Ground-based Measurements of Solar Ultraviolet Radiation
In Tibet: Preliminary Results
Pu Bu Ci Ren (1) Fred Sigernes (2) Yngvar Gjessing (1)
(1) Geophysical Institute, University of Bergen, Allegaten 70, N-5007, Bergen, Norway
(2) University Courses on Svalabard, P.O. Box 156, N-9170, Longyearbyen, Norway
Abstract: Measurements were carried out on Tibetan Plateau (Lhasa, 3648 meters above
sea level) with a multi-channel moderate bandwidth filter instrument in the period from June
24 to October 30, 1996. The analysis of preliminary data shows that hourly mean CIE weighted
biologically effective UV dose rate at local noon can reach to as much as 390.5 mW/m2 with a
average of 222.5 mW/m2; Due to the combined effect of cloud cover variation with ozone variation,
the increasing trend of this UV dose rate exists during this period of measurements. The relation
between the ratio of irradiance close to noon at 340nm to that at 305nm and ozone column amount shows
that the variation of ozone amount may also causes the variation of UV radiation. In addition, UV dose rate
is more sensitive to the variation of solar zenith angle than total global radiation.
The importance of solar ultraviolet radiation and its impact on the human beings have been extensively discussed
during last few decades [ Scotto et al, 1988; Blumthaler and Ambach, 1990], and both broadband and spectral measurements
of UV radiation have been performed at several sites around the world [Blumthaler et al, 1991; Frederick et al, 1993;
Bais et al, 1993]. However, neither long-term broadband nor long-term spectral measurements of UV radiation in Tibetan
Plateau have been reported, where UV radiation is extremely important for the population and plants in this region
because of the combined effect of high altitude with low latitude.
Tibetan Plateau is the highest plateau in the world with average height about 4,000 metres above sea level.
Due to the strong solar radiation caused by shallow atmosphere and the special nature of surface, Tibetan Plateau
becomes a strong heating source to the atmosphere, which can modify the atmospheric circulations on local, regional
and global scales [Yeh et al, 1979]. Especially, in the meridional monsoon circulation, which becomes much stronger
over Tibetan Plateau compared to other corresponding longitudes, such as the Rocky Mountains or the Andes Mountains.
In addition, the works of Zhou et al.[ 1994] and Han Zou,  showed that there is a region of low ozone existing
over Tibetan Plateau in summer. They also found decreasing trend of ozone over Tibetan Plateau from TOMS 1978-1991
year-round data. Therefore, study of UV radiation on Tibetan Plateau has essential importance in mainly three aspects:
1) to obtain its spatial extents and temporal changes in high altitude area; 2) to understand the biological effect
of UV radiation; 3) combined with radiative transfer model, to determine the total ozone concentration and trends and
compare it with the results from satellite measurements.
In this paper, we report on the first ground-based solar UV radiation measurements over Tibetan Plateau.
The characteristics of resulting CIE weighted biologically effective UV dose rate over Tibetan Plateau and the relation
between the trend of UV dose rate and the trend of ozone column amount measured by Earth Probe TOMS are also discussed.
A multi-channel filter instrument (NILUV) was installed on the roof of Institute of Tibetan Plateau Atmospheric
and Environment Science Research (ITPAESR) in Lhasa (29°40'N, 90°08'E; 3648 m above sea level). This instrument was
produced by Norwegian Institute of Air Research ( NILU), it consists basically a Teflon diffuser, interference filters
and photo-diodes as detectors. It has three channels, the UV-B channel is centred at 305nm, while the UV-A channels
are centred at 320nm and 340nm. The filters have 10nm full width at half maximum(FWHM). One minute interval
was adopted in this measurements.
The effective UV dose rate is determined by a linear combination of the irradiances measured by 305nm channel
and 340nm channel. The coefficients used in this linear combination are calculated with the radiative transfer model.
The conclusion from the tests of sensitivity of these coefficients to ozone profile and surface albedo shows that
only a single set of coefficients is needed to determine CIE-weighted UV dose rate [Dahlback,1996].
The above instrumental technique was constructed and tested by Dahlback , and the comparison with high
resolution spectral radiometer which was done during a one week period with variable cloudiness in San Diego, USA,
showed that for solar zenith angle (SZA) <80°, the relative difference was 1.4˝3.2% under all sky conditions, and only 0.6˝1.5% under near clear sky condition; [Dahlback, 1996].
The hourly mean total global radiation used in this paper is observed simultaneously at Tibet Meteorological
Observatory, which is about 100 metres from our UV measurements site. In addition, The ozone column amount data is
obtained from the products of Earth Probe TOMS, which is available since July 25, 1996, and the average value of
four pixels around location of UV radiation measurements, Lhasa, is taken as Ozone column amount over Lhasa.
Results and Discussion
Tibetan Plateau is located in middle latitude area (29°N 40°N), and its high altitude makes solar radiation
penetrate through atmosphere in a shorter pass from the sun to the ground. The highest solar irradiation is received
on ground level ( 1178.0 w/m2, maximum of hourly mean at local noon in this data set). According to the analysis of
measurements from June 24 to November 30, 1996, the hourly mean of UV dose rate at local noon (06:00 UT) can reach to
as much as 390.5 (mW/m2), and the average value is 225.4 (mW/m2).
