Introduction

Naturally occurring radioactive materials (NORMs) existing in sufficient quantity with their decay products are significant sources of radiation exposure for the human population. These NORMs can be found across the earth’s crust causing terrestrial environmental radioactivity. The external exposure to humans is mainly due to the primordial radionuclides of the 238U and 232Th decay series as well as 40K (UNSCEAR (United Nations Scientific Committee on the Effects of Atomic Radiation), 2000). A mitochondrial DNA mutation associated with the exposure to natural radioactivity has been reported for the human inhabitants along the Chavara-Neendakara coast of Kerala, India (Forster et al., 2002). Such type of data escalates the importance to understand the distribution of NORMs in the environment in association with radiological hazards and human health risks.

Regions with enormous amounts of the primordial radionuclides associated with rocks, soil, and sand are identified as high background natural radiation areas (HBNRAs). There are a few well-known HBNRAs; however, the sources of natural radionuclides vary according to the local geology and geological processes. For example, in Ramsar, Iran, and Niška Banja, Serbia, these radionuclides are associated with travertine formations; in the Abu Rusheid area of Egypt, they are linked to mylonitic rocks with U mineralization; and in the Dornogobi Province of southeastern Mongolia, the presence of U deposits close to the ground surface has resulted in high dose rates (Ghiassi-nejad et al., 2002; Omori et al., 2019; Sahoo et al., 2023; Sakr et al., 2023). In coastal regions of India, high background natural radiation is associated with beach placer deposits of heavy minerals (density (ρ) > 2.9 g cm−3) such as ilmenite, monazite, rutile, and zircon (Mohanty et al., 2003; Veerasamy et al., 2020). Especially the thorium-bearing mineral monazite is considered to be the highest contributor to the background radiation in the beach sands of India (Mohanty et al., 2003; Singh et al., 2007; Veerasamy et al., 2020).

The streams and rivers of these coastal areas play a major role in the formation of beach placer deposits. The hinterland high-grade metamorphic and igneous rocks are weathered and eroded by the streams and rivers, and the detritus composed of sand, silt, clay, and heavy minerals is carried and deposited in a variety of coastal environments including deltas, barrier islands, lagoons, foreshore, and backshore. The sediments are then reworked by waves, tides, longshore currents, and winds leading to an effective sorting of the mineral grains based on their size and density (Van Gosen et al., 2014). For instance, the Rushikulya River in Odisha erodes the Eastern Ghat Mobile Belt (EGMB) composed of rock types such as khondalite, charnockite, granite gneiss, and pegmatites and the river is believed to be the source of heavy minerals and high background natural radioactivity along Chhatrapur beach (Sulekha Rao & Misra, 2009; Veerasamy et al., 2020). Similarly, sediments of the Bentota River in Sri Lanka have been found to be rich in monazite and that has resulted in forming seasonal monazite-rich beach sand deposits along the Kaikawala and Beruwala coastal regions (Rupasinghe et al., 1983).

River sediments (mainly as sand) also have an essential role in construction projects as an important mixture component for building materials in Tamil Nadu, India (Ramasamy et al., 2011). The activity concentration of radionuclides and radiological hazard parameters have been estimated and found to be equal to or less than the world average values in some sediments of the major rivers of Tamil Nadu (Narayana et al., 2016; Ramasamy et al., 2011; Thangam et al., 2020). Natural radionuclide distribution in the coastal sands and soils of the Kanyakumari HBNRA region has been extensively studied (Malathi et al., 2005; Natarajan et al., 2023a; Punniyakotti & Ponnusamy, 2018). However, there is little information about natural radionuclide data in river sediments from this area.

Rivers are a possible source for the high radioactivity observed along the Kanyakumari coastal area. Therefore, in this study, the sediments from rivers and streams of Kanyakumari were analyzed for the activity concentration of natural radionuclides and the radiological hazard parameters were estimated based on data of the natural radionuclides. In addition to it, spatial distribution maps of the natural radionuclides and dose rate were generated to visualize the distribution pattern.

