Content uploaded by Gauri Datar
Author content
All content in this area was uploaded by Gauri Datar on Nov 28, 2019
Content may be subject to copyright.
Draft version November 26, 2019
Causes of the Diurnal Variation observed in γ-ray Spectrum using NaI (Tl) Detector
Gauri Datar,1Geeta Vichare,1C. Selvaraj,2Ankush Bhaskar,3,4 and Anil Raghav5
1Indian Institute of Geomagnetism, Navi Mumbai, 410218, India
2Equatorial Geophysical Research Laboratory, Tirunelveli, 627011, India
3Heliophysics Science Division, NASA/Goddard Space Flight Center, Greenbelt, 20770, USA
4University Corporation for Atmospheric Research, Boulder, CO, 80307 USA
5University of Mumbai, Department of Physics, Santacruz (Mumbai), 400098, India
ABSTRACT
Present paper studies the γ-ray spectrum observed using NaI(Tl) scintillation detector over a year.
The paper reports the presence of a distinct diurnal pattern in the total number of γ-ray counts de-
tected by the NaI detector. The counts start decreasing after sunrise and show gradual recovery after
sunset. The amplitude of this variation is quite significant ( >10%) at the observation site Tirunelveli,
South India. However, further investigation based on different energy ranges reveals that the men-
tioned diurnal pattern is actually present only in the energies related to the terrestrial background
radioactivity. The study demonstrates that the pattern is associated with the radioactivity of isotopes
of radon (222Rn and 220 Rn ) and their daughter radionuclides. The amplitude of the diurnal variation
is found to have seasonal dependence, with the lowest amplitude during June–October (∼2%), and
highest in April–May months (∼14%). The dependence of the amplitude of the diurnal variation on
various atmospheric parameters, namely temperature, relative humidity, and vertical wind is examined.
The observed diurnal pattern is attributed to the generation of the inversion layer. The distribution
of concentration of radon and thoron progenies in an air column changes with the vertical mixing and
atmospheric boundary layer (ABL), resulting in the diurnal variation of γ-rays.
Keywords: NaI (Tl) detector; Diurnal pattern in γ-ray; Background radioactivity; 222Rn
1. INTRODUCTION
The surrounding atmosphere contains abundant radiation, with a daily influx from the Sun, cosmic rays (CR), in
addition to radiation of terrestrial origins. The near-Earth environment is abundant with a vast number of particles of
various energies and a broad spectrum of radiation. One of the primary components of radiation affecting humans is
γ-rays of varying energies. Due to very high penetration power, γ-rays can damage human organs and bone marrow,
thus making their high exposure biologically hazardous. However, all of us are regularly exposed to a certain level
of radiation, i.e., ‘Background radiation’, that collectively represents all types of radiation together that are present
in our surroundings at all times. The significant natural sources of γ-rays on Earth are (1) terrestrial radioactive
decay, and (2) secondary particles generated during the interaction of CR with atmospheric nuclei (on the Effects of
Atomic Radiation 2000). However, there are other natural sources as well that have a relatively lesser contribution,
e.g., lightning, terrestrial γ-ray flashes (TGF) in thunderstorm produce γ-rays (Fishman et al. 1994;Briggs et al.
2013). Furthermore, fission in nuclear reactors, high energy physics experiments, nuclear explosions, and accidents are
some of the artificial sources of γ-rays.
Even though background radiation is present everywhere on Earth, the extent can differ from place to place, especially
where human-made radiation can be in high concentration. Background radiation is measured in such areas to monitor
the radiation dose received by the general population. The most significant radionuclides that enter the body are of
terrestrial origin. Terrestrial radioactivity is present mainly due to radiations from certain radioactive isotopes present
in the Earth’s crust since its formation. Among the radioactive isotopes, thorium (232Th), uranium (235 U, 238U), and
Corresponding author: Gauri Datar
datar.gouri@gmail.com, gdatar16@iigs.iigm.res.in
arXiv:1911.05137v2 [physics.ao-ph] 25 Nov 2019
2Datar et al.
the daughter radionuclides formed in their decay chains are the major contributors (the decay chains are shown in
the Appendix). These isotopes and their daughter radionuclides emit γ-rays of different energies until a stable isotope
is reached in the chain (Knoll 2010). Among these radionuclides, radon gas and its decay products constitute the
principal part of the dosage that we inhale constantly. Radon levels in a particular geographic region depend on the
uranium and thorium content of the soil. The 222 Rn decay products affecting environmental γ-ray intensity are 214Pb
and 214Bi.
As 222Rn is chemically inert, and it has a well-defined source and sink (Jacobi & Andre 1963); it is considered as
a useful tracer for studying the vertical mixing of the pollutants in the atmospheric boundary layer (ABL) (e.g., Liu
et al. (1984); Galeriu et al. (2011)). Atmospheric abundance of radon is connected with its exhalation rate from the
soil, which is estimated by Wilkening et al. (1972) as the mean flux of 0.75 atom cm−2sec−1with a range from 0.01
to 2.5 atoms cm−2sec−1or 0.72 to 1.2 atoms cm−2sec−1for ice-free terrains (e.g., Griffiths et al. (2010)). It is also
affected by factors such as soil composition, moisture content, porosity, and permeability Szegvary et al. (2009); Minty
& Wilford (2004); Manohar et al. (2013); Arora et al. (2017). Besides, to understand how the radon concentration
varies with height, the vertical profile of radon concentration has been studied for a range of altitudes (near-surface to
stratosphere), and geographical locations (inland, coastal, islands, and open ocean), e.g., Williams et al. (2011) (and
other studies enlisted therein in Table 1), Baldoncini et al. (2017).
