure time was optimized according to the morphological and cave climatic features and seasonal dynamics in the caves [11,14]. This allowed for radon activity to be included in the scientific program of ELK for MIKS in model karst geosystems [3]. For this purpose, since 2011, the construction of a specialized network for monitoring the activity of radon, BGSpeleo-RadNet, has been underway [5]. It is jointly maintained by the ELK and DRD NPhI and already includes 46 karst caves (including 13 tourist ones) with 85 measurement points. Caves with a duration of radon monitoring of over 1 year are the subject of this study (see Section 2).

BGSpeleo-RadNet provides objective information in three scientific and applied directions:

• Interactions between natural radiation processes and modern karstogenesis;
• Connection between radon and seismotectonic activity;
• Radiation protection of workers in karst show caves.

Since 2016, instrumental radon monitoring (periodic or continuous) has been carried out in some of the caves, which is described in Section 3.

The aim of this article is to present and analyze the summarized results of the conducted radon monitoring in Bulgarian karst caves. On this basis, conclusions have been drawn about the interactions between radon activity and modern karstogenesis in the conditions of different types of karst and cave morphology. Special attention is paid to the risks of radiation exposure in show caves, and scientifically substantiated prevention measures are proposed.

2. Objects of the Study

In 37 of the karst caves in the BGSpeleo-RadNet network, radon activity monitoring has lasted for over 1 year. These caves are the subject of the present study (Figure 1). In 21 of them, the monitoring was carried out from 1 to 5 years, and in 5 caves, monitoring was carried out over 10 years. The selected caves have a very diverse morphology and water regime (Table 1) and are representative of the different types of karst in Bulgaria, as well as the different natural and geographical regions in which they were formed. The entrances of the caves are at altitudes ranging from 126 m a.s.l. (No. 3 Moesian Plain) to 2525 m a.s.l. (No. 4 in Pirin Mountain). Most of the caves (23) are in the lowland–hilly altitudinal belt (up to 600 m a.s.l.). In the low mountain belt (600–1000 m a.s.l.), there are six caves, in the mid-mountain belt (1000–1600 m a.s.l.) there are seven, and in the high mountain belt (over 1600 m a.s.l.), there is only one. Some general characteristics of the caves are shown in Table 1.

The activity of radon in cave systems is also strongly influenced by the climate on which the cave climate and the cave ventilation regime depend. The Balkan Peninsula has a transitional climate between oceanic and continental and between subtropical and temperate [37]. The territory of Bulgaria falls into two main climatic zones, the border between which is the Stara Planina Range. Northern Bulgaria has a transitional climate between a subtropical and temperate climate, and southern Bulgaria has a subtropical climate. Territories with an altitude above 800–1000 m have a mountainous variant of these climates (with lower temperatures and more precipitation) and are distinguished by less favorable conditions for karstogenesis. Of the caves subject to this study, 20 are under the influence of a subtropical climate, of which 12 are under the influence of its mountainous variant. Seventeen of the caves fall into the transitional zone between a subtropical and temperate climate, of which only two have a mountainous climate. Trends in recent climate change in the Balkan Peninsula suggest an increase in the intensity of karst processes [38], but their impact on radon activity in cave systems also depends on many other factors. The ventilation regime is the leading factor, but in addition to the climate, it also depends on the cave morphology. The cave systems in which radon monitoring is carried out are representative of different morphological types, which is evident from Table 1. According to the dimensions of the studied galleries, 16 of the caves are over 1000 m long, and 3 have a difference in elevation over 100 m. Multi-storey caves predominate (20 in total). There are 12 complex labyrinthine caves, 12 sub-horizontal caves, and 13 abyssal caves (cascade). According to cave genesis, 20 of the caves are sinkholes (abyssal), 14 are spring caves, and 3 are sinkhole–spring caves. Most of the caves (23) have underground waters (cave rivers and lakes). The caves are also distinguished by their location and the number of entrances, which also control the ventilation regime. Seventeen of the caves are multi-entrance, and in eight of them, additional entrances were dug during the construction of tourist infrastructure (No. 1, 8, 9, 25, 26 and 31) or during the capture of cave water for water supply (No. 18). The entrance to the Venetsa Cave (No. 2) was discovered during quarry development.

Table 1. General information about the karst caves, the object of the study ([1,39] and authors’ own data).

Cave and Status 1	Cave Location	Cave Entrance/s	Cave Dimensions, m	Type of Rock	Cave Morpho- and Hydro-Characteristics 2
(Number as per Figure 1)	(v.—Village; c.—City; m.—Municipality)	(height, m a.s.l.)	(total length/denivelation)	(rock age)
MOESIAN PLAIN				Limestone:
1. Magura—ShC (1961); NM (1960)	v.Rabisha, m.Belogradchik	371/364	3645/+44, −25	(J3-K1)	AC, HC, LC, E-3 (2) // DryC
2. Venetsa—ShC (2015); NM (1971)	v.Oreshets, m.Dimovo	350	220/−28	(J3-K1)	AC, CC, MC, E-1 (1) // DryC
3. Devetashka Cave—ShC (2013); NM (1996)	v.Devetaki, m.Lovech	126	2442/+96, −25	(K1)	SC, HC, E-8 // CR and DryC
4. Mandrata	v.Chavdartsi, m.Lovech	197	530/+12	(K1)	SC, HC, E-1 // DryC and CR
5. Urushka Maara	v.Krushuna, m.Letnitsa	180	1700/+25	(K1)	SC, HC, MC, E-1 // CR, CL
6. Boninska Dupka	v.Krushuna, m.Letnitsa	269	5550/+33, −22	(K1)	SC, HC, MC, E-1 // CR, CL
7. Vodopada Cave—in NM (1995)	v.Krushuna, m.Letnitsa	215	2225	(K1)	SC, HC, LC, E-1 // CR, CL
8. Orlova Chuka—ShC (1942, 1961); NM (1962)	v.Pepelina, m.Dve mogili	170	13,437/+33, −12	(K1)	AC-SC, HC, LC, E-2 (1) // DryC
9. Bisserna Cave—ShC (2019); in NM (1980)	c.Shumen, m.Shumen	404/411	2716/+19,5	(K2)	SC, HC, CC, MC, E-2 (1) // CR, CapC
STARA PLANINA MOUNTAIN				Limestone:
10. Ledenika—ShC (1961); NM (1960) in NP	c.Vratsa, m.Vratsa	835	320/+16, −21	(J3)	AC, DC-UC, CC, MC, E-2 // CL, DryC
11. Temnata Dupka—NM (1962) in NP	v.Milanovo, m.Svoge	398	9000/+33, −21	(T2)	SC, CC, LC, MC, E-1 // DryC and CR, CL
12. Kolkina Dupka	v.Zimevitsa, m.Svoge	1290	24,034/−561	(T2)	AC, CC, LC, MC, E-1 // DryC and CR
13. Saeva Dupka—ShC (1967); NM (1962)	v.Brestnitsa, m.Yablanitsa	500	230/+3, −17	(J3-K1)	AC, DC-UC, CC, MC, E-1 // DryC
14. Gradezhnishka Cave—NM (1962)	v.Glogovo, m.Teteven	456	908/+15	(T-J)	SC, HC, E-2 // CR, CapC
15. Emenska Cave—in NM (1980)	v.Emen, m.Veliko Tarnovo	249	3113/+8, −40	(K1)	AC, CC, LC, MC, E-1 // DryC and CR
16. Bacho Kiro Cave—ShC (1937); NM (1962)	c.Dryanovo, m.Dryanovo	350	3600/+65	(K1)	AC-SC, LC, MC, E-1 // DryC
17. Andaka	c.Dryanovo, m.Dryanovo	250	5000/+45	(K1)	SC, HC, LC, MC, E-1 // CR, CapC
UPPER THRACIAN LOWLAND				Limestone:
18. Chirpan Bunar—NM (1974)	v.Granit, m.Bratya Daskalovi	155	445/+3	(Pg2)	SC, HC, E-2 (1) // CR, CapC
STRANDZHA MOUNTAIN				Marbled limestone (J)
19. Labyrinth Cave—in NP (1995)	v.Stoilovo, m.Malko Tarnovo	263	202/−9		AC, DC, LC, E-1 // DryC
20. Bazat—in NP (1995)	v.Stoilovo, m.Malko Tarnovo	265/270	58/−5		AC, HC, E-2 // DryC
21. Golyamata Vapa—in NP (1995)	v.Stoilovo, m.Malko Tarnovo	300	450/−125		AC, CC, LC, E-1 // DryC and CR
22. Varadat—in NP (1995)	v.Stoilovo, m.Malko Tarnovo	280	43/−12		AC, DC-UC, E-1 // DryC
23. Brezhanka—in NM(2007) in NP(1995)	v.Mladezhko, m.Malko Tarnovo	200	67/−5		AC, CC, E-3 // CL
PIRIN MOUNTAIN				Marble (Pt)
24. Abyss No. 9–11—in NtP(1962)	c.Bansko, m.Bansko	2525	2200/−407		AC, CC, MC, E-2 // DryC
WESTERN RHODOPE MOUNTAINS				Marble (Pt)
25.Yagodinska Cave—ShC (1982); in NM (1971)	v.Yagodina, m.Borino	990,997/1020	10,500/ +36		AC-SC, CC, LC, MC, E-2 (2) // CR and DryC
26. Dyavolsko Garlo—ShC (1977); in PL (1963)	v.Trigrad, m.Devin	1054/1150	548/−89		AC-SC, CC, MC, E-2 (1) // CR
27. Ukhlovitsa—ShC (1983); NM (1979)	v.Mogilitsa, m.Smolyan	1040	550/−25		AC, CC, MC, E-1 // DryC
28. Goloboitsa-1	v.Koshnitsa, m.Smolyan	860	2145/+22, −44		SC, HC, MC, E-2 // CR and DryC
29. Nadarska Cave	v.Nadartsi, m.Smolyan	1151	55/+10		AC, HC, E-1 // DryC
30. Lepenitsa—ShC (2010); NM (1962)	c.Rakitovo, m.Rakitovo	990	2000/+21		SC, HC, MC, E-2 // CR and DryC
31. Snezhanka—ShC (1966); NM (1961) in BR	c.Peshtera, m.Peshtera	880	348/−18		AC, DC, MC, E-2 (1) // DryC
32. Novata/Starata Cave—in BR (1961)	c.Peshtera, m.Peshtera	550	955/+7		SC, HC, LC, E-2 //DryC and CL
33. Yubileyna—NM (2003) in BR (1961)	c.Peshtera, m.Peshtera	580	814/+3, −18		AC, MC, E-1 //DryC and CR
34. Gargina Dupka—in NM (2003)	v.Mostovo, m.Asenovgrad	940	525/+38		SC, CC, MC, E-2 // CR and DryC
35. Ivanova Voda—in PL (1980)	v.Dobrostan, m.Asenovgrad	1337	695/−113		AC, CC, E-1 // CL
36. Topchika—NM (1970) in BP (1962)	v.Dobrostan, m.Asenovgrad	982	727/−61		AC, CC, MC, E-1 // DryC
37. Chelevechnitsa	v.Orehovo, m.Chepelare	1130	329/+3, −12		AC, DC, E-1 // DryC