Materials and methods

Study area

Kanyakumari is located at the southernmost tip of India and has two perennial rivers, the Tamiraparani (TR) and the Pazhaiyar (PR). The TR originates as the Kothaiyar River on the Agastiar Hills of Western Ghats and flows on the western slopes for a short distance before taking a southwesterly direction. The Paraliyar River originating from the same hills flows in a southwesterly direction and unites with the Kothaiyar River near Moovattumugam to flow as the TR until it joins the Arabian Sea near Thengapattinam without forming any delta (Fig. 1a). The PR originating from the Mahendragiri Hills flows on the easterly slopes before taking a southeasterly direction for 30 km to enter the Arabian Sea near Manakudi by forming an estuary. The Kothaiyar, Paraliyar, and TR collectively flow for a length of 60 km from the foothills of the catchment area to reach the mouth (the place where the river enters the sea) and they are grouped as the TR in this study (Fig. 1b).

Fig. 1
figure 1

Kanyakumari district maps a showing rivers and streams and b showing sediment sampling locations (n = 44) (Generated using ArcGIS 10.6 software) (Numeric value in the river Sample ID are assigned from low to high for the catchment area to the mouth of the river) 

Full size image

Kanyakumari receives rainfall from both southwest and northeast monsoons which benefits the rivers; however, most of its precipitation is due to the cyclonic activities in the Bay of Bengal (CGWB, CGWB (Central Groundwater Board of India), 2008). The geology of Kanyakumari is comprised of Trivandrum Block (TB) and Nagercoil granulites (NG) of the Southern Granulite Terrain and the composition of these rocks is discussed in detail elsewhere (Rajesh et al., 2011; Santosh et al., 2003). The catchment area of the rivers includes Precambrian crystalline rocks of charnockites, khondalites, and migmatite gneisses. The rivers erode these rocks which are tonalite-granodiorite in composition with apatite, ilmenite, monazite, and zircon as major accessory minerals (Rajesh et al., 2011). Along with these two major rivers, there are a few small seasonal streams (SS) with a length not exceeding 20 km which form and enter the Arabian Sea. All these streams might once have been the channels of rivers like the TR and Paraliyar as there is evidence for river migration such as oxbow lake (Fig. 2). The width of these stream channels exceeds 100 m in some places and these channels are primarily used for agricultural activities since most of Kanyakumari is rough terrain, unsuited to farming (Fig. 3a, b). Figure 3c shows the cross section at a SS sampling location indicating the river depositional sequence.

Fig. 2
figure 2

Satellite image of an oxbow lake adjacent to the Tamiraparani River demonstrating migration of the river (Google Earth Version 9.191.0.0)

Full size image
Fig. 3
figure 3

Satellite images of a streams and b agricultural activities on the old river channels (Google Earth Version 9.191.0.0). c Cross section at a stream sampling location showing the river depositional sequence

Full size image

Sample preparation

Sediment samples were collected from the foothills of the catchment area to the mouth of the rivers at 18 locations for the TR and at 14 for the PR. Sediment samples were also collected at 12 locations for the SS among the agricultural lands (Table 1). All the sampling locations were within latitude N 8.114 to 8.446 and longitude E 77.161 to 77.488 and sampling was done during March 2022 (summer season). For the TR and PR, the interval between each sampling location along the main channel was approximately maintained at 2 km. Since both rivers have tributaries joining them, sediment samples from those tributaries were also collected, respectively (Fig. 1b). The SS sediment samples were collected from the branches before and after confluence (Fig. 1b). The samples from each location were collected at a depth of 0–5 cm using a Wildco® hand core sediment sampler and stored in a labelled polyethylene bag before being transported to the laboratory in Japan. The sediment samples were brought to Japan after acquiring proper permission from the Ministry of Agriculture, Forestry and Fisheries based on plant protection laws. In the laboratory, each sediment sample was first air-dried and then oven-dried at 105 °C for 24 h until complete removal of moisture had been realized. Plant root materials and rock fragments were removed by passing the dried samples through a 2-mm sieve. The samples were packed and sealed in U8 cylindrical containers (d = 48 mm, h = 58 mm) and left for 4 weeks to attain secular equilibrium among the U series radionuclides.