Fig. 1. Time series of biological effective UV dose rates from Lhasa, Tibet in the period from June 24 to
November 30, 1996 at solar zenith angle (SZA) 60° and ozone column amounts (DU) measured by Earth Probe TOMS from
July 25 to November 30. The dashed line represents the fitted line for ozone column amount(DU). The solid line and
dotted line are fitted lines for morning and afternoon UV dose rate at solar zenith angle 60° respectively .
AM and PM represent morning measurements and afternoon measurements at the same solar zenith angle.
The intensity of solar UV radiation on ground level is strongly influenced by the concentration of
ozone in atmosphere, equally important parameters are cloud cover and solar zenith angle. For eliminating the
influence of solar zenith angle, the ten minutes average of UV dose rate are calculated at solar zenith angle
10°˝1°, 30°˝1°, 50°˝1° and 60°˝1° respectively. Figure .1. shows the time series of UV dose rate measured at solar
zenith angle 60°and ozone column amount measured by Earth Probe TOMS. The decreasing trend of ozone column amount
clearly exists during this measurement. Using linear regression analysis, a daily ozone trend was found to be
-0.12˝0.09 DU/day with 95% confidence during period from July 25 to November 30. The same analysis was used to
the corresponding UV dose rate time series at solar zenith angle 60°, and a daily trend of UV dose rate was found to be
0.15˝0.16 mW/m2/day( or 0.11˝0.18 mW/m2/day) for morning value ( or afternoon value). However, It is difficult to
deduce that the corresponding increasing trend of UV dose rate is due to the decreasing of ozone amount, since the data
set used here includes data under different cloud cover conditions, which cause different modulating effects to UV dose
Fig.2. Time series of the ratio of irradiance at 340nm to that at 305nm with solar zenith angle
60° and ozone column amount (DU) in Lhasa from June 24 to November 30.
However, as scattering by clouds is essentially independent of wavelength, and the absorption of ozone
decreases with the increase of wavelength almost in orders of magnitude[Dan, 1991], much of the variability associated
with cloud cover can be cancelled in the ratio of irradiance at 340nm to that at 305nm. In order to assess the effect
of ozone variation to UV radiation, the ratio of irradiance at 340nm to that at 305nm at certain solar zenith angle,
which was measured during this experiment, is used in this paper(Fig.2). It shows that the corresponds of this ratio
variation at solar zenith angle 60° to ozone column amount variation improve with time, and the correlation analysis
between this ratio and ozone column also shows that linear correlation coefficient in the period from October 14 to
November 30 (R=0.62,N=47) increases to 2 times of that in the period from July 25 to September 10 (R=0.31,N=47).
The reason can be explained by the fact that the time when UV irradiances was measured is more close to the ozone
observing time by satellite at the end of this time series. The ozone column amount is observed almost constantly by
Earth Probe TOMS at about 11:20 am (LT) , but in the period of this UV measurement from July 25 to November 30, the
time when solar zenith angle is equal to 60°, changes from 7:30 to 10:30 (LT). Therefore, according to the availability
of data and the closeness of UV irradiance measurement time to ozone observation time, the ratio at solar zenith angle
30° is chosen to analyse the effect of ozone variation on UV radiation instead during the time period from July 25 to
September 23, and the correlation coefficient between this ratio and ozone amount increases from 0.31 to 0.49.
The scatter plot in Figure.3 shows that the positive correlation exists between this ratio and ozone column amount
during these two periods, even if the standard deviation of ozone amount during these two periods are only 4.8 and 8.0
respectively. It indicates that the variation of ozone amount also has effect on the variation of UV radiation at
Earth's surface during the period of measurements in Lhasa.
Fig.3. Scatter plot between ratio of irradiance at 340nm to that at 305nm and ozone column amount (DU)
measured from Earth Probe TOMS. (a). Measurements from October.14 to November.30 at solar zenith
angle is 60° . (b). Measurements from July 25 to September 22 at solar zenith
angle is 30°.
Based on the hourly mean total global radiation measured at the same time, the hourly mean percentages of UV
dose rate in the total global radiation are also calculated. Figure 4. shows the relation between hourly mean
percentages and hourly mean solar zenith angles. It is clear that the percentage of UV dose rate in total global
radiation generally decrease with the increasing of solar zenith angle, that is due to the long optical depth in the
ozone layer at high solar zenith angle, the UV dose rate is more sensitive to the change of solar zenith angle than
total global radiation. In addition, except the high values which appeared when solar zenith angle becomes high than
85° can be associated with the ratio between two small values, the high values also happen at other solar zenith angle,
and a general increasing trend of these high values with solar zenith angle exists. It probably results from the
difference between effects of cloudiness on UV dose rate and that on total global radiation, and It can be explained
by the fact that the reducing effect of cloudiness is considerably greater on total global radiation than on UV dose
rate, while it is also systematically greater at high solar zenith angles compared with lower solar zenith angles
[Blumber.M, 1994]. On the other hand, some part of deviation from the decreasing trend with solar zenith angle ,
specially these lower values, can be attributed to the variation of ozone amount at same solar zenith angle during
this time period, but Its effects are expected to be smaller compared to that of the cloudiness.