Many studies have investigated the relationship between rainfall and 222Rn concentration (Takeuchi & Katase 1982;
Minato 1983;Horng & Jiang 2003;Gusev et al. 2015). The diurnal variation in the concentration of radon and its
relation with the metrological parameters of temperature, relative humidity, and pressure has been studied by many
researchers such as El-Hussein et al. (2001); Schubert & Schulz (2002); Desideri et al. (2006); Galeriu et al. (2011);
Banjanac et al. (2012); Victor et al. (2019). Further, the variations of radon concentration in the atmosphere in
different seasons have been studied, and it was found by Tchorz-Trzeciakiewicz & Solecki (2018) that the highest
radon concentrations were during autumn and the lowest during winter. Besides the diurnal and seasonal variations
in the atmospheric radon, there have been reports on the diurnal variations of the radon concentration in the upper
soil layers and at the soil-air interface with meteorological parameters (Schubert & Schulz 2002;Victor et al. 2019).
As discussed earlier, the other significant source of γ-rays on Earth is cosmic rays in the atmosphere. CR continually
arrive at Earth, and after entering the atmosphere, they tend to collide with atmospheric atoms and molecules. Further,
the spallation process produces a variety of lighter particles, which are called secondary cosmic rays (SCR). In general,
muons and neutrinos originate in the decay chain of charged mesons, while electrons and photons are produced as a
result of the decays of neutral mesons (Patrignani et al. 2016). Photons are generated due to annihilation processes
as well (e.g., e−e+,µ−µ+). Among SCR, the most abundant particles at sea level are muons, neutrons, neutrinos,
photons. Muons with energies greater than 250 MeV are the most abundant secondary particles, while photons are
predominant in the energy range between 100 keV to ∼10 MeV.
Chin & Standil (1968a) studied the variation of the γ-rays (photon component of SCR) in the energy range 3.8 –
183 MeV using NaI (Tl) spectrometer. They observed the dependence of the time of maximum amplitude on energy
using 16 energy bands in the range 3.8 – 183 MeV. For energies <50 – 60 MeV, Chin & Standil (1968a) observed
that the times of maximum intensity move towards later hours with increasing energy, and the amplitudes decrease
with energy. In their other paper, using the same dataset, Chin & Standil (1968b) found that the photon intensity
depends only on the atmospheric pressure, and does not vary significantly with the ground temperature. Katase
et al. (1982) reported the presence of a diurnal pattern in the intensity of γ-rays obtained using Ge (Li) spectrometer.
Recently, using NaI(Tl) observations, Vichare et al. (2018) described the diurnal variation of the γ-ray counts, and
a similar pattern was also noticed by Raghav et al. (2013,2015). However, none of these have studied the source of
this variation. In fact, there are only a few studies to date, focusing on the diurnal pattern of the γ-ray spectrum.
The present paper aims to conduct an in-depth analysis of the diurnal pattern observed in the γ-ray flux and the
parameters responsible, using γ-ray data spanning one year from January 2017 to December 2017. Such long term
data also facilitates examining the seasonal dependence, if any.
2. DATA DESCRIPTION
The γ-ray flux data used in the present analysis is obtained from a rectangular cuboid sized NaI (Tl) scintillation
detector, of dimension 10.16 cm ×10.16 cm ×40.64 cm (4” ×4” ×16”), installed at Tirunelveli (Geographic Coordi-
nates: 8.71◦N, 77.76◦; Geomagnetic Coordinates: 0.28◦N, 151.01◦E) in South India. The detector is surrounded by a
lead shield box of 5 cm thickness, which is open from the top. The entire set up is kept inside an air-tight, temperature-
Causes of the Diurnal Variation observed in γ-ray Spectrum using NaI (Tl) Detector 3
Figure 1. Background spectrum observed in NaI(Tl) data; x-axis shows energy in MeV, and the corresponding counts on y-axis
are accumulated in one hour.
controlled cabin, to minimize the thermal electronic noise due to the ambient temperature variations. Each minute, a
data file is generated containing energy histogram (counts corresponding to different energies) accumulated over one
-minute interval. The system is equipped to acquire an energy spectrum of photons ranging from ∼150 keV to ∼10
MeV. The energy spectra are calibrated regularly (weekly) using standard sources (60Co and 137 Cs) and background
radioactivity peaks marked in Figure 1. More details of the Equatorial secondary cosmic ray observatory (ESCRO)
and experimental set-up can be found in Vichare et al. (2018). Atmospheric parameters, recorded using an in-house
automatic weather station (AWS) and the India Meteorological Department (IMD-AWS) data at Tirunelveli, have
been used in the present study. Vertical velocity (m/s) parameter is obtained by modifying vertical velocity (Pa/s)
reanalysis data by Copernicus Atmosphere Monitoring Service Information [2017]. As mentioned in the previous sec-
tion, SCR and the radioactivity of primordial radionuclides and their daughter nuclides have a significant contribution
in the γflux measured on the ground. Therefore, in the energy spectrum obtained using NaI (Tl), a contribution from
the γcomponent of SCR is present along with distinct photopeaks due to background radioactivity over a continuum.