Table 1 Activity concentration (Bq kg−1) of 226Ra, 232Th, and 40K with absorbed dose rate (nGy h−1) for Kanyakumari River and stream sediment samples  
Full size table

Instrumentation and calibration

A coaxial P-type high-purity germanium (HPGe) detector (CANBERRA GX4018) with a range of 0–4000 keV was used for gamma spectrometry measurements. Energy and efficiency calibration of the HPGe detector was carried out using a multi-nuclide standard source supplied by the Japan Radioisotope Association (JRIA) with gamma energies ranging from 60 to 1333 keV and the overall uncertainty was found to be less than 10%. The precision of the method was checked using the reference material Irish Sea Sediment IAEA – 385. The sample counting time was pre-set at 80,000 s. 214Pb (351.99 keV) and 214Bi (609.31 keV) were considered for the estimation of 226Ra activity concentration. 208Tl (583.14 keV) and 228Ac (911.20 keV) peaks were considered for the calculation of 232Th, assuming that the daughter radionuclides are in radioactive equilibrium with parent radionuclides and 40K (1460.8 keV) was considered for the direct estimation of 40K activity concentration (Natarajan et al., 2023a). The minimum detection levels (MDL) for 214Pb and 214Bi were 2 ± 1 Bq kg−1 and 2 ± 1 Bq kg−1, 208Tl and 228Ac were 3 ± 2 Bq kg−1 and 3 ± 1 Bq kg−1, and the MDL for 40K was 11 ± 5 Bq kg−1, respectively. The specific activities (Bq kg−1) were calculated using Eq. (1).

$$A = {~}^{C}\!\left/ \!{~}_{pwt\varepsilon }\right.$$
(1)

Here, A is the specific activity, C is the net count above the background, p is the absolute emission probability, w is the dry weight of sample (kg), t is the measurement time (80,000 s), and ε is the absolute efficiency of the detector (Hassan et al., 2010).

Spatial analysis

The spatial distribution maps were prepared using ArcGIS software (v 10.6). The inverse distance weighted (IDW) method in the software Spatial Analyst Tool was used to determine the values for unsampled locations. The IDW method presumes that every measured point has an effect that decreases with distance and the nearest value is considered rather than the farthest away one. Although inverse distance weighting (IDW), ordinary kriging (OK), and ordinary co-kriging (OCK) methods are more commonly used, for this study, the IDW method was employed since it has been found to be the best conventional interpolation technique with some measure of certainty and predictive accuracy (Li & Heap, 2011). The boundary for the spatial distribution starts from the foothills, from where the first sampling location was found.

Radiological hazard parameters

The radiological hazard parameters such as absorbed dose rate (D), annual effective dose equivalent (AEDE), and radium equivalent (Raeq) were estimated based on the activity concentrations of 226Ra, 232Th, and 40K in order to assess the radiological risk of the river and stream sediments. The radiological effects are directly related to D (nGy h−1) and D was calculated using Eq. (2) suggested by the UNSCEAR (UNSCEAR (United Nations Scientific Committee on the Effects of Atomic Radiation), 2000).

$$D=0.462{C}_{{\mathrm{Ra}}}+0.604{C}_{{\mathrm{Th}}}+0.0417{C}_{{\mathrm{K}}}$$
(2)

Here, CRa, CTh, and CK are the activity concentrations of 226Ra, 232Th, and 40K, respectively.

The river sediments are mainly used as materials to be mixed with cement for building construction. As people spend more time indoors than outdoors, the annual dose to any individual can be assessed by calculating the annual effective dose equivalent (AEDE) outdoors and indoors by Eqs. (3) and (4) as given by the UNSCEAR (United Nations Scientific Committee on the Effects of Atomic Radiation) (2000).

$${AEDE}_{{\mathrm{out}}}=D \times 8760 \times 0.2 \times 0.7 \times {10}^{-6}$$
(3)
$${AEDE}_{{\mathrm{in}}}=D \times 8760 \times 0.8 \times 0.7 \times {10}^{-6}$$
(4)

Here, AEDEout and AEDEin (mSv y−1) are for the outdoor and indoor effective dose equivalents, respectively; D is the absorbed dose rate; 0.7 Sv Gy−1 is used to convert the absorbed dose rate (nGy h−1) to the annual effective dose equivalent. Finally, 0.2 and 0.8 are the occupancy factors of outdoors and indoors.