¹ Cave status (announcement year): ShC—show cave; NM—natural monument; PL—protected locality; BR—biosphere reserve; NP—natural park; NtP—national park; ² Morphology: abyssal cave—AC; spring cave—SC; sub-horizontal cave—HC; descending cave—DC; upwards cave—UC; cascading cave—CC; labyrinthine cave—LC; multi-storey cave—MC; entrance number—E-2 (1) (in brackets—artificial entrance) // Hydrology: cave river—CR; cave lake—CL; captured cave for water use—CapC; dry cave—DryC.

Table 1. General information about the karst caves, the object of the study ([1,39] and authors’ own data).

Cave and Status 1	Cave Location	Cave Entrance/s	Cave Dimensions, m	Type of Rock	Cave Morpho- and Hydro-Characteristics 2
(Number as per Figure 1)	(v.—Village; c.—City; m.—Municipality)	(height, m a.s.l.)	(total length/denivelation)	(rock age)
MOESIAN PLAIN				Limestone:
1. Magura—ShC (1961); NM (1960)	v.Rabisha, m.Belogradchik	371/364	3645/+44, −25	(J3-K1)	AC, HC, LC, E-3 (2) // DryC
2. Venetsa—ShC (2015); NM (1971)	v.Oreshets, m.Dimovo	350	220/−28	(J3-K1)	AC, CC, MC, E-1 (1) // DryC
3. Devetashka Cave—ShC (2013); NM (1996)	v.Devetaki, m.Lovech	126	2442/+96, −25	(K1)	SC, HC, E-8 // CR and DryC
4. Mandrata	v.Chavdartsi, m.Lovech	197	530/+12	(K1)	SC, HC, E-1 // DryC and CR
5. Urushka Maara	v.Krushuna, m.Letnitsa	180	1700/+25	(K1)	SC, HC, MC, E-1 // CR, CL
6. Boninska Dupka	v.Krushuna, m.Letnitsa	269	5550/+33, −22	(K1)	SC, HC, MC, E-1 // CR, CL
7. Vodopada Cave—in NM (1995)	v.Krushuna, m.Letnitsa	215	2225	(K1)	SC, HC, LC, E-1 // CR, CL
8. Orlova Chuka—ShC (1942, 1961); NM (1962)	v.Pepelina, m.Dve mogili	170	13,437/+33, −12	(K1)	AC-SC, HC, LC, E-2 (1) // DryC
9. Bisserna Cave—ShC (2019); in NM (1980)	c.Shumen, m.Shumen	404/411	2716/+19,5	(K2)	SC, HC, CC, MC, E-2 (1) // CR, CapC
STARA PLANINA MOUNTAIN				Limestone:
10. Ledenika—ShC (1961); NM (1960) in NP	c.Vratsa, m.Vratsa	835	320/+16, −21	(J3)	AC, DC-UC, CC, MC, E-2 // CL, DryC
11. Temnata Dupka—NM (1962) in NP	v.Milanovo, m.Svoge	398	9000/+33, −21	(T2)	SC, CC, LC, MC, E-1 // DryC and CR, CL
12. Kolkina Dupka	v.Zimevitsa, m.Svoge	1290	24,034/−561	(T2)	AC, CC, LC, MC, E-1 // DryC and CR
13. Saeva Dupka—ShC (1967); NM (1962)	v.Brestnitsa, m.Yablanitsa	500	230/+3, −17	(J3-K1)	AC, DC-UC, CC, MC, E-1 // DryC
14. Gradezhnishka Cave—NM (1962)	v.Glogovo, m.Teteven	456	908/+15	(T-J)	SC, HC, E-2 // CR, CapC
15. Emenska Cave—in NM (1980)	v.Emen, m.Veliko Tarnovo	249	3113/+8, −40	(K1)	AC, CC, LC, MC, E-1 // DryC and CR
16. Bacho Kiro Cave—ShC (1937); NM (1962)	c.Dryanovo, m.Dryanovo	350	3600/+65	(K1)	AC-SC, LC, MC, E-1 // DryC
17. Andaka	c.Dryanovo, m.Dryanovo	250	5000/+45	(K1)	SC, HC, LC, MC, E-1 // CR, CapC
UPPER THRACIAN LOWLAND				Limestone:
18. Chirpan Bunar—NM (1974)	v.Granit, m.Bratya Daskalovi	155	445/+3	(Pg2)	SC, HC, E-2 (1) // CR, CapC
STRANDZHA MOUNTAIN				Marbled limestone (J)
19. Labyrinth Cave—in NP (1995)	v.Stoilovo, m.Malko Tarnovo	263	202/−9		AC, DC, LC, E-1 // DryC
20. Bazat—in NP (1995)	v.Stoilovo, m.Malko Tarnovo	265/270	58/−5		AC, HC, E-2 // DryC
21. Golyamata Vapa—in NP (1995)	v.Stoilovo, m.Malko Tarnovo	300	450/−125		AC, CC, LC, E-1 // DryC and CR
22. Varadat—in NP (1995)	v.Stoilovo, m.Malko Tarnovo	280	43/−12		AC, DC-UC, E-1 // DryC
23. Brezhanka—in NM(2007) in NP(1995)	v.Mladezhko, m.Malko Tarnovo	200	67/−5		AC, CC, E-3 // CL
PIRIN MOUNTAIN				Marble (Pt)
24. Abyss No. 9–11—in NtP(1962)	c.Bansko, m.Bansko	2525	2200/−407		AC, CC, MC, E-2 // DryC
WESTERN RHODOPE MOUNTAINS				Marble (Pt)
25.Yagodinska Cave—ShC (1982); in NM (1971)	v.Yagodina, m.Borino	990,997/1020	10,500/ +36		AC-SC, CC, LC, MC, E-2 (2) // CR and DryC
26. Dyavolsko Garlo—ShC (1977); in PL (1963)	v.Trigrad, m.Devin	1054/1150	548/−89		AC-SC, CC, MC, E-2 (1) // CR
27. Ukhlovitsa—ShC (1983); NM (1979)	v.Mogilitsa, m.Smolyan	1040	550/−25		AC, CC, MC, E-1 // DryC
28. Goloboitsa-1	v.Koshnitsa, m.Smolyan	860	2145/+22, −44		SC, HC, MC, E-2 // CR and DryC
29. Nadarska Cave	v.Nadartsi, m.Smolyan	1151	55/+10		AC, HC, E-1 // DryC
30. Lepenitsa—ShC (2010); NM (1962)	c.Rakitovo, m.Rakitovo	990	2000/+21		SC, HC, MC, E-2 // CR and DryC
31. Snezhanka—ShC (1966); NM (1961) in BR	c.Peshtera, m.Peshtera	880	348/−18		AC, DC, MC, E-2 (1) // DryC
32. Novata/Starata Cave—in BR (1961)	c.Peshtera, m.Peshtera	550	955/+7		SC, HC, LC, E-2 //DryC and CL
33. Yubileyna—NM (2003) in BR (1961)	c.Peshtera, m.Peshtera	580	814/+3, −18		AC, MC, E-1 //DryC and CR
34. Gargina Dupka—in NM (2003)	v.Mostovo, m.Asenovgrad	940	525/+38		SC, CC, MC, E-2 // CR and DryC
35. Ivanova Voda—in PL (1980)	v.Dobrostan, m.Asenovgrad	1337	695/−113		AC, CC, E-1 // CL
36. Topchika—NM (1970) in BP (1962)	v.Dobrostan, m.Asenovgrad	982	727/−61		AC, CC, MC, E-1 // DryC
37. Chelevechnitsa	v.Orehovo, m.Chepelare	1130	329/+3, −12		AC, DC, E-1 // DryC