Fig.4. The change of the ratio of hourly mean UV dose rate to hourly mean total global radiation with
hourly mean solar zenith angle from June 24 to October 30 in Lhasa.
The amount of UV radiation reaching to ground is mainly determined by solar zenith angle, ozone amount and
cloud cover. Figure.5 shows the scatter plot between morning UV dose rates and afternoon UV dose rates at solar zenith
angle 10°˝1°, 30°˝1° and 60°˝1°. The morning UV dose rate and afternoon UV dose rate are defined as UV dose rate
measured at the same solar zenith angle in morning and at afternoon. The gradient and correlation coefficient of
the regression line is 0.82˝0.27 and 0.78 respectively, indicating slightly higher UV dose rates exist in the morning
compared with that at afternoon, which can mainly be attributed to the observed more frequently and heavily cloud cover
at afternoon compared to prenoon during summer monsoon season.
Fig.5. Scatter plot between measured morning and afternoon UV dose rate with same solar zenith
angle during time period from June 24 to November 30 in Lhasa. Fitted line are also shown(solid line).
The principal results obtained by these first ground-based measurements of UV radiation in Lhasa during
time period from June 24 to November 30 may be summarized as follows:
Acknowledgement: This work was financially supported by Norwegian Department for Development (NORAD).
The authors are grateful to our colleagues in Institute of Tibetan Plateau Atmospheric and Environmental Science
Research (ITPAESR) in Lhasa for helping us install the instrument, specially Mr. Cuduo who is in charge of down loading
data and maintaining this instrument.
- 1. The UV dose rate was found extremely high during the period of measurements. Maximum value at local
noon was close to 390 mW/m2 with a total average of 225 mW/m2.
- 2. Due to combined effects of cloud cover variation and ozone variation, The biologically effective
UV dose rate at solar zenith angle 60 °shows about 0.13˝0.17 mW/m2 /day daily trend during this period of measurements.
- 3. The good relation exists between the ratio of irradiance close to noon at 340nm and that at 305nm and
ozone column amount measured from Earth Probe TOMS, indicating that the variation of ozone amount may also causes
the variation of UV radiation at surface during this period of measurements.
- 4. The UV dose rate in the morning is slightly more intense compared to that at afternoon with same
solar zenith angle.
- 5. Further study about the relation between the spectral irradiance and ozone amount and long time
spectral measurement of UV radiation are needed, to verify the seasonal and long-term trend of UV radiation and
the ozone amount over Tibetan Plateau.
Ambach,W., M. Blumthaler, G.Wendler, A comparison of ultraviolet radiation measured at an arctic and an alpine
site, Solar Energy(2), 121-126,1991.
Bais, A.F., C. S. Zerefos, C. Meleti, I. C. Ziomas and K.Tourpali, Spectral measurement of solar UVB
radiation and its relations to total Ozone, So2 and clouds, J. Geophys. Res., 98, 5199-5204,1993.
Blumthaler,M., and W. Ambach, Indication of increasing solar ultraviolet-B radiation flux in Alpine regions,
Science, 248, 206-208, 1990.
Blumthaler, M., W. Ambach and M. Salzgeber, Effects of cloudiness on global and diffuse UV irradiance in
a high- mountain area, Theor. Appl. CLimatol.,50, 23-30, 1991
Dahlback. A., Measurements of biological effective UV-doses, total ozone abundance and cloud effects with
multi-channel moderate bandwidth filter instruments, Appl.optic. Vol. 35, No. 33, 1996.
Dan Lubin, The ultraviolet radiation environment of the Antarctic Peninsular: The role of ozone
and cloud cover, Journal Of applied Meteorology, Vol.30, 1991.
Frederick, J. E., P. F. Soulen, S. B. Diaz, I. Smolskaia and C. R. Booth, Solar ultraviolet irradiance observed
from southern Argentina: September 1990 to March 1991, J. Geophys.Res.,98, 8891-8897, 1993.
Han Zou, Seasonal variation and trends of TOMS ozone over Tibet, Geophys. Res. Let., 23, 1029-1032, 1996.
Mckinlay,A.F. and B.L.Diffey, A reference action spectrum for ultra-violet radiation: Risks and Regulation, Elsevier, Amsterdam, 83-87, 1987.
Scotto,J., G. Cotton, F.Urbach, D. Berger, and T.Fears, Biologically effective ultraviolet radiation: surface measurements in the United States, 1974 to 1985, Science, 239,762-764,1988.
Yeh, T. C and Y. X. Gao, Meteorology of Tibetan Plateau ( in Chinese ), lpp., Scientific Press, Beijing, China, 1979.
Zhou, Xiuji and Chao Luo, Ozone valley over Tibetan Plateau, Acta Meteorologica Sinca, 8(4), 505-506, 1994.