Radium equivalent activity (Raeq) is a widely used hazard index. This single index allows assessment of the exposure to radiation since the radionuclides 226Ra, 232Th, and 40K are not uniformly distributed in the river and stream sediments. Raeq can be calculated using Eq. (5) (UNSCEAR (United Nations Scientific Committee on the Effects of Atomic Radiation), 2000) and it is expressed in Bq kg−1.

$${Ra}_{{\mathrm{eq}}}= {C}_{{\mathrm{Ra}}}+1.43{C}_{{\mathrm{Th}}}+0.077{C}_{{\mathrm{K}}}$$
(5)

Here, CRa, CTh, and CK are the activity concentrations of 226Ra, 232Th, and 40K, respectively.

Results and discussion

Activity concentration

Activity concentrations of 226Ra, 232Th, and 40K (Bq kg−1) of the river and stream sediments are given in Table 1. The activity concentrations of 226Ra, 232Th, and 40K (Bq kg−1) for the TR sediment samples (n = 18) were in the range of 18 to 386 with a mean of 100 Bq kg−1, 39 to 2362 with a mean of 773 Bq kg−1, and 65 to 632 with a mean of 340 Bq kg−1, respectively. The activity concentrations of 226Ra, 232Th, and 40K (Bq kg−1) for the PR sediment samples (n = 14) ranged from 4 to 91 with a mean of 33 Bq kg−1, 12 to 1372 with a mean of 313 Bq kg−1, and 209 to 718 with a mean of 512 Bq kg−1. The activity concentrations of 226Ra, 232Th, and 40K (Bq kg−1) for the SS sediments (n = 12) ranged from 12 to 368 with a mean of 95 Bq kg−1, 69 to 2163 with a mean of 557 Bq kg−1, and 28 to 659 with a mean of 211 Bq kg−1. The mean activity concentrations of 232Th (Bq kg−1) of both rivers and stream sediments were found to be relatively higher than the world average value of 45 Bq kg−1 as well as the Indian average value of 68 Bq kg−1 (UNSCEAR, 2008; Punniyakotti & Ponnusamy, 2018). In case of 226Ra activity concentration, the TR and SS sediments were found to be higher than the world average value of 33 Bq kg−1 and Indian average value of 28 Bq kg−1, whereas the mean 226Ra activity concentration of PR sediments was normal. However, the mean 40K activity concentration was a little higher than the world average value of 412 Bq kg−1 only for the PR sediments.

The mean activity concentration of radionuclides followed the order of 232Th > 40K > 226Ra in both TR and SS sediments, whereas it was 40K > 232Th > 226Ra for the PR sediments. In the river and stream sediments, 226Ra activity concentration was lower than that of 232Th and 40 K; 226Ra is known to be preferentially incorporated into the aqueous phase by means of alpha recoil, and this could be a possible reason for the depletion of 226Ra relative to 232Th in the river and stream sediments (Powell et al., 2007). The soils of the Odisha HBNRA were suspected to be the cause of the increased U concentration in the groundwater (Veerasamy et al., 2023). The river water may also feed the groundwater by influent flows during high flow times. Water is an essential component for the survival of life and the people of Kanyakumari rely more on the groundwater for freshwater needs (Raja et al., 2021). However, the average uranium concentration in the groundwater of Kanyakumari was found to be 2 µg L−1 which was less than the permissible limit recommended by the WHO (30 µg L−1) and the Atomic Energy Regulatory Board (AERB) of India (60 µg L−1) (Raja et al., 2021).

The high activity concentration of 40K may be owing to the presence of light silicate and felsic groups of minerals as 40K mainly has a terrestrial origin (Natarajan et al., 2023a). The high concentrations of 40K and 232Th compared to 226Ra have also been observed in other river sediments of Tamil Nadu such as the Cauvery, Ponnaiyar, and Thamirabarani (Tirunelveli) Rivers (Narayana et al., 2016; Ramasamy et al., 2011; Thangam et al., 2020). A similar observation was reported in the Nile River (Egypt) sediments and it was stated that the clay content might be a contributing factor for the high activity concentrations of 40K, while conversely the lower concentrations were associated with sandy content (El-Gamal et al., 2007). The major rock types present in the catchment area of the TR and PR might also influence the activity concentrations of 226Ra, 232Th, and 40K since they are rich in heavy minerals like ilmenite, monazite, and zircon as the major accessory phase. Natural radiation levels as high as 45,000 nGy h−1 have been reported in hinterland Putteti syenite rock units and the western part of the Kanyakumari district also exhibited a prominent radioactivity zone (Singh et al., 2007).