¹ Cave status (announcement year): ShC—show cave; NM—natural monument; PL—protected locality; BR—biosphere reserve; NP—natural park; NtP—national park; ² Morphology: abyssal cave—AC; spring cave—SC; sub-horizontal cave—HC; descending cave—DC; upwards cave—UC; cascading cave—CC; labyrinthine cave—LC; multi-storey cave—MC; entrance number—E-2 (1) (in brackets—artificial entrance) // Hydrology: cave river—CR; cave lake—CL; captured cave for water use—CapC; dry cave—DryC.

The caves, the objects of the study, have different statuses. The majority of them are protected areas, and 13 are tourist caves, with 7 (No. 13, 16, 25, 26, 27, 30 and 31) being managed by a non-profit legal entity, 5 (No. 1, 2, 3, 8 and 10) being managed by the municipal administration and 1 (No. 9) being managed by the directorate of a nature park [11,14]. Tourist caves have year-round access for visitors, with the exception of only three of them (No. 8, 9 and 30), which are open only during the warm half of the year.

The radon monitoring points in the cave systems were selected according to their morphology and cave climatic zones, determined through repeated expeditionary meteorological measurements along the cave profiles [3]. The accessibility of the points for conducting periodic instrumental measurements and replacing the detectors was also considered. In the tourist caves, monitoring points were also located along the tourist routes in order to determine the radioactive risks from radon activity to visitors and cave guides. The power supply and security of the tourist caves made it possible to organize continuous instrumental monitoring of radon in five of them (No. 9, 13, 16, 25 and 27, Figure 1).

3. Methods

Monitoring of radon activity in Bulgarian karst caves is carried out with 2 types of measurements, using the following:

• Passive detectors (cumulative, with a longer exposure period, usually up to several months). The results are reported with some delay (transportation, treatment and evaluation of detectors).
• Instrumental (active) devices (discrete, short-term, from several hours to several days, or continuous monitoring, from several months to several years and more). Data is obtained instantaneously (in situ or on-line).

3.1. Passive Measurements of Radon Activity Concentration

Track-etch passive detectors (TED, [40]) are used in this work. These detectors are commonly used for long-time monitoring (e.g., [25,26,27,41]) because they have proven to have long-term resistance to the very high humidity often seen in caves. These are integral detectors, and the method provides average concentration values during the exposure time. Because of the simple design and low cost, they have advantages for screening when it is necessary to cover a large number of locations simultaneously and repeatedly.

During exposure, the detectors were placed inside plastic chambers. Based on previous experience, small (~20 mL, see Figure 2) cylindrical plastic vials with a thin window were used.

The window area of 2 cm² was closed hermetically by thin (6 µm) polyethylene foil [11]. This foil enabled the penetration of radon into the measuring space by diffusion whilst stopping the Rn daughter products. α-Particle-sensitive CR-39 plastic discs (Page Track Analysis Systems Ltd., Bristol, BS49 4RF, UK) with a diameter of 16 mm and a thickness of 0.35—0.55 mm were used as track detectors. Exposure intervals in the studied caves varied between 1 week and 4–5 months (average of approximately 6 weeks). Processing of the exposed detectors was carried out using a combined chemical and electrochemical etching method, increasing the primary alpha particle tracks [42]. The above-described measurement arrangement was calibrated in the Rn chamber of the National Institute for Nuclear, Chemical and Biological Protection, Kamenná, near Příbram, Czech Republic [43], at different values of the time integral of Rn activity within 8–35 kBq.m⁻³.h. The response R was 2.37 ± 0.35 tracks.cm⁻² per kBqm⁻³.h, i.e., 56.9 ± 8.3 per kBq.m⁻³.d.

The exposed and etched detectors (Figure 2) were scanned using an Epson Perfection 4990 Photo high-resolution (4800 dpi) scanner, and α-track densities ρ, in tracks per cm², were estimated from binary pictures using the ImageJ software (version 1.54k) [44]. In the case of track densities >5000 cm², the effectiveness of the track counting by the SW decreased due to track overlapping; therefore, the experimentally determined correction function was used:

ρ = ρ_sw × (1 + (6.321 × 10⁻⁶) × ρ_sw + 1 × 10⁻⁹ × ρ_sw²)

(1)

where ρ_sw is the track density obtained by Image J.

The mean value of the radon activity concentration over the period of exposure t_exp (days), hereafter referred to as A_V,exp (kBq.m⁻³), was then calculated according to the following formula:

$${\mathrm{A}}_{\mathrm{V},\mathrm{e}\mathrm{x}\mathrm{p}}={\displaystyle \frac{\mathsf{\rho }}{\mathrm{R}.{\mathrm{t}}_{\mathrm{e}\mathrm{x}\mathrm{p}}}}$$

(2)

Since May 2017, monitoring of radon in the soil gas above the Saeva Dupka Cave has been organized with passive track detectors. Three points in the soil above the Concert Hall were selected, in which small-diameter pipes with a tightly closing upper opening were drilled in. TEDs were lowered into the pipes to a depth of 10 to 30 cm by two methods: in the diffusion chambers and in Al holder fixings, two pairs of detectors with an air gap of 1 mm and 7 mm between them were used, respectively [45]. The duration of exposure of the detectors was the same as in the cave.

The relative uncertainty of TED measurements was ~15–20%, considering a relative statistical deviation of ≤5% (at least 400 tracks collected on each detector) and a relative response deviation of 14.7% (R = 2.37 ± 0.35 tracks.cm⁻² per 1 kBqm⁻³.h).

3.2. Instrumental Measurements of Radon Activity Concentration

The track-etch passive detectors used are not suitable for detailed short-term measurements. Therefore, instrumental measurement of radon activity concentration was also applied in some of the studied caves. Portable AlphaE instruments (Bertin Technologies, 78180 Montigny-le-Bretonneux, France) were used. Objects of periodic measurements (with an interval of 10 min) lasting from 1 to 2 days were used in 5 caves (No. 2, 9, 10, 11 and 14, Figure 1) in 2016 and in 2 caves (No. 10 and 27, Figure 1) in 2022. The measurements were concentrated at the points also used for the measurements with TED. For the spatial distribution of radon in the cave systems, instrumental measurements were also carried out along the profiles of the caves. Since 2018, continuous monitoring with AlphaE has been organized in Saeva Dupka (Concert Hall), which continued until 2023, and in Bacho Kiro (Pop Hariton Hall), which continues in 2025. The data recorded by the instruments are downloaded in situ every 2–3 months. At that time, the instrument’s battery is recharged. Instrumental measurements provide data on the diurnal fluctuations in radon concentration and the influence of short-term ventilation pulses in the cave systems.

At the suggestion and with the assistance of the ELK, in November 2022, a team from the University of Plovdiv joined in with continuous instrumental measurements of radon in 4 of the caves in the BGSpeleo-RadNet network: Saeva Dupka, Ukhlovitsa and Bisserna (since November 2022) and the Yagodinska Cave (since February 2024). For this purpose, a specialized METER.AC instrument is used, equipped with an AlphaSensor^® device (RadonTec GmbH, 89426 Wittislingen, Germany), involving a 5 cm alpha-ray detector based on a Lucas cell (Figure 3A). The AlphaSensor device, together with other sensors (for temperature, atmospheric pressure and relative humidity, as well as a small GM counter to measure the background beta and gamma radiation and the front-end electronics as well), is mounted inside a perforated plastic container aligned with a 12 μm PP foil to prevent Rn progeny and water droplets from entering (Figure 3B). At the same time, radon gas enters freely into the internal volume by diffusion. The instruments are installed at one of the main points in the caves, where radon monitoring is carried out with passive detectors, and in the Bisserna cave, they are installed at two points, which are located in the upper dry and lower wet cave floors. The measurements are carried out in the continuous mode, with an interval of 10 min. The raw data is available online at [46].

3.3. Comparison Between Passive and Instrumental Measurements of Radon Activity Concentration

The basis for the comparison is the parallel measurement of the radon activity concentration in the caves using both methods (passive with TED and instrumental). They were carried out periodically at the same points and showed good agreement (Pearson’s correlation coefficient r = 0.887, TED_av/METER.AC_av = 0.966) in the Saeva Dupka Cave (Figure 4). The situation is similar in the Bacho Kiro and Uhlovitsa Caves, where similar comparative measurements were also made.