For the convenience, the decay chains of 232Th, 235 U, and 238U are displayed in the Appendix section. One photopeak
of 40K, two of 208Tl, three of 214 Bi, and one of 214Pb are clearly observed in the present data, as evident from Figure 1.
Among which, 40K is a primordial radionuclide, 208Tl is a daughter radionuclide of 232Th produced in the decay chain
via 220Rn, while 214 Bi and 214Pb are daughter radionuclides of 238U produced in the decay chain via 222Rn (Refer to
the 232Th, 235 U, and 238U decay chains visualised in the appendix). Except for 214Pb (energy 352 keV) that is too
feeble in our spectrum, all the other observed photopeaks lie in between 500 keV – 2.7 MeV.
3. OBSERVATIONS
3.1. Diurnal pattern and its energy dependence
When total γ-ray counts accumulated over each minute is plotted against time during the day, it shows a distinct
pattern similar to that reported by Vichare et al. (2018). The pattern shows a decrease in the γ-ray counts after the
sunrise and recovery after the sunset. To study this variation with respect to energy, we applied filters with different
cut-off energies. The application of a particular cut-off filter gives the total number of photons from all the energies
above that particular energy, detected by the NaI (Tl) detector. Here, we have used 200 keV, 500 keV, 1 MeV, 1.5
MeV, 2.7 MeV, and 5 MeV as different energy cut-offs. The original spectrometer data of 1 min resolution is averaged
over 5 min period. Temporal profiles of the counts with different energy cut-offs, on 29 December 2017 are shown in
Figure 2.
As explained above, the counts with 200 keV cut-off contain particles with energies above 200 keV. Similarly, for
subsequent panels, counts are shown for the corresponding cut-off energies. It is observed from the plots depicted in
Figure 2that a clear pattern is seen in the plots with cut-off energies from 200 keV to 1.5 MeV. Around 1-2 UT, counts
4Datar et al.
Figure 2. γ-ray data on 29 December 2017 is depicted for total six energy cut-offs ranging from ∼200 keV to ∼10 MeV plotted
in subplots with energy as mentioned above each subplot. Number of counts accumulated in one minute are indicated on y-axes.
start decreasing, and after 13 UT, the increase is started. Note that the local time corresponding to the onset of the
decrease is near sunrise, while the beginning of the increasing trend corresponds to the local sunset. The counts for
energy cut-offs higher than 2.7 MeV do not show the pattern described above; instead, they show no variation with
the time of a day, indicating the absence of any diurnal pattern for the energies above 2.7 MeV.
Therefore, it is appropriate to examine the variation in different energy bands or energy ranges. As mentioned
above, the background radioactivity dominates in 500 keV – 2.7 MeV, hence we have considered this range as one of
the energy band. Energies lower than this band are grouped as a lower energy range, i.e., 200 keV – 500 keV. The
counts in the energy range between 2.7 – ∼10 MeV is selected as a higher energy band. As the number of counts
in an energy band varies due to the background radioactivity variations, it is appropriate to consider the percentage
variations calculated with respect to the minimum count in that energy band. Figure 3shows the diurnal variation of
the % change in counts in three energy bands. The amplitude of the diurnal variation is defined as the maximum %
variation during the day.
In Figure 3, the x-axis shows time in UT; the corresponding energy range is mentioned in each panel. We have
shown sunrise and sunset (red and blue dashed lines) and solar noon (yellow dashed line) times at the observation site
on the presented day (29 December 2017). It is observed that the pattern similar to that observed in Figure 2that
is present in the first two energy bands, 200 – 500 keV and 500 keV – 2.7 MeV. The % of counts starts decreasing
after around 1 UT and follows a decreasing trend up to around 13 UT. After sunset terminator, the counts start to
increase, indicating recovery of the counts. It should be noted that though the pattern in the first two energy bands
looks similar, there is a significant difference in the amplitude of the diurnal variation in both the energy bands. In
the middle energy band (500 keV-2.7 MeV), it is about 12%, while in the lower energy band, it is less than 2%. This
indicates that the major contribution to the diurnal pattern of γ-rays is coming from the middle energy band, which is
primarily associated with the background radioactivity. There are a few γ-flux peaks, e.g., 214Pb (of energy 352 keV),
along with the Compton component of other higher energy photopeaks due to background radioactivity in the lower
energy band as well, which might have caused the observed diurnal pattern in that band. It can be observed from the
bottom panel of Figure 3that the higher energy band (2.7 – 10 MeV) does not show any discernible diurnal pattern.
Thus, it is observed that the significant diurnal variation exists only in the middle energy band, and hence, hereafter,
the diurnal variations only for 500 keV – 2.7 MeV energy band are presented.
3.2. Seasonal variation
Causes of the Diurnal Variation observed in γ-ray Spectrum using NaI (Tl) Detector 5
Figure 3. Temporal variation of γ-ray counts on 29 December 2017 for three different energy bands: 200 – 500 keV, 500 keV
– 2.7 MeV, 2.7 – 10 MeV. Variation in percentage in counts per minute are indicated on y-axes. Vertical red, blue, and yellow
dashed lines indicate local sunrise, sunset, and noon times, respectively.