The spatial distributions of the radionuclides 226Ra, 232Th, and 40K are shown in Fig. 4a, b, and c, respectively. A heterogenous distribution of radionuclides was seen. The background radiation levels in the Kanyakumari district were mainly associated with the 232Th because of the presence of thorium-bearing minerals. The spatial distribution map of Fig. 4b indicated that the very high concentrations of 232Th were limited to the western part of this district.

Fig. 4
figure 4

Spatial distribution maps of activity concentrations of a226Ra, b 232Th, and c.40K. d Absorbed dose rate distribution map (Generated using ArcGIS v 10.6)

Full size image

The world average ratios for 232Th/226Ra, 232Th/40K, and 226Ra/40K are 0.86, 0.08, and 0.09, respectively (UNSCEAR (United Nations Scientific Committee on the Effects of Atomic Radiation), 2000) . The mean 232Th/226Ra, 232Th/40K, and 226Ra/40K ratios of TR and PR sediments were 6.98, 5.99, and 0.68 and 8.10, 0.94, and 0.08, respectively, and for SS sediments, the ratios were 5.65, 6.20, and 1.10, respectively. 232Th/226Ra and 232Th/40K ratios were considerably higher than the world average values, and that showed the dominance of the Th-rich minerals in the sediments. The mean 226Ra/40K ratio of the PR sediments was slightly lower than the world average due to the predominance of light minerals over U-rich minerals and the elevated 232Th/226Ra of TR, PR, and SS sediments could indicate the leaching of 226Ra due to an oxidizing environment (Khan et al., 2022).

Radiological risk assessment

The absorbed dose rate (D) in air at 1 m above the ground was estimated using Eq. (2) and it ranged from 50 to 1518 nGy h−1 with a mean of 527 nGy h−1 for TR sediment samples, from 28 to 880 nGy h−1 with a mean of 226 nGy h−1 for PR sediment samples, and from 52 to 1478 nGy h−1 with a mean of 390 nGy h−1 for SS sediment samples. The mean absorbed dose rates of river and stream sediments were found to be relatively higher than the world average value of 58 nGy h−1 reported by the UNSCEAR (2008). The spatial distribution of the absorbed dose rate is shown in Fig. 4d. The comparison of the activity concentrations of natural radionuclides 226Ra, 232Th, and 40K (Bq kg−1) and absorbed dose rate (nGy h−1) for the river and stream sediment samples from Kanyakumari with some of the river sediments from other locations in India (Tamil Nadu, Kerala, and Odisha) as well as a few other countries like Bangladesh, China, Egypt, Nigeria, Slovak Republic, Thailand, and United States are given in Table 2.

Table 2 Comparison of activity concentrations of 226Ra, 232Th, and 40K and absorbed dose rate of the present study with sediments of other rivers of India and of other countries
Full size table

The radiological hazard parameters such as Raeq (Bq kg−1), AEDEout (mSv y−1), and AEDEin (mSv y−1) were also estimated along with the absorbed dose rate using Eqs. 3, 4 and 5. The Raeq for the TR sediments ranged from 110 to 3574 Bq kg−1 with a mean of 1230 Bq kg−1; for the PR sediments, it ranged from 57 to 2070 Bq kg−1 with a mean of 520 Bq kg−1; and for the SS sediment samples, it ranged from 120 to 3465 Bq kg−1 with a mean of 910 Bq kg−1. The mean Raeq of all the sediment samples was found to be 920 Bq kg−1, which was relatively higher than the UNSCEAR world average value of 370 Bq kg−1 (Fig. 5). The annual effective dose equivalent outdoor (AEDEout) and indoor (AEDEin) values for the river and stream sediment samples from Kanyakumari were in the range of 0.03 to 1.9 mSv y−1 with a mean of 0.5 mSv y−1 and 0.1 to 7.4 mSv y−1 with a mean of 1.9 mSv y−1, respectively. Both the AEDEout and AEDEin were higher than the world average values of 0.07 mSv y−1 and 0.45 mSv y−1 (UNSCEAR, 2008).