4. Results and Discussion

4.1. Concentrations and Regime of Radon Activity in the Studied Caves

The summarized results for the radon activity concentration in the karst caves, which were subject to monitoring, are shown in

Table 2. An overview of Rn monitoring results in Bulgarian karst caves.

Cave (Number as per Figure 1)	Cave Monitoring Point	Cave Climatic Zone	Time	Concentration Lowest	Highest	Annual Average
			Years	Bq.m−3	Bq.m−3	Bq.m−3
MOESIAN PLAIN
1. Magura	point 1	Transitional	2.7	423	687	569
	point 2	Internal	5.4	397	982	640
2. Venetsa	point 1 (Office)	Entrance zone	5.0	124	3472	897
	point 2	Transitional	5.7	145	6887	2177
	point 3	Internal	5.6	202	4459	1830
3. Devetashka Cave	point 1	Transitional	2.2	94	184	111
4. Mandrata	point 1	Transitional	3.8	87	330	169
5. Urushka Maara	point 1	Transitional	2.4	236	420	327
6. Boninska Cave	point 1	Transitional	2.3	337	2158	621
	point 2	Internal	6.5	897	2678	1470
7. Vodopada Cave	point 1	Transitional	5.5	189	5973	1169
	point 2	Internal	11.4	204	7774	1212
8. Orlova Chuka	point 1	Transitional	4.5	507	1939	1247
	point 2	Internal	2.3	927	3224	1805
	point 3	Internal	2.1	738	3012	1716
9. Bisserna Cave	point 1	Transitional	5.9	130	1971	677
	point 2	Internal	6.5	317	5903	1889
	point 3	Internal	3.9	748	3839	1518
STARA PLANINA MOUNTAIN
10. Ledenika	point 1	Transitional	10.7	167	2559	572
	point 2	Internal	5.8	163	2591	534
	point 3	Internal	5.6	264	5783	2432
11. Temnata Dupka	point 1	Transitional	1.9	460	1516	788
12. Kolkina Dupka	point 1 (−112 m)	Transitional	1.5	893	3515	1383
	point 2 (−213 m)	Internal	1.5	795	2385	1090
	point 3 (−267 m)	Internal	1.2	895	1900	1101
13. Saeva Dupka	point 1 (outside)	Entrance zone	10.1	8	200	38
	point 2	Transitional	12.8	104	3382	984
	point 3	Transitional	12.6	112	4061	1284
	point 4	Internal	13.5	176	13,376	1792
	point 5	Internal	13.5	167	4231	1214
14. Gradezhnishka Cave	point 1	Transitional	1.7	314	1466	1111
15. Emenska Cave	point 1	Transitional	4.2	104	862	416
16. Bacho Kiro Cave	point 1	Transitional	8.2	168	1677	672
	point 2	Internal	8.5	177	3118	493
	point 3	Internal	11.2	676	3949	1723
	point 4	Internal	12.6	629	6136	2472
17. Andaka	point 1	Transitional	2.4	137	1250	761
UPPER THRACIAN LOWLAND
18. Chirpan Bunar	point 1	Transitional	5.7	548	2917	1035
STRANDZHA MOUNTAIN
19. Labyrinth Cave	point 1	Internal	1.6	460	7937	3851
20. Bazat	point 1	Transitional	2.0	127	785	253
21. Golyamata Vapa	point 1	Internal	3.1	443	5677	2758
22. Varadat	point 1	Internal	2.9	224	5999	1567
23. Brezhanka	point 1	Transitional	2.0	164	4027	1008
PIRIN MOUNTAIN
24. Abyss No. 9–11	point 1	Transitional	1.0	177	177	177
	point 2	Transitional	1.0	219	219	219
	point 3	Internal	1.0	312	312	312
WESTERN RHODOPE MOUNTAIN
25. Yagodinska Cave	point 1	Transitional	6.1	560	1961	940
	point 2	Transitional	6.1	437	1853	827
26. Dyavolsko Garlo	point 1	Transitional	4.2	84	132	107
	point 2	Transitional	4.2	44	248	93
27. Ukhlovitsa	point 1	Transitional	10.3	102	1922	717
	point 2	Internal	10.6	162	2112	723
28. Goloboitsa	point 1	Internal	2.0	358	934	433
	point 2	Internal	2.0	334	702	469
29. Nadarska Cave	point 1	Transitional	2.1	66	595	324
30. Lepenitsa	point 1	Internal	8.1	365	2943	1139
	point 2	Internal	8.3	376	4280	1492
30. Lepenitsa	point 1	Internal	8.1	365	2943	1280
	point 2	Internal	8.3	376	4280	1369
31. Snezhanka Cave	point 1	Transitional	9.2	420	1484	752
	point 2	Internal	9.0	514	2456	1066
32. Novata/Starata Cave	point 1	Internal	4.4	331	1426	721
33. Yubileyna Cave	point 1	Internal	6.9	638	1813	1016
34. Gargina Dupka	point 1	Transitional	1.6	262	2772	1221
35. Ivanova Voda	point 1	Transitional	6.3	129	680	344
36. Topchika	point 1	Internal	5.0	577	2159	969
37. Chelevechnitsa	point 1	Internal	6.0	1045	4496	2506

. The average values vary widely. Values of over 2000 Bq m⁻³ were observed in the inner-cave climatic zones of the Venetsa Cave in the Moesian Platform, the Ledenika and Bacho Kiro Caves in the Stara Planina, the Labirinta and Golyamata Vapa Caves in Strandzha and the Chelevechnitsa Cave in the Western Rhodopes. Seasonal variation in the radon activity concentration was observed in all the monitored caves. As an example, nine of them are used, which are representative of all the geographical zones studied and have a longer monitoring period (Figure 5). The maximum is in summer–autumn (August–September/October), when values over 6000 Bq.m⁻³ were measured, e.g., in the Bacho Kiro (6136 Bq m⁻³), Venetsa (6887 Bq.m⁻³), Vodopada (7774 Bq.m⁻³), Labirinta (7937 Bq.m⁻³) Saeva Dupka (13,376 Bq.m⁻³) Caves. As expected, the lowest radon concentrations during this period were found in the high caves in Pirin and the Western Rhodopes. However, some of the low-lying caves with active year-round ventilation also fall into this group (e.g., Magura in the Moesian Platform).

The results of the integrated monitoring confirmed that the maximum radon activity concentrations show a seasonal delay of about 1 month. This is due to the ventilation regime in the cave system, which depends on its morphology and elevation. There is a continuous accumulation of cave gases before the initiation of the ventilation and, as a consequence, their peak concentrations coincide with the period of September–October. The typical examples shown in Figure 5 are the Saeva Dupka, Bisserna, Lepenitsa and Ukhlovitsa Caves. A similar seasonal delay is established by the authors in a previous study that was also conducted on the cave climatic elements [3,4,11,38]. The lowest values of the radon activity concentration were observed during the cold half of the year, especially in December–March (below 200–300 Bq.m⁻³). The large annual fluctuations in the radon activity concentration in the caves are confirmed by the difference between the measured highest and lowest values. In some of them, it exceeded 6000 Bq.m⁻³, such as in the Venetsa (6742 Bq.m⁻³), Labirinta (7477 Bq.m⁻³), Vodopada (7570 Bq.m⁻³) and Saeva Dupka (13,200 Bq.m⁻³) Caves. This is due to the leading role of cave ventilation in the annual radon regime, which will be analyzed in the next section.

The large variations in the radon activity concentration in the studied caves also depend on their geographical location, morphological specificity and water regime (see Section 2 and Table 1). For example, the caves in Strandzha, which are at an altitude below 300 m and are under the influence of a subtropical climate, have a high radon activity concentration (Table 2). The combination of high altitude and a submerged type of cave system with a continuous and active invasion of cold external air predetermines the lowest radon activity concentration values. Typical examples are the Abyss No. 9–11 Caves in Pirin (average concentration between 177 and 312 Bq.m⁻³) and the Ivanova Voda Cave in the Western Rhodopes, formed in a ponor (see [47]) of a karst uvala (average concentration 344 Bq.m⁻³). Spring caves have higher concentrations compared to descending cascade caves, e.g., Lepenitsa and Snezhanka in the Western Rhodopes (Figure 5C). The highest concentrations are observed in U-shaped caves, in which radon actively accumulates during the period without cave ventilation. Typical examples are the Saeva Dupka and Ledenika Caves (Figure 5B).

Another specific case caused by the impact of underground karst waters is the Goluboyitsa-1 Cave (Table 2). The accessible initial cave part is limited between two siphons, with the entrance opening at the beginning of summer when the level of the cave river decreases. The radon measurements carried out in the cave gallery between the two siphons found low values (between 433 and 574 Bq.m⁻³).