Table 1. List of fair weather days in 2017 used in the analysis
Month Day 1 Day 2 Day 3
January 10 14 23
February 05 11 18
March 13 24 28
April 01 08 27
May 01 05 06
June 08 10 22
July 22 25 29
August 03 21 27
September 03 21 30
October 02 17 20
November 11 21 24
December 11 28 29
In order to examine the robustness of the observed pattern on statistical grounds, three days from each month are
selected for the year 2017, based on fair-weather conditions such that the maximum wind speed does not exceed 5 m
s−1.
Figure 4shows the daily variation in the energy band 500 keV – 2.7 MeV for the months January to December 2017.
Each subfigure depicts the averaged diurnal pattern on three different days, indicated in Table 1. It is observed that
in January and February, the γ-ray counts on individual days follow the pattern of decrease around 1 UT and increase
after around 13-14 UT. The amplitude of variation in January and February is seen to be around 5.6% and 7.2%,
respectively. For March, April, and May 2017, a range of variation is 9.2%, 10.3%, 10%, respectively. In June 2017,
the variation fell to 3.7%, while in July and August, the pattern becomes very weak with just 1.8% and 1.7% variation,
6Datar et al.
Figure 4. Each panel depicts the average of % variation of γ-ray counts with energies between 500 keV – 2.7 MeV on three
different days in each month from January to December 2017. Vertical red, blue, and yellow dashed lines indicate local sunrise,
sunset, and noon times, respectively.
Figure 5. Observed diurnal variation of the γ-ray flux on individual days; x-axis shows day of year (DOY)
respectively. In September and October, it slightly increases to 3.1% and 3.2%, respectively. Although the range of
variation is small for months from June to October, the diurnal pattern is still similar. In November, the variation
increases to 6.4%; for December, it further increases to 9.5%. To visualise the trend of change in percentage variation,
Figure 5shows the range/amplitude of the diurnal variation on individual days (Listed in Table 1) against the day of
the year (DOY). It is clear from the figure that the amplitude of diurnal variation is minimum in June-October months
and maximum in April-May months. Thus γ-flux recorded on the ground exhibits a considerable seasonal dependence.
4. ANALYSIS OF INDIVIDUAL BACKGROUND PHOTOPEAK
Causes of the Diurnal Variation observed in γ-ray Spectrum using NaI (Tl) Detector 7
Figure 6. The diurnal pattern1 (average of three days) of (i) 208Tl (2.614 MeV), (ii) 214Bi(1.764 MeV), (iii) 40 K (1.461 MeV),
and (iv) 214Bi(0.609 MeV) for each month from January to April 2017.
It is evident from Figure 3that the diurnal variation is most prominently seen in 500 keV – 2.7 MeV energy band. As
background radioactivity is dominant in this energy range, one can examine whether any particular radioactive elements
are causing the observed pattern. In order to inspect this, we have estimated the counts in different photopeaks of
radioactivity by calculating the area under the photopeak and checked whether it exhibits any diurnal pattern.
As discussed in ‘Data description’ (Section 2), there are several photopeaks between 500 keV – 2.7 MeV. To identify
the peaks clearly, we have accumulated data for two hours. For the selected days, total flux is calculated (accumulated
counts) with a two-hour window at each hour. Thus, the two-hour sliding window has a one-hour overlap so that 23
data points per day are obtained. Following four photopeaks are chosen for further analysis, as they are significant and
very clear in the spectrum: 214Bi (0.609 MeV), 40K (1.461 MeV), 214 Bi(1.764 MeV), 208Tl (2.614 MeV). By calculating
the area under a photopeak using the trapezoidal method of integration, we estimated the number of photons emitted
by the element corresponding to that particular photopeak.
Figures 6,7, and 8together depict the variation in the counts under the photopeaks of radionuclides 214Bi (0.609
MeV, 1.764 MeV), 40K (1.461 MeV), and 208 Tl (2.614 MeV). The mean of three days’ counts for a month is presented
in a subfigure for the annotated radionuclides. It is observed from the plots that the temporal variation of the counts
under almost all the selected photopeaks shows a pattern similar to that described in the previous section. The pattern
8Datar et al.
Figure 7. Same as Figure 6, but for each month from May to August 2017.
is quite robust and intense for two peaks of Bi and one peak of K, but not very clear in some cases for 208 Tl. But
more or less, a similar diurnal pattern is observed in all the selected photopeaks. It is noted that the percent change
in the counts varies for different photopeaks in a given month. The amplitude for two 214Bi peaks is quite large (e.g.,
∼15-20% during March-April months) and that for 208Tl is very small (∼3-4%) as compared to other peaks. The
diurnal amplitude for 40K is intermediate (∼6-8%). From Figures 6,7, and 8, it can be observed that the percent
change in the counts of each photopeak varies from month to month in year 2017. Except for 208 Tl, the other three
photopeaks show maximum variation during April, and the amplitudes are lowest during June-October. In June,
214Bi(1.764 MeV) shows 7%, 40 K shows 3%, and 214Bi (0.609 MeV) shows 6% variation. The diurnal pattern also
seems to deviate from the general trend in the months of June as well as July. In July, 214Bi(1.764 MeV) shows 3.5%,
40K shows 1.5%, and 214 Bi (0.609 MeV) shows 1.5% variation. The diurnal pattern is significantly disturbed in the
month of July for all the photopeaks. The pattern is restored from October onwards. For 208Tl peak, the daily range
of variation is ∼3-4% and do not show any seasonal dependence. The seasonal variation in each photopeak is depicted
in Figure 9. It is evident from the figure that the 214Bi (1.764 MeV) has the strongest seasonal dependence, 40K has
moderate dependence, and 208 Tl has almost no seasonal dependence. Notably, an increase in the month of April is a
feature present in all four photopeaks. It can be remarked that the seasonal dependence of the 214 Bi peaks is similar
Causes of the Diurnal Variation observed in γ-ray Spectrum using NaI (Tl) Detector 9
Figure 8. Same as Figure 6, but for each month from September to December 2017.
to that of γ-flux depicted in Figure 5, indicating that the diurnal pattern observed in the γ-flux is mainly due to the
214Bi radionuclide.