Fig. 5
figure 5

Radium equivalent activity for the sediment samples from Kanyakumari

Full size image

The absorbed dose rate as well as the activity concentration of radionuclides in the river and stream sediments of Kanyakumari were high relative to all the river sediments except for the 238U and 232Th activity concentrations of the Rushikulya and Mahanadi Rivers in Odisha (Table 2). The reported high concentrations of titanium (Ti) and cerium (Ce) in the Rushikulya and Mahanadi River sediments have been attributed to the presence of heavy minerals such as ilmenite (Ti dominant) and monazite (Ce dominant) (Mohanty et al., 2023). Our recent studies on the beach sands of Kanyakumari determined that the dose rate was mainly from 232Th and the high concentrations of Ti and Ce in beach sands propounding the natural radiation were mainly due to ilmenite and monazite (Natarajan et al., 2023a, 2023b).

The respective absorbed dose rates for 226Ra and 232Th had a strong positive correlation (R2 = 0.78 and 0.99) which may be due to the similar origins. The contribution of 232Th to the absorbed dose rate was 87%, which was similar to the contribution for the Kanyakumari beach sands (Natarajan et al., 2023a). The highest 232Th activity concentration was for sample TR-6 (Table 1), which was Paraliyar River bed sediment collected in close proximity to the Perunchanai Dam (Fig. 1b). The 226Ra and 232Th activity concentrations of samples TR-7 and TR-8 sharply decreased compared to TR-6 values whereas for TR-9 sample, the activity concentrations were somewhat lower compared to TR-6. There was no gradual increase or decrease in the activity concentration from the foothills of the catchment area reaching the mouth of the river rather; the 226Ra and 232Th activity concentrations fluctuated highly unlike the activity concentrations of 238U, 232Th, and 40K in the sediments of the Ponnaiyar River (Ramasamy et al., 2011). Flow velocity of a river has a direct relationship to the sediment carrying capacity; when the flow velocity decreases, the river loses its ability to carry the heavy minerals first and a normal graded bedding depositional sequence is followed (sand at the bottom and mud at the top) (Earle, 2019). Bedrock exposure in the river channel may cause sediments to deposit in pockets and river erosional features like potholes were witnessed in upstream areas of rivers flowing in Kanyakumari. These are some of the factors that could control the uneven distribution of radionuclides along the river channels.

On the other hand, almost all the SS sediment samples from the small streams flowing adjacent to the agricultural lands showed an elevated activity concentration of 232Th as well as the absorbed dose rate than UNSCEAR world average values (Table 1). Soil samples from other parts of Kanyakumari also showed a high-level activity concentration of 232Th and an absorbed dose rate (Malathi et al., 2005). Similar observations paved a way to study the fate of radionuclides in food crops grown in the Kanyakumari region, since the radionuclides can transfer from soil to plant by root uptake. Major foods in the diet of Kanyakumari residents include rice and tapioca and the Th activity concentrations in both were reported to be higher than the activity concentrations of other radionuclides (Shanthi et al., 2009). For an adult in this region, the daily radionuclide intake and daily dose through ingestion of food crops were nearly 127.7 Bq day−1 and 2.3 µSv day−1 which were relatively higher when compared to values for ingestion of food crops grown in areas where the background radiation is low (Shanthi et al., 2009).

Statistical approach

The standard deviation values were similar to or higher than the mean value for the activity concentration of the radionuclides as well as for the radiological hazard parameters, and this indicates a low degree of uniformity of the distribution of radionuclides. The overall kurtosis of 40K was found to be platykurtic, suggesting a near uniform distribution, while the leptokurtic nature of the activity concentrations of 226Ra and 232Th and the absorbed dose rate shows more variance with infrequent extreme deviations (Fig. 6).

Fig. 6
figure 6

Frequency distributions of activity concentrations of a 226Ra, b 232Th, and c.40K. d Absorbed dose rate frequency distribution (n = 44)

Full size image

The skewness shows the asymmetry of the probability distribution. The 40K activity concentration of the river and stream sediments showed a normal symmetrical distribution (Fig. 6c). However, the 226Ra and 232Th and absorbed dose rates were right-skewed with an asymmetric tail extending towards the values that were more positive. Based on the frequency distributions, about 75% of the activity concentration of 226Ra, 232Th, and 40K for the collected samples were under 106 Bq kg−1, 685 Bq kg−1, and 586 Bq kg−1 (Fig. 6a, b, c), whereas 50% of the absorbed dose rates were in the range of 115 to 485 nGy h−1 (Fig. 6d).