Of interest is the concentration of radon in the Bacho Kiro Cave, which is high year-round (average value between 1723 and 2472 Bq.m⁻³) in the internal-cave climatic zone and has very small variations in the intra-annual regime compared to the other studied caves (Figure 5B). In addition, despite the high concentration of radon, the concentration of the other cave gas measured—CO₂—is very low year-round (between 650 and 2300 ppm) (for comparison, see CO₂ data for Saeva Dupka, (see Section 4.2). This is due to the morphological features of the cave (Table 1) and the terrigenous rocks overlying the limestones [48], which isolate the cave system from the soil cover and hinder the flow of soil gas into the cave. Integrated monitoring in the Bacho Kiro Cave revealed the role of other sources of radon, which maintain its high concentration year-round: 1. Old alluvial sediments deposited in the cave with a thickness of up to 4 m, brought in by the Dryanovska River (the cave is of the sinkhole-spring type by genesis). These deposits are of marl-sand origin and are not carbonate. 2. Inflow from deep radon from the active fault zone in which the cave system is developed (see Section 4.6).

River sediments as an additional source of radon are also typical for the Yagodinska Cave, formed in an allogenic type of karst (according to L. Jakucs, [49]). It is also of a sinkhole-spring type, and the waters of the Buynovska River have brought thick deposits of granodiorites and biotite and amphibole–biotite granites into the cave galleries [50], which contain natural terrigenous radionuclides. Their influence has also been proven by parallel monitoring of the gamma background. In the Yagodinska Cave, it is approximately 2 times higher (0.13 μSv/h) than in other caves in Bulgaria, where it is lower than the natural gamma background in the nearby regions [11]. This additional source of radon in the Yagodinska Cave (river deposits) is probably also the reason for the high concentration of radon maintained in the tourist part of the cave, although it is strongly ventilated throughout the year due to the daily opening of the two artificial entrances (see Section 4.3). On the other hand, this ventilation regime is the reason for the high diurnal variations in the radon activity concentration. For comparison, in the Magura Cave, which has similar year-round ventilation through the tourist entrances, the average radon activity concentration does not exceed 640 Bq.m⁻³. The impacts of human activity on the radon concentration in the caves are analyzed in the next section.

The comparison between Table 1 and Table 2 reveals a number of other interesting relationships between the specifics of the cave systems and the radon concentration in them. High concentrations in some studied caves create health risks from radiation exposure to guides (see Section 4.7).

4.2. Role of Cave Ventilation in Radon Emissions

From previous studies of the authors, as well as other researchers (see Section 1), it has been established that the concentration of radon in cave systems depends primarily on the ventilation regime. The main generator determining the direction and intensity of ventilation is the difference in the density of air masses inside and outside the cave, which is caused by the difference in their temperatures. Additional factors are the morphology of the cave systems, the number and location of cave entrances, the location and scope of cave climatic zones, the geographical features of the region in which the cave system is developed and the features of atmospheric circulation and climate. The physical processes through which ventilation is carried out are convection and advection.

The studied cave systems belong to two main models of ventilation: dynamic and static. According to the morphology of the cave systems, some of them combine both ventilation models. The dynamic model includes abysses and abyssal caves, as well as multi-entrance caves. Ventilation patterns can also change as a result of human activity in caves: drilling new or expanding existing cave entrances; closing cave entrances with doors or walling them up, including when capturing karst springs; changes in cave morphology as a result of construction activities, etc. Typical examples are given by some of the investigated tourist caves: Dyavolsko Garlo Cave (entrance tunnel dug for tourists), Yagodinska Cave (entrance and exit tunnels have been dug), Snezhanka Cave (natural entrance closed and new one opened), Devetashka Cave (artificial partition walls erected and tunnels dug during the management of the cave as a fuel storage location), Magura Cave (changes in the morphology of the exit opening and an additional entrance dug for a wine cellar built in the cave) and Bisserna Cave (entrance tunnel dug in the upper dry gallery, artificial tunnel built under the bat colony and shaft opened to the lower water gallery). Changed natural ventilation is also found in Venetsa Cave, which was discovered in a quarry and subsequently developed as a tourist site. Another specific case is the Goluboyitsa-1 Cave in the Western Rhodopes, in which natural ventilation is controlled by the level of karst groundwater—when it rises, it form siphons and closes the contact with the cave entrance.

The seasonal change in the direction and intensity of ventilation in cave systems is a leading factor determining the gas composition of the cave atmosphere throughout the year. Ventilation processes have two clearly defined periods: active (continuous ventilation during the cold half of the year from October–November to March–April) and static (no ventilation during the warm half of the year from June–July to September). This duration of the periods is averaged for the majority of the studied caves, which are located below 1100–1200 m above sea level. Of these, the caves in the hypsometric zone below 600 m have a longer static period by about 1 month without ventilation. During the cold half of the year, when the cave temperature is higher than outside, in the descending and abyssal cave systems, there is an active flow of cold external air, which pushes out the warmer and less dense cave air through the cave entrances and open rock cracks. During the warm half of the year, the ventilation processes in the cave systems are of low intensity, which causes the cold, denser and heavier cave air to remain in the caves and for cave gases to accumulate. However, in ascending cave systems, a “leakage” of cold cave air occurs through the cave entrances. In spring caves, this is also “assisted” by the movement of water in the cave river. Between these two ventilation periods, there are transitional stages of “switching on” and “switching off” of the natural “fan” when the outside air temperature varies around the temperature of the cave air. The transitional stage is longer in spring (until mid-June) and is much shorter in autumn (lasting about a month). During the transitional stages, active ventilation phases occur within 1–2 weeks (depending on atmospheric circulation). They are the cause of the short-term fluctuations in the radon activity concentration. Figure 6 presents the results from three of the studied caves, in which continuous instrumental monitoring is carried out. They illustrate the dependencies between the temperature regime of the external air and cave air and the radon activity concentration very well. So, the seasonal delay between the radon activity concentration and external temperature discussed in Section 4.1 can be calculated quantitatively as a difference between the mass centers of the corresponding time series, and this is 33 days for the Saeva Dupka Cave, 18 days for the Bisserna Cave (point 3) and 20 days for the Ukhlovitsa Cave.

The integrated monitoring conducted confirmed that ventilation in cave systems also determines the regime of carbon dioxide (CO₂) in the cave air. It is measured on expeditions with a portable hand-held carbon dioxide meter GM70 (VAISALA), and in the Saeva Dupka and Bisserna Caves, continuous measurements have been organized using GMP343 and GMP222 carbon dioxide probes from VAISALA [3]. The authors’ published studies have already provided specific results that show that the regime in the concentrations of the two cave gases is similar. Similar conclusions have been made by other researchers [16,20,21,51], etc. An example of this similarity is their long-term parallel monitoring in the Saeva Dupka Cave (Figure 7). In some years, more significant differences in the course of the peak values of the two gases are observed (e.g., 2018, 2023, 2024). This is due to the different factors that influence their genesis and migration in cave systems. Furthermore, Saeva Dupka is a tourist cave, and an additional source of CO₂ in it is from the breathing of visitors [3,11]. They are most numerous during the active tourist season (May–September, with over 30,000 visitors), which coincides with the period without ventilation in the cave.

The extent to which the patterns and duration of ventilation and the values of the radon activity concentration depend on the morphological features and location of the respective cave systems can be assessed by comparing Table 1 and Table 2. The intensity of ventilation in cave systems and, accordingly, the variations in radon concentration are also determined by the cave climate. Depending on the size and spatial development of the studied cave systems, cave climatic zones with varying degrees of impact of external climatic factors are distinguished in them. Three zones predominate: entrance (dynamic—with the most active fluctuations in the values of cave climatic indicators and in the composition of the air due to active ventilation), transitional and internal (static—with a relatively constant cave climate and the weakest manifestations of ventilation). This directly affects the radon concentrations and can be traced in Figure 8. It reflects the course of radon values in the transitional and internal cave climatic zones, respectively, for the monitoring period in six cave systems that are representative of the different regions. As expected, the largest difference in the radon activity concentration in the two zones is during the static period without ventilation—between 1 and 3 kBq.m⁻³, and in the Saeva Dupka Cave, up to 13 kBq.m⁻³ (2020). In the period with active ventilation, this difference is already minimal.

4.3. Diurnal Variations in Radon Activity Concentration

Diurnal variations were tracked through instrumental monitoring (see Section 3.2). This covers different seasons and periods with different ventilation regimes in the caves. Here, we present results from five caves that are representative of the main morphographical regions in Bulgaria.

In the Saeva Dupka Cave, the diurnal fluctuations in the radon activity concentration during the cold period with active ventilation are from 50 to 150 Bq.m⁻³, and in the summer, when there is no ventilation of the cave air, the values are between 400 and 500 Bq.m⁻³. The diurnal fluctuations in the temperature of the cave air throughout the year are of the order of 0.1 °C, which was also found in the other studied caves. In the autumn, when very active pulses of cave ventilation occur, sharp decreases in the radon concentration are observed in Saeva Dupka within a few days. Here are a few examples: from 4300 to 1200 Bq.m⁻³ (16–19 October 2023); from 3000 to 250 Bq.m⁻³ (9–14 November 2023); from 5500 to 2200 Bq.m⁻³ (30 September–2 October 2024) and from 3500 to 210 Bq.m⁻³ (17–21 October 2024). This trend of sharp autumn decreases has already been established and has also been established for the concentration of CO₂ [3].