5. DISCUSSION
The present work investigates the diurnal variation of the γ-ray spectrum recorded by the NaI (Tl) scintillation
detector. It is demonstrated here that there exists a diurnal pattern in the γ-ray flux measured near the ground. The
γ-ray counts start decreasing at a time coinciding with local sunrise, and the decrease is observed up to local sunset,
after which, the counts start increasing. This diurnal pattern is observed almost on all fair weather days throughout the
year, thus indicating robustness. There are few earlier works reporting the similar diurnal trend noticed in the γ-rays
detected by Ge(Li) detector (Katase et al. 1982) and NaI(Tl) detector (Raghav et al. 2013,2015;Vichare et al. 2018).
The present study carried out a detailed analysis of this pattern and attempted to identify the possible responsible
factors.
In this context, we examined the pattern of the γ-flux of different energies. As mentioned in the introduction, the
γ-rays observed on the ground can have contribution due to background radioactivity and secondary particles produced
in the cosmic ray interaction with the atmospheric particles. The terrestrial background radioactivity is, in general,
observed in the energy range between 500 keV to 2.7 MeV. Therefore, it is appropriate to inspect the response in
10 Datar et al.
Figure 9. Seasonal variation in the amplitude of individual photopeaks: x-axis indicates month of year 2017; y-axis indicates
the amplitude of average diurnal variation in counts under the photopeak in %.
different energy ranges to identify the responsible contributor in the observed diurnal pattern. It is observed that a
strong diurnal pattern is unambiguously present in the 500 keV– 2.7 MeV energy range, and completely absent in the
higher energies (>2.7 MeV). As the γ-rays (photons) due to cosmic ray interactions can exist in the entire 100 keV– 10
MeV energy range, it can be concluded that the observed diurnal variation in the γ-ray flux could be only due to the
background radioactivity. Normally, in the energy spectrum collected at the observation site, four very distinct peaks,
including 40K (1.461 MeV), 208 Tl (2.614 MeV), and two peaks of 214Bi (0.609 MeV, 1.764 MeV) are clearly visible.
We found that the diurnal variations of the counts present in these photopeaks exhibit a diurnal pattern similar to
that observed for the γ-flux. The pattern is quite robust and intense for the two peaks of 214Bi and 40 K, whereas a
weak pattern is observed for 208Tl. Roughly a similar diurnal pattern is observed in all the selected photopeaks.
Then the next question is what is the source for the observed diurnal pattern in the γ-ray flux. Katase et al. (1982)
ascribed the diurnal variation of γ-ray intensity to the fluctuations in the atmospheric concentration of 222Rn, which
induce the variation of concentrations of 214Pb and 214 Bi components in the atmosphere, resulting in the variation of γ-
ray flux. They attributed diurnal variation to the generation of ‘inversion layer’ and its dissipation after sunrise. Their
findings were based on the data collected on some days during three months (January, July, and November). There
are many papers reporting the diurnal variation in the concentration of radon [El-Hussein et al. (2001); Schubert &
Schulz (2002); Desideri et al. (2006); Galeriu et al. (2011); Banjanac et al. (2012); Victor et al. (2019)]. These reported
patterns are similar to that of γ-ray flux. Usually, the radon gas is exhaled from the ground, and its intensity depends
on the geological and soil properties of the observation site. It is generally considered that the diurnal variation in
radon concentration near the ground occurs due to the difference in the ABL or mixing height during day and night.
Because of atmospheric heating and cooling, contraction and expansion of the air envelope occur. Being in the gaseous
form, transportation of the radon gas in the air after its exhalation from the soil is quite plausible. On fair weather
days, once exhaled from the soil, depending on the atmospheric temperature and mixing height, radon can reach up to
greater heights (day times) or can get accumulated in the lower heights (night times). Under non-turbulent conditions,
during night-times as the temperature of the surface-adjacent atmospheric layer is cooled down, the exhaled radon
does not get transported to the higher heights, and an inversion layer is formed keeping the radon accumulated near
the surface. When the sun rises again, the temperature starts increasing; thus, a thin but stable boundary layer formed
due to cooling gets perturbed. As a result, even though the exhalation from soil continues, the concentration detected
at different heights during different times differ depending on the mixing height or ABL (Galeriu et al. 2011). This
pattern is well-defined during sunny days or warm months in general (El-Hussein et al. 2001;Desideri et al. 2006), as
the solar heating-induced vertical mixing is more efficient in the warmer season.
Causes of the Diurnal Variation observed in γ-ray Spectrum using NaI (Tl) Detector 11
Figure 10. Scatter plots between the diurnal amplitudes of γflux and temperature variations (left) and relative humidity
(RH)(right) for individual days (labels show the corresponding month).