The principal component analysis (PCA) is one of the most powerful mathematical tools to reduce the dimensionality of data sets and it helps in understanding the correlation of the different variables which could lead to source discrimination. To carry out the PCA and for an easy interpretation, varimax rotation using Kaiser normalization was applied which helps in maximizing the component loadings variance and eliminating invalid components (Dragović & Onjia, 2006). The factor analysis supports extracting the eigenvectors from the correlation matrix and the two significant components obtained are given in Table 3. The scree plot using eigenvalues for the sediment samples is shown in Fig. 7. The sharp falloff from the first to the second component (Fig. 7) indicated that the first component (PC 1) accounted for most of the data variability (Kumar et al., 2012). About 97.15% of the total variance was explained by these two components, PC 1 accounted for 86.80% and PC 2 accounted for 10.35%. The PC 1 was distinctive due to the positive loading of 226Ra and 232Th, whereas PC 2 was controlled by the positive loading of 40K (Table 3).

Table 3 Varimax rotated components of the variable
Full size table
Fig. 7
figure 7

Scree plot of eigenvalues for Kanyakumari River and stream sediments

Full size image

Figure 8 shows the effect of components PC 1 and PC 2 on the radionuclides 226Ra, 232Th, and 40K of sediments from selected sampling locations as the factor score. The factor score can distinguish the influence of each radionuclide; here, PC 1 was the loading of 226Ra and 232Th and PC 2 was influenced by 40K. The highest PC 1 scores can be seen in the sediment samples collected upstream for the TR (TR–3, –4, and –6) (Fig. 8a), and that shows these sediment samples were enriched with 226Ra and 232Th. Most of the sediment samples from the PR had high PC 2 scores (Fig. 8b) due to the direct influence of the low activity concentration of 226Ra and the high activity concentration of 40K. Nearly all SS sediment samples showed similar factor scores for both PC 1 and PC 2 except SS–2 (Fig. 8c), which was controlled by the high 232Th activity concentration (Table 1). Figure 9 shows the cluster plot generated for the sediment samples using the PCA results. The sediment samples from the rivers and seasonal streams are grouped separately and the ellipses represent the 95% confidence level. The overlapping of all three groups might indicate a similar origin for the sediments.

Fig. 8
figure 8

Factor score of sediments from a TR, b PR, and c SS (TR, Tamiraparani River; PR, Pazhaiyar River; SS, seasonal stream)

Full size image
Fig. 9
figure 9

Principal component analysis (PCA) with the cluster of radionuclides (TR, Tamiraparani River; PR, Pazhaiyar River; SS, seasonal stream)

Full size image

Conclusion

The activity concentrations of the natural radionuclides 226Ra, 232Th, and 40K and their associated risk parameters were evaluated for river and stream sediment samples collected in the Kanyakumari HBNRA. The present investigation, the first of its kind for the Kanyakumari area, revealed that the mean activity concentrations of 226Ra and 232Th for the sediments were relatively higher than the UNSCEAR world average values. The radiological hazard indices such as absorbed dose rate (D), Raeq, AEDEout, and AEDEin were estimated and the means of all these indices were found to be higher than the UNSCEAR world average values. The high 232Th/226Ra ratio in the sediments could explain the removal of 226Ra in the particulate phase due to the terrestrial oxidizing environment and the high ratios supported the presence of Th-bearing minerals. Due to the presence of clay minerals, the 232Th/40K ratio of PR sediments was relatively lower than the TR and SS sediments; however, the 232Th/40K ratio was higher than the world average. The results of the multivariate statistical analysis of the sediments were able to explain the uneven distribution of the radionuclides, even though the sources of the sediments were similar. Based on these findings, the Kanyakumari sediments were concluded to be enriched with 232Th and that had a significant influence on the natural radioactivity of the area. This study is expected to be useful in understanding whether these river sediments pose any radiological risk when they are used for construction purposes. The results can help in understanding the source of natural radionuclides deposited along the Kanyakumari coast, and a geochemical and mineralogical study would provide further details.