In the Bisserna Cave, the daily fluctuations in the radon activity concentration are from 150 to 600 Bq.m⁻³. In the transitional cave climatic zone of the upper dry floor of the cave, sharp increases in the radon activity concentration have also been established (up to 5000 Bq.m⁻³, see also Figure 8). They are typical of the cold period and are caused by impulses of cave ventilation and “suction” of radon from the lower wet level, in which the radon concentration is high year-round.

The Bacho Kiro Cave is distinguished by higher values of daily fluctuations in the radon activity concentration (between 1200 and 1800 Bq.m⁻³), which occur year-round in the inner-cave climatic zone (the long tourist route). Sharp increases in the radon activity concentration (up to 3600 Bq.m⁻³) were also detected, followed by a decrease (up to 2200 Bq.m⁻³) within 1 to 2–3 days. These coincide with manifestations of seismic activity, which is discussed in Section 4.6. This confirms the specificity of the radon activity and regime of concentration in this cave (see also Section 4.1).

In the Ukhlovitsa Cave, which is of the same morphological type as Saeva Dupka, the diurnal fluctuations in the radon activity concentration vary from 80 to 400 Bq.m⁻³. In this cave, sharp autumn decreases in the radon activity concentration also occur due to ventilation pulses, for example, from 1800 to 400 Bq.m⁻³ (7–8 October 2024).

The other Rhodope cave—Yagodinska—is a typical example of a cave withvery active diurnal fluctuations in the radon concentration, which occur year-round (Figure 9). In winter, they vary around 250 Bq.m⁻³, but during periods of heat waves in the external atmosphere, they reach up to 1250 Bq.m⁻³. In summer, the fluctuations are between 500 and 1000 Bq.m⁻³. The reason for this is the active year-round ventilation in the cave, which is due to the two artificial tourist entrances that have been dug and which are used year-round [11,14]. A similar situation of anthropogenically induced diurnal fluctuations in the concentration of cave gases has been established through monitoring in the Magura and Dyavolsko Garlo Caves, as well as in the initial part (transitional cave climatic zone) of the Bisserna Cave (Figure 5).

These results challenge the published claims [52] that “there is no difference in diurnal radon concentrations in investigated caves” (six Bulgarian show caves, including Saeva Dupka, Bacho Kiro and Ukhlovitsa). This conclusion is based on only one-time (2–3 days long) “continuous diurnal radon measurements in caves” conducted in connection with determining effective doses for cave guides (see Section 4.7).

4.4. Relationships Between Radon in Soil and in Cave Systems

Soil is an important source of radon in cave systems, especially in those developed in the epikarst. A typical example is the Saeva Dupka Cave, which is why it was chosen to clarify the relationships between radon concentrations in the soil above the cave and in the soil gas. For this purpose, parallel monitoring with TED was carried out in the Concert Hall of the cave (point 4, Table 2) and in the soil above it at three points (see Section 3.1). The thickness of the rock layer above the cave hall is 20–25 m. The soil is Rendzic Leptosol, LPk, with a thickness of 0.2–0.5 m. It is overgrown with woody and scrub vegetation, with a predominance of Carpineta orientalis and Carpinus orientalis.

The results of the parallel measurements of the radon activity concentration and cave air and external air temperatures (Figure 10) show a clear correlation between the radon concentration in the Saeva Dupka Cave and the external temperature, which determines the ventilation regime of the cave (see also Figure 8C).

The main source of radon in the Saeva Dupka Cave is soil gas. It is noteworthy that the concentration of radon activity in the soil has a more complex and even reverse trend with respect to that in the cave, with a number of “jumps”, and it is especially active (with sharp increases) during the cold half of the year. This indicates a predominant influence of other factors (precipitation, temperature) or their combination (humidity, frost), which determine the permeability of the soil. This depends both on the annual distribution and intensity of precipitation, on the structure and composition of the soil–vegetation cover and on the temperature regime. During the cold half of the year, the duration and thickness of the snow cover is also an important factor, which stops radon exhalation into the external atmosphere. In periods of drought, typical of the warm half of the year, this exhalation is especially active and leads to a decrease in radon concentrations in the soil.

Soil is also a factor in increased radon emanation from the rock massif. The increase in soil temperature and humidity activates the productivity of CO₂ and the acidification of soil solutions. This intensifies corrosion processes, especially in the soil–rock contact zone, and it expands the system of rock cracks. As a result, the area for emanation increases, and radon exhalation is activated in the cave system.

4.5. Global Changes and Radon Emissions in Cave Systems

In the program of the scientific project ProKARSTerra-GlobalChange (see Acknowledgements), models of the impacts of global climate change were developed in seven of the model karst geosystems of the ELK, in which most of the caves with radon monitoring are developed: three geosystems in northern Bulgaria in the climatic zone with a transitional subtropical–temperate climate and four in the Western Rhodopes with a mountainous subtropical climate [37,53]. This study covers the period after 1979 and uses mainly statistical methods and cluster analysis. The European Center for Medium-Range Weather Forecasts (ECMWF), ERA5-Land reanalysis [54], was used as the main source of data for the climate elements in the studied areas. The resolution of this data is 0.1 × 0.1° (9 × 9 km), and it allows for the climatic characterization of relatively small areas, such as the model karst geosystems [3]. Based on a comparison with data from meteorological stations of the Bulgarian National Institute of Meteorology and Hydrology in the studied areas, some corrections were introduced, especially for the average annual precipitation.

Based on the trends in the average annual values for the studied period, the average annual air temperatures have increased everywhere, and this increase is statistically significant. The specific values range from 0.3 to 0.5 °C/decade [53]. Regarding precipitation, no statistically significant trends in the average annual values have been identified over the past four decades, but changes in the sub-annual course have been observed. These trends have a specific effect on karst processes and on the cave environment and cave gases.

It was clarified in Section 4.2 that the main factor affecting the regime in the concentration of radon is cave ventilation. Therefore, global climate change will have an impact on the activity of radon in caves through its effects on the ventilation regime. The increase in atmospheric air temperatures implies a prolongation of the period without cave ventilation, as a result of which an increase in the concentration of radon in caves is expected. However, this effect is not clearly manifested in the studied caves because the increase in temperatures is primarily during the summer period, as well as through “hot waves” during the cold period. However, these waves have a short-term impact on the concentration of radon because cave ventilation continues to be active. Another effect is an increase in the variations in the values of the radon activity concentrations during the transitional seasons, when the alternation between “hot” and “cold waves” causes active ventilation phases of “on” and “off” of the cave “fan”. This is due to abrupt climate changes, known as temperature flips, which, according to the latest global studies, will become more frequent and more extreme in the future.

The periods of radon monitoring in the studied caves are still too short for more in-depth analyses and forecasts of the impacts of global changes. However, the statistical models compiled for the Saeva Dupka Cave, which has the longest monitoring history, and the forecasts made on their basis show that by 2070 the radon concentration in the cave will increase, primarily during the warm half of the year [38]. Against the background of this trend, both periods of active growth and periods of decline will occur (for example, see Figure 6B). However, during the cold half of the year, low values of the radon concentration in the cave system will remain, because active intrusion of external air will continue due to the large difference that will remain between the temperatures of the cave and external atmospheric air. An additional argument is the fact that the rock massif in which the cave system is developed reacts much more slowly to external temperature warming changes. This has been confirmed by the long-term cave climatic monitoring [3].

Global changes in the precipitation regime also have an effect on radon activity. Against the background of the statistically established weak change in the average annual precipitation in the studied karst regions, periods of prolonged drought (mainly summer–autumn) and short-term intense precipitation have become more frequent in recent years. In addition, the duration of snow cover is rapidly decreasing, especially in the low hypsometric zones, as well as the freezing of the surface soil layer. As a result, the period of active exchange (diffusion) between atmospheric and soil gas is extended, as well as the exhalation of radon from the soil in karst cavities and caves.

4.6. Radon Emissions in Caves as Precursors of Seismotectonic Activity

Caves develop along systems of rock cracks, which in active tectonic zones are of great depth and can be conductors of radon of an endogenous origin [55,56]. Radon emissions as precursors of seismic activity are discussed in [57,58]. They discuss and cite a number of publications that examine pre-seismic radon anomalies, and attempts are made to link them to geological and geophysical processes in fault tectonic zones.