Another approach is based on the fact that the temperature difference between soil and air is different during the
day and night, which can result in the difference in the exhalation of radon from the ground. Schubert & Schulz (2002)
reported that during night-time, the temperature of the soil is higher than the temperature of the air above; hence,
there is a more pronounced upward-directed convective flux of radon. This increases the overall diffusive transport,
which results in the maximum concentration in the morning. When the situation is reversed, a downward convective
flux emerges, reducing the overall upward diffusive radon transport. This also explains why there is a decline in γflux
during the daytime (Schubert & Schulz 2002). However, Gogolak (1977) have pointed out that the diurnal variations in
the exhalation rate have a small effect on concentration profiles compared to that of changes in the vertical diffusivity.
Recently, Victor et al. (2019) presented results on diurnal and seasonal variations of 222Rn using a radon detector
(RTM 2200) at Pune, India. They have explained the results in terms of surface meteorology, soil moisture, soil
temperature, vertical wind velocity, local topography, and development of ABL. They report that 222Rn concentration
increased with soil moisture content of <16%, rapidly dropped down for dry soil values and then remained almost
constant with increasing soil moisture (Victor et al. 2019); while Hosoda et al. (2007) had showed experimentally that
radon exhalation rates increased with increase in moisture content up to 8% but then showed a decreasing tendency
for soil moisture >8%. According to Schery & Whittlestone (1989), when temperature increases or moisture content
rises due to rainfall in a very dry, clay-rich soils, more radon will be exhaled.
In the present study conducted during January-December 2017, a strong seasonal dependence of the diurnal am-
plitude (indicated by the daily range of variation) is observed. In the months between June-October, the diurnal
amplitude is very weak, although the pattern is still present. The seasonal variation of the diurnal variation is visu-
alised in Figure 5. The seasonal variation of the two photopeaks of 214Bi and 40 K also follow a similar trend, whereas
208Tl shows very weak seasonal variation. Thus the present analysis indicates that the diurnal pattern observed in γ
flux is essentially due to 214Bi radionuclide present in the background radioactivity and some contribution from 40K
and 208Tl. Referring to the decay chains mentioned in the Appendix, observed two photopeaks of 214Bi are the daugh-
ter radionuclide of 222Rn. Also, 208 Tl is a daughter radionuclide of another radon isotope, 220Rn, known as ‘thoron
gas’. The radiation dose from 220Rn and its daughter products is much smaller than 222Rn and its daughters (on the
Effects of Atomic Radiation 2000); this could also be because of the short half-life of 220 Rn (55 s, against 3.8 days of
222Rn). Thus, the weak variation observed in 208Tl can be attributed to the lesser concentration of 220 Rn than 222Rn.
However, 40K does not fall in the decay chain. Here, one should note that there exists some increase in the counts of
the energies lower than the actual photopeak due to the Compton effect. Likewise, associated with the photopeak of
214Bi (1.764 MeV), there is a major Compton scattering component present in the lower energies coinciding with the
40K (1.461 MeV) photopeak, which can affect the counts under the photopeak of 40K. Thus, the pattern reflected in
the counts under 40K (1.461 MeV) could be actually the Compton component of 214Bi (1.764 MeV).
Thus, the present study has demonstrated that the diurnal pattern in the γ-rays is because of the isotopes of radon
(220Rn, 222 Rn), and their daughter radionuclides. It is not only 222Rn but also 220 Rn that contributes to the diurnal
variations, in spite of its weak and short-living nature. El-Hussein et al. (2001)) had studied the radon progeny
12 Datar et al.
Figure 11. Vertical velocity (m/s) from January to December 2017 (’Contains modified Copernicus Atmosphere Monitoring
Service Information [2017]’).
concentration in the atmosphere for diurnal and seasonal changes using one-year data; however, they had used α-
spectrometry. Although they conclude the same about diurnal variation, they suspect that the observed considerable
seasonal variations are due to meteorological factors. They also found that in cloudy and windy weather, the difference
between the short-lived radon progeny concentration during the day and night was small. Their remark about wind is
that turbulence is increased by a strong wind, which tends to distribute radon progeny over a large area and to a great
vertical depth in the atmospheric air. However, this does not explain the low variation we observed on days when the
wind is not exceeding 5 m s−1(viz. days from June to October).
To understand the effect of meteorological parameters on the amplitude of γflux, it will be interesting to see the
range of temperature and humidity against the amplitude of γ. In Figure 10, range of γflux variation (∆), in % and
the maximum difference in temperature variation in a day (∆T), in oC as well as relative humidity (RH) is compared
for individual days. The days are selected from the list in Table 1as per the availability of temperature data, either
from in-house AWS or IMD AWS. Additionally, three more days are selected apart from the list with 5 <∆T < 7
oC, to include the lower range of temperature below 7 oC. A least square linear fit is shown by the solid line. Diurnal
amplitude and ∆Tdo show a good correlation with the Pearson correlation coefficient (CC) of 0.609. The data points
which are out of the general linear trend are from the months June to October (indicated by the labels) and have very
low diurnal amplitude. For the relative humidity (Figure 10b), the CC value is 0.797.
Victor et al. (2019) observed that the diurnal variations of radon concentration during winter, post-monsoon, and
pre-monsoon seasons were similar in shape but with decreasing amplitudes. However, during the monsoon season,
radon concentration was the lowest and remained almost constant throughout the day (Victor et al. 2019). They
explained their results based on the variations in soil moisture due to the monsoon and capping effect of precipitation.