During the integrated monitoring in the Bulgarian karst caves, we identified two seismotectonic periods: an active one with a very clear pulse of tectonic pressure (2012–2016), followed by a passive period of relative tectonic quiet, which continues into 2025 [3,59]. During the first period, there was a direct relationship between seismic activity, anomalies in radon-222 concentrations and fault displacements recorded by TM71 dilatometers. This relationship was particularly clear in the Bacho Kiro Cave [59]. Anomalies in the radon concentration were recorded at both monitoring points (1 and 2, Table 2) along the “long route”, where cave ventilation is weakly manifested. The manifestations of active seismotectonics required us to shorten the period of TED exposure (from 1 week to 20 days). This made it possible to register seven cases of a sharp pre-seismic increase in the radon activity concentration (by 1.8 to 3.6 kBq.m⁻³), followed by a decrease (by up to 2.2 kBq.m⁻³). At the same time, the TM71 dilatometer in the Pop Hariton Hall also registered fault displacements along monitored fault (Figure 11). The reason was a series of earthquakes with a magnitude ranging from 4.1 to 6.9 in a vast region of the eastern part of the Balkan Peninsula. The results of the parallel climate monitoring conducted in the cave exclude the possibility that the detected anomalies in the radon concentration were related to ventilation effects from external meteorological influences. Given the location of the Bacho Kiro Cave in the Balkan Mountains, the pulses in the radon concentration were due to both the activity of the Vrancea seismic zone in Romania (from the north) and the seismic zones in southern Bulgaria and in Turkey and Greece (from the south). The reason is the structure of the tectonic zone in which the Bacho Kiro Cave is located. It is under the influence of the collision between the main tectonic plates and microplates in the region [59]. The main fault structures along which the cave system is developed are oriented northeast–southwest. A TM71 dilatometer was installed in a cave on such a fault (310/70). The active endogenous influences to which the cave system is exposed are also the reason for the specific regime and the year-round high concentrations of radon in it (see also Section 4.1).

The period after 2017 is one of a relative tectonic lull, and the observed anomalies in the radon concentrations and the fault displacements in the Bulgarian caves, the objects being monitored, are less pronounced. This is also a consequence of the continuing relatively low seismic activity in the territory of Bulgaria. This was also confirmed by the measurements of the TM71 dilatometers both in the Bacho Kiro Cave (Figure 12) and in the four other Bulgarian caves [3]. During this period, more significant fault movements in the Bacho Kiro Cave occurred in the interval from 29 May 2019 to 20 March 2021. They recorded an opening of the fault (x) and a relative subsidence of the NW tectonic block, in which part of the cave system is developed. Anomalies in the radon activity concentration (from 500 to 1500 Bq.m⁻³) were also detected, which correlate with regional earthquakes with a magnitude of 2.5–3.5. In 2022, on 1 May, an earthquake with a magnitude of 4.6 was registered in the region of the city of Plovdiv, which coincided with an increase in the radon activity concentration in the cave by 1500 Bq.m⁻³.

The combined monitoring of fault displacements and radon activity in caves developed in fault zones can be a very valuable precursor of seismic activity [58]. For example, the Bacho Kiro Cave has already shown potential as an underground station for studying seismic hazard [3,58]. For this purpose, active devices for continuous recording of radon activity concentrations are more effective. They provide identifiable time indicators comparable to recorded earthquakes. Our experience (see Section 3) has shown that the Alpha series instruments are suitable for continuous instrumental monitoring of radon in specific cave conditions. Another advantage is their ability to monitor the recordings online. In attempts to correlate anomalies in radon activity concentrations and seismic activity, another very important and often misleading factor must be considered—cave ventilation (see Section 4.2). This proves the need to conduct integrated monitoring in caves [3].

4.7. Radiation Risk in Show Caves and Individual Effective Doses

Show caves are specific workplaces for cave guides. The Bulgarian Radiation Protection Regulation [13] introduced a reference level of 300 Bq.m⁻³ for the average annual volume activity of radon in the air at designated workplaces in closed rooms (Article 94). This necessitates the application of article 95, paragraph 1 of the regulation to “assess the individual effective dose of the workers in these workplaces”. The results of the long-term monitoring of radon in show caves are indicative that in eight of them, the radon activity concentration is above the reference level (Table 2).

This is an reason for these caves, based on our research, to calculate the individual effective doses for the tour guides working in them [11,14]. For this purpose, the specificity of the caves as workplaces and the actual staying of the tour guides in them must be considered. Tourist tours in the caves are on routes of different lengths and with different durations of stay for talks at certain points of it. They have different radon concentrations, which also vary throughout the year in connection with the cave ventilation regime. This requires that when determining the effective dose, the relevant specific and objective information regarding these radiation exposure factors must be used.

Due to the lack of developed methodologies for radiation control in Bulgarian caves, we apply experience from Czech show caves. According to the recommendation of the Czech State Office for Nuclear Safety and ICRP Publication 115 [60], the personal effective dose limit is 6 mSv over 1 year (2000 working hours). The real annual effective dose E_a [mSv] from Rn exposure to professional staff (cave guides) can be calculated using the following formula:

$${\mathrm{E}}_{\mathrm{a}}=\mathrm{j}.{\displaystyle \frac{{\int }_{1\mathrm{J}\mathrm{a}\mathrm{n}}^{31\mathrm{D}\mathrm{e}\mathrm{c}}{\mathrm{A}}_{\mathrm{V}}\left(\mathrm{t}\right)\mathrm{d}\mathrm{t}}{2.{10}^6\mathrm{B}\mathrm{q}.{\mathrm{m}}^{-3}}}.6\mathrm{m}\mathrm{S}\mathrm{v}$$

(3)

where A_V(t) [Bq.m⁻³] is the variable for the Rn concentration throughout the year or season, and j presents the individual “cave factor”. It includes the share of the free fraction of short-lived daughter products of radon decay fixed on aerosol particles (<5 nm). For the accurate determination of j, specialized measurements are required in each cave (including the unbound fraction f_p, [8]). Since such measurements are still pending in Bulgaria, the Czech example is used, according to which the usual value of j is ~1.5, and, in the conservative approach, j = 2 [33]. When calculating E_a [mSv], the real duration of guides staying in the cave throughout the year must be considered. For the purposes of the numerical solution of the time integral of the Rn concentration, the numerator in Equation (3) must be expressed as:

$${\int }_{1\mathrm{J}\mathrm{a}\mathrm{n}}^{31\mathrm{D}\mathrm{e}\mathrm{c}}{\mathrm{A}}_{\mathrm{V}}\left(\mathrm{t}\right)\mathrm{d}\mathrm{t}={\sum }_{\mathrm{k}=1}^{365}{\mathrm{A}}_{\mathrm{k}}{\mathrm{t}}_{\mathrm{k}}={\sum }_{\mathrm{k}=1}^{365}{\mathrm{A}}_{\mathrm{k}}{{\mathrm{n}}_{\mathrm{k}}\mathrm{t}}_0$$

(4)

where k is a day in the year, A_k is the corresponding daily Rn concentration and t_k is the daily time spent in the cave (n_k is the number of entries in day k, t₀ is the duration of a single visit).

Equation (2) is valid in the case of a stable spatial Rn concentration along the entire tourist route in the cave. However, in most of the studied caves, the radon activity concentrations show significant variations. For example, in the Saeva Dupka Cave, a typical tour with tourists usually consists of three stops with guided talks (e.g.) in areas L1, t₀(L1), L2, t₀(L2) and L3, t₀(L3), in which there are different radon activity concentrations of A_k(L1), A_k(L2) and A_k(L3), respectively. Therefore, Equation (2) must be reformulated as

$${\int }_{1\mathrm{J}\mathrm{a}\mathrm{n}}^{31\mathrm{D}\mathrm{e}\mathrm{c}}{\mathrm{A}}_{\mathrm{V}}\left(\mathrm{t}\right)\mathrm{d}\mathrm{t}={\sum }_{\mathrm{k}=1}^{365}\left[{\mathrm{A}}_{\mathrm{k}}\right(L1\left){\mathrm{t}}_{\mathrm{k}}\right(L1)+{\mathrm{A}}_{\mathrm{k}}\left(L2\right){\mathrm{t}}_{\mathrm{k}}\left(L2\right)+{\mathrm{A}}_{\mathrm{k}}(L3){\mathrm{t}}_{\mathrm{k}}(L3)]$$

(5)

resp.

$${\int }_{1\mathrm{J}\mathrm{a}\mathrm{n}}^{31\mathrm{D}\mathrm{e}\mathrm{c}}{\mathrm{A}}_{\mathrm{V}}\left(\mathrm{t}\right)\mathrm{d}\mathrm{t}={\sum }_{\mathrm{k}=1}^{365}{\mathrm{n}}_{\mathrm{k}}\left[{\mathrm{A}}_{\mathrm{k}}\right(L1\left){\mathrm{t}}_0\right(L1)+{\mathrm{A}}_{\mathrm{k}}\left(L2\right){\mathrm{t}}_0\left(L2\right)+{\mathrm{A}}_{\mathrm{k}}(L3){\mathrm{t}}_0(L3)]$$

(6)

The actual annual effective dose, Ea [mSv], from exposure to Rn for professional personnel (cave guides) is obtained by substituting Equation (5) into Equation (3). In Saeva Dupka Cave, a guide spends ~350 h per year underground with ~500 entries. This gives an effective annual dose of Ea = 6.5 mSv using A_k,max, according to the conservative approach, and resp. 3.0 mSv using a more realistic value of A_k,ave [11].