In our case as well, it is clearly evident from the observations in Figures 4and 5that the percentage change is minimum
from June to October. However, the observation site in the present work receives rainfall normally in the months of
October and November during the northeast monsoon, while occasional rains occur from June to September–when the
southwest monsoon prevails in the Indian subcontinent (Panneerselvam et al. 2007). Thus the main monsoon months
at the observation sites of Victor et al. (2019) and ours are different, and yet the lower amplitudes of radon and its
daughter nuclides occur during the same months, which is intriguing. The explanation for the present observed results
is sought in terms of the winds, as presented below.
Moderate-to-intense southwesterly winds dominate during one half (AprilSeptember) of the year over the observation
site, and moderate north-easterlies during the other half (OctoberMarch) of the year (Panneerselvam et al. 2007).
From the vertical wind velocity profile presented in Figure 11, from the end of May to September, negative values, i.e.,
Causes of the Diurnal Variation observed in γ-ray Spectrum using NaI (Tl) Detector 13
downward winds prevail at altitudes 0.5 km – 2 km. This suppresses the vertically upward propagation of radon, and
the radon exhaled from the soil remains in this thin layer of the atmosphere. This would result in the lower amplitude
of diurnal variation, as radon would remain in a thin layer in the lower atmosphere during the day as well as night. The
mixing height for radon would not change from day-time to night-time as compared to the months when the vertical
wind was positive, i.e., upwards, and mixing heights would vary extensively for day and night. Intermittent rain spells
can increase the soil moisture releasing radon until it reaches a certain limit (as specified by Hosoda et al. (2007); Victor
et al. (2019)). After a spell of rainfall, γ-ray flux can be enhanced due to rainout and washout processes (Takeuchi &
Katase 1982); however, those are short term enhancements and are not considered in the diurnal amplitudes.
As mentioned earlier, the observation site receives rainfall mainly due to northeastern monsoon, while southwest
monsoon is not experienced significantly. However, the neighbouring state Kerala experiences a heavy (southwest
and northeast) monsoon. The orographic feature of the Western Ghats creates such a situation. Particularly, we can
consider the case of Agasthyamalai hills situated at ∼45 km away (West) from the observation site. Although the
tallest peak in Agasthyamalai hills is ∼1800 m; generally, these hills have an elevation of around 1000 m or even
less. Even though the southwest monsoon does not cross these hills to much extent, strong horizontal winds can travel
across the mountain range, affecting the atmosphere on the other side (i.e., Tamil Nadu). The intense southeasterly
horizontal winds can affect the vertical motion of air parcels in an air column, consequently producing downward
dominated vertical winds. This would explain why wind observations at Tirunelveli, Tamil Nadu are similar to those
made from Pune, Maharashtra. Even though the observation site does not experience the southwest monsoon, the
winds carrying them affect the atmospheric column over the observation site, resulting in the downward vertical wind,
which further suppresses the exhaled radon. Thus, during the southwest monsoon period (June to October), radon
movement in the air column is restricted in a thin layer, giving smaller diurnal variation amplitudes in total γ-ray flux.
6. CONCLUSION
The present work carried out to understand the diurnal variation of the γ-ray spectrum and its causes recorded by
NaI (Tl) scintillation detector at Tirunelveli, South India concludes the following:
•There exists a distinct and significant ( >10%) diurnal pattern in the total number of γ-ray counts.
•The counts start decreasing after sunrise and show gradual recovery after sunset.
•The diurnal pattern is present only in the energies related to the terrestrial background radioactivity (500 keV
– 2.7 MeV).
•The study demonstrates that the pattern is associated with the radioactivity of isotopes of radon and their
daughter radionuclides.
•The diurnal variation is predominantly due to 222Rn and its daughter radionuclides. Thoron gas i.e., 220 Rn,
though in small proportion, also contributes to the observed diurnal pattern via its daughter product 208Tl.
•The amplitude of the diurnal variation is found to have seasonal dependence, with the lowest amplitude during
June-October (∼2%), and the highest in April-May months (∼14%).
•Good correlation of the amplitude of diurnal variation with the range of temperature variation (CC = 0.609) as
well with the relative humidity (CC = 0.797) is observed.
•The pattern arises due to changing contribution from radioactivity of radon progenies due to generation of the
inversion layer. The distribution of concentration of radon and thoron gases in an air column changes with the
vertical mixing and ABL.
•γ-rays with energies >2.7 MeV resulting from the cosmic ray interactions with the atmospheric particles do not
exhibit any detectable diurnal pattern.
DATA AVAILABILITY
The γ-ray data from NaI(Tl) detector used for the current study is available on request through proper per-
missions from the Director, Indian Institute of Geomagnetism (IIG). The vertical wind dataset is available at
[https://cds.climate.copernicus.eu/cdsapp#!/dataset/reanalysis-era5-pressure-levels] under public domain.
14 Datar et al.
ACKNOWLEDGEMENTS
The experimental set up at Tirunelveli is operated by the Indian Institute of Geomagnetism. This work is supported
by the Department of Science and Technology, Government of India. Authors are grateful to corresponding authorities
of IMD and ECMWF for providing respective datasets. We acknowledge that neither the European Commission nor
ECMWF is responsible for any use that may be made of the Copernicus Information or Data this publication contains.