With respect to the radiation safety of guides, this confirms the importance of continuous online monitoring of radon activity. At the same time, the results obtained prove that the exposure of ordinary visitors to radiation can be considered negligible [14].

Table 3. Estimated real annual effective dose of the cave guides in 5 show caves in Bulgaria (by [11,14]).

Tourist Cave (Number as per Figure 1)	Monitoring Point 1	Personal Effective Dose (mSv) 2
		Conservative Approach: Max Av (Cave Factor j = 2)	Optimal Approach: Average Av (j = 1.5)
2. Venetsa Cave	Third Hall (2)	11.6	4.8
	Fifth Hall (3)	6	2.8
	Total in the cave	17.6	7.6
13. Saeva Dupka Cave	Stack Hall (2)	0.7	0.4
	Concert Hall (4)	4.7	2.0
	Cosmos Hall (5)	1.1	0.6
	Total in the cave	6.5	3.0
16. Bacho Kiro Cave	Ritual Hall (1)	0.3	0.2
	Concert Hall (2)	0.2	0.1
	Total for short route	0.5	0.3
	Pop Hariton Hall (3)	0.7	0.3
	Reception Hall (4)	2.1	1.2
	Total for long route	2.8	1.5
	Total in the cave	3.3	1.8
27. Ukhlovitsa Cave	Lower Floor (1)	1.1	0.6
	Sinter Waterfall (2)	2.5	1.3
	Total in the cave	3.6	1.9
31. Snezhanka Cave	The Great Hall (1)	0.84	0.42
	The Magic Hall (2)	0.86	0.48
	Total in the cave	1.7	0.9

¹ In brackets is the number of the point according to Table 2. ² Effective doses exceeding the 6 mSv reference value are marked in red.

shows the calculated real effective doses of tour guides in five of the show caves for which we were provided with the necessary information in terms of the duration and annual distribution of the tour guides’ staying in the caves. Personal effective doses were calculated using either max values, with A_V, j = 2 (conservative approach), or average values, with A_V, j = 1.5, (optimal approach).

The highest effective doses were calculated for the Venetsa Cave. In this cave, we found that tour guides are also exposed to radon when they are in the office—there, they are exposed to an additional dose of 17.6 mSv (at j = 2) or 7.6 (at j = 1.5). The office is built at the very entrance to the cave, and the cave air also circulates in it. The situation is even more worrying given the limited number of tour guides (two to three) and the great tourist interest in the cave. This required us to send a statement with specific recommendations to the Dimovo Municipality, which manages the cave, immediately after our calculations [11]. As a result, a new office was provided for the tour guides without any connection to the cave entrance. There is also a potential risk of radiation exposure to tour guides in one of the most visited show caves, Saeva Dupka (avg. 50,000 tourists per year), especially if the number of tour guides serving it is reduced. In the Bacho Kiro Cave, despite the year-round high concentration of radon, no health risk for tour guides was identified. The reason is due to the applied work schedule and the mode of visits (a walk without a tour guide on the “short route” and an hourly visit on the “long route”, with groups of only over 15 visitors).

According to the radon monitoring results, our observations during periodic expedition surveys and the number and work schedule of the tour guides, effective doses above the reference value are also expected for the show caves Orlova Chuka, Bisserna and Ledenika. The fact that the radon concentration in most of the show caves has a clearly pronounced summer–autumn maximum, which coincides with the maximum number of tourist visits and with the period of the longest duration of tour guides staying in the caves, should also be considered [11,14]. This requires a well-thought-out work schedule and, if necessary, the appointment of additional tour guides during this season. For this purpose, the institutions that manage and operate show caves in Bulgaria are periodically notified with official letters and expert opinions about the radon monitoring results and the calculated effective doses for cave guides.

In 2023, effective doses for six Bulgarian tourist caves were also published, calculated by the National Center for Radiobiology and Radiation Protection, Sofia [52]. For five of these caves, we made calculations based on our long-term research (Table 3 in this work). The comparison of the results reveals serious discrepancies. For example, the doses for the Venetsa and Saeva Dupka Caves in this publication are significantly lower, and in the Bacho Kiro Cave, the dose is much higher (the specific work schedule of the tour guides is not considered in [52], see above in the text). The cited team of authors used their own data from one-time measurements with passive detectors with an excessively long exposure period (1 year), which is why the effective doses calculated by them are not considered objective. This is also confirmed by inaccurate information in the questionnaires that they used (Table 3 in [52]). The true duration of the tourist tours and the stay (time distribution) of the tour guides in the caves are as important as the measured Rn concentrations for the correctness of the effective dose calculation.

This example confirms the need to organize and maintain professionally proven continuous monitoring of radon in Bulgarian show caves, which would be the basis for the objective calculation of effective doses of cave guides. According to paragraph 2 of article 95 of the Radiation Protection Regulation [13], “when the individual effective dose of workers due to radon exposure exceeds 6 mSv for a period of one year, it is treated as in a situation of planned exposure and employers take appropriate radiation protection measures applicable to occupationally exposed persons”. This is within the competence of the relevant institutions responsible for workplace safety. In this aspect, there are other serious problems regarding the management of tourist caves in Bulgaria, which are due to gaps in Bulgarian legislation: 1. the Bulgarian legislation has not yet introduced a workplace called a “Show Cave” with all its ensuing consequences; and 2. the profession “Cave guide in a show cave” is absent from the register of professions in Bulgaria. We have already outlined these problems and possible measures to solve them in our publications [3,11,14], but they have not yet been discussed at the necessary level.

5. Conclusions

The long-term monitoring conducted in this study in cave systems representative of the different types of karst in Bulgaria revealed that the activity of radon in them depends on a combination of various factors. They regulate both the levels of the radon concentration and its annual regime. It has a well-pronounced seasonality, which is due to the natural ventilation processes in the cave systems due to the difference in the temperatures of the outside air and the cave air. The most radon accumulates during the warm (summer) season in U-shaped caves (up to 13 kBq.m⁻³). The lowest concentrations are found in sinking caves in mountainous areas, as well as in ventilated multi-entrance caves, including tourist ones with additionally dug entrances. The monitoring confirmed that the main sources of radon are the rocks in which the cave systems are formed, the soil above them, cave sediments and speleothems. In some of the caves, an additional source is from the flow of geogas from the earth’s crust through cracked fault zones. Degassing of dripping cave water can also contribute to the increase in the concentration of ²²²Rn [51]. Radon emissions are also possible from cave lakes and flowing cave rivers and streams.

The parallel measurements of the radon and CO₂ concentrations have established that both gases in the cave air react in the same way to the ventilation processes in the cave systems and have a similar regime in terms of seasonal fluctuations [3,38]. Such a regime has been established in other caves around the world (e.g., [15,51]). Therefore, in many studies conducted, radon is used as an indicator for monitoring ventilation processes (e.g., [17]). Our experience has proven that for this purpose, measurements of the CO₂ concentration are more effective. They are much easier to perform (expeditionary and stationary) both at selected points and along the entire profile of the caves, and the equipment used is more resistant to cave conditions, especially the high humidity of the cave air. In addition, CO₂ monitoring provides additional valuable information on the distribution, regime and dynamics of the radon concentration in cave systems.

Active modern global climate change, which also affects karst areas in Bulgaria, is expected to affect radon activity. The most obvious trend in the main climate indicators is the sustained increase in air temperature, especially during the warm half of the year. As a result, a summer–autumn increase in radon concentrations is expected [38,53]. However, the effect of global changes on the various factors of radon activity has different trends, which are often opposite. This, as well as the relatively short period of the MIKS, makes it difficult to develop effective models of expected changes in radon activity and concentrations in cave systems in Bulgaria at this stage.

The high concentrations of radon found in the most visited show caves confirm the need to maintain continuous monitoring in them and to monitor the real effective doses for the staff by cave guides. It is recommended that each of them uses a portable active alpha dosimeter (e.g., AlphaE) during their stay in the cave. Daily (weekly, monthly) dose readings should be recorded in each guide’s personal card. After completing their work shift in the cave, he/she should pass the dosimeter to the next guide. This would limit the number of dosimeters needed for each tourist cave to two or three. The calculation of the effective dose (Section 4.6) is applicable if the guide has correctly kept a work diary with time data for his stay in the cave. However, this method is less accurate. Due to the short stay of tourists in show caves, there is no health risk from radiation exposure for them.

The MIKS developed and applied in karst geosystems in Bulgaria [3] has also proven its effectiveness for radon activity, because it depends on many factors that should also be monitored. In addition, MIKS is based on the ProKARSTerra methodological platform, which is the basis for integration between scientific research, management, business and education. Additionally, integration in all its aspects is extremely important and requires serious resources, incl. in organizing and maintaining networks for integrated monitoring in model cave systems. Given the important scientific and applied significance of the monitoring results, such as radon activity, interest and support from the relevant state institutions is also necessary.