AB acknowledges the support by the NASA Living With a Star Jack Eddy Postdoctoral Fellowship Program,
administered by UCARs Cooperative Programs for the Advancement of Earth System Science (CPAESS).
AUTHOR CONTRIBUTIONS STATEMENT
All authors participated in designing the experiment. GD carried out the data analysis. GD, GV, AB, and AR
contributed to the analysis of results, discussion, and writing the manuscript.
ADDITIONAL INFORMATION
Competing interests: The authors declare no competing interests.
Causes of the Diurnal Variation observed in γ-ray Spectrum using NaI (Tl) Detector 15
REFERENCES
Arora, B. R., Kumar, A., Walia, V., et al. 2017, Journal of
Asian Earth Sciences, 149, 49
Baldoncini, M., Alb´eri, M., Bottardi, C., et al. 2017,
Atmospheric Environment, 170, 259
Banjanac, R., Udoviˇci´c, V., Dragi´c, A., et al. 2012, in Paper
presented at the First East European Radon
Symposium–FERAS
Briggs, M. S., Xiong, S., Connaughton, V., et al. 2013,
Journal of Geophysical Research: Space Physics, 118,
3805
Chin, F., & Standil, S. 1968a, Planetary and Space Science,
16, 15
—. 1968b, Planetary and Space Science, 16, 7
Desideri, D., Roselli, C., Feduzi, L., & Meli, M. A. 2006,
Journal of radioanalytical and nuclear chemistry, 270, 523
El-Hussein, A., Mohamemed, A., El-Hady, M. A., et al.
2001, Atmospheric Environment, 35, 4305
Fishman, G. J., Bhat, P., Mallozzi, R., et al. 1994, Science,
264, 1313
Galeriu, D., Melintescu, A., Stochioiu, A., Nicolae, D., &
Balin, I. 2011, Romanian Reports in Physics, 63, 115
Gogolak, C. 1977, Variation of radon and radon daughter
concentrations with height above ground, Tech. rep.
Griffiths, A., Zahorowski, W., Element, A., & Werczynski,
S. 2010, Atmospheric Chemistry and Physics, 10, 8969
Gusev, A. A., Martin, I. M., Alves, M. A., & de Abreu,
A. J. 2015, Modeling Earth Systems and Environment, 1,
18
Horng, M.-C., & Jiang, S.-H. 2003, Radiation
measurements, 37, 603
Hosoda, M., Shimo, M., Sugino, M., Furukawa, M., &
Fukushi, M. 2007, Journal of nuclear science and
technology, 44, 664
Jacobi, W., & Andre, K. 1963, Journal of Geophysical
Research, 68, 3799
Katase, A., Narahara, Y., Ishihara, Y., Tanaka, K., &
Matsuyama, H. 1982, Journal of Nuclear Science and
Technology, 19, 918
Knoll, G. F. 2010, Radiation detection and measurement
(John Wiley & Sons)
Liu, S., McAfee, J., & Cicerone, R. 1984, Journal of
Geophysical Research: Atmospheres, 89, 7291
Manohar, S., Meijer, H., & Herber, M. 2013, Atmospheric
environment, 81, 399
Minato, S. 1983, Journal of Radioanalytical and Nuclear
Chemistry, 78, 199
Minty, B., & Wilford, J. 2004, Exploration Geophysics, 35,
312
on the Effects of Atomic Radiation, U. N. S. C. 2000,
Sources and effects of ionizing radiation: sources, Vol. 1
(United Nations Publications)
Panneerselvam, C., Nair, K., Selvaraj, C., et al. 2007,
Earth, planets and space, 59, 429
Patrignani, C., Agashe, K., Aielli, G., et al. 2016, Chinese
physics C, 40
Raghav, A., Bhaskar, A., Yadav, V., & Bijewar, N. 2015,
Astrophysics and Space Science, 355, 347
Raghav, A., Bhaskar, A., Yadav, V., et al. 2013, Journal of
Geophysical Research: Space Physics, 118, 6426
Schery, S., & Whittlestone, S. 1989, Journal of Geophysical
Research: Atmospheres, 94, 18297
Schubert, M., & Schulz, H. 2002, Health Physics, 83, 91
Szegvary, T., Conen, F., & Ciais, P. 2009, Atmospheric
Environment, 43, 1536
Takeuchi, N., & Katase, A. 1982, Journal of Nuclear
Science and Technology, 19, 393
Tchorz-Trzeciakiewicz, D., & Solecki, A. 2018, Atmospheric
Environment, 174, 54
Vichare, G., Bhaskar, A., Datar, G., et al. 2018, Advances
in Space Research
Victor, N. J., Siingh, D., Singh, R., Singh, R., & Kamra, A.
2019, Journal of Atmospheric and Solar-Terrestrial
Physics, 195, 105118,
doi: https://doi.org/10.1016/j.jastp.2019.105118
Wilkening, M., Clements, W., & Stanley, D. 1972, Radon
222 flux measurements in widely separated regions, Tech.
rep.
Williams, A. G., Zahorowski, W., Chambers, S., et al. 2011,
Journal of the Atmospheric Sciences, 68, 155
16 Datar et al.
APPENDIX
Decay chains of radioactive isotopes of Thorium and Uranium (232Th, 235 U, and 238U)
