Comparative analysis of cloud effects on ultraviolet-B and broadband solar radiation: Dependence on cloud amount and solar zenith angle

Atmospheric Research 168 (2016) 149–157

Contents lists available at ScienceDirect

Atmospheric Research journal homepage: www.elsevier.com/locate/atmos

Comparative analysis of cloud effects on ultraviolet-B and broadband solar radiation: Dependence on cloud amount and solar zenith angle M. El-Nouby Adam a,b,⁎, Emad A. Ahmed b a b

Quality Assurance Unit (ALI), King Saud University, Riyadh 11491, Saudi Arabia Faculty of Science (Physics Department), South Valley University, Qena 83523, Egypt

a r t i c l e

i n f o

Article history: Received 13 December 2014 Received in revised form 8 March 2015 Accepted 11 September 2015 Available online 25 September 2015 Editor: J.L. Sanchez Keywords: Egypt Ratio of UVB to broadband solar radiation Cloud modification factor Solar zenith angle Clouds

a b s t r a c t The role played by clouds in atmospheric radiative transfer is fundamental. However, studies of their influences on solar radiation at Qena, Egypt, are lacking. This paper discusses the effect of clouds on the ratio of ultraviolet-B (UVB, 295–2800 nm) to broadband solar radiation (G, 280–320 nm) under all cloud conditions. To achieve this aim, the dependence of cloud modification factors for both UVB and G (CMFUVB and CMFG respectively) on cloud amount (CA) was compared for each category. From 2000 to 2009, the dataset used included hourly values for both UVB and G under all cloud conditions. The results indicated that the average UVB/G ratio for all cloud conditions was 0.25 ± 0.10%. This average represents 76% of the corresponding value under cloudless sky conditions (0.33 ± 0.11%). Accordingly, it indicates that clouds do not transmit UVB and G equally. However, on average, the percentage decrease in CMFUVB from its corresponding value of CMFG was 20 ± 5%. Further analysis for the hourly values of both CMFUVB and CMFG was employed. These hourly values are distinguished by three different behaviors: CMFUVB ≌ CMFG (9% of all cases), CMFUVB N CMFG (11% of all cases), or CMFUVB b CMFG (80% of all cases). The average UVB/G values for each group were 0.27 ± 0.09%, 0.33 ± 0.12%, and 0.23 ± 0.09% respectively. Based on the results of earlier studies, an attempt was made to explain these outcomes. In a more thorough study of CMFUVB and CMFG, the dependence of these variables on SZA was investigated. This analysis has shown that both CMFUVB and CMFG decrease as the SZA increases. For each group, the sensitivities of CMFUVB to SZA were −0.27 ± 0.03, −0.32 ± 0.06, and −0.18 ± 0.03 (respectively); moreover, the sensitivities of CMFG to SZA were −0.27 ± 0.03, −0.65 ± 0.05, and −0.18 ± 0.02 (respectively). © 2015 Elsevier B.V. All rights reserved.

1. Introduction Numerous literatures provided a detailed description of the electromagnetic energy emitted by the Sun and reviewed the factors controlling the intensity of this energy (e.g. Iqbal, 1983). In addition, the beneficial and damaging effects of ultraviolet-B (UVB) on humans, ecosystems, animals, plants, and materials were mentioned (e.g. McKinlay and Diffey, 1987; UNEP, 1989; Kripke, 1992; Webb, 1998; Juzeniene et al., 2011). Adam (2014) studied the role of the atmosphere as a filter, absorbing and scattering portions of the solar spectrum and shielding the biosphere from damaging doses of UVB. In this research, the variability of the ratio of UVB (280–320 nm) to broadband solar radiation (G, 295–2800 nm) at the Earth's surface (UVB/G) under all sky conditions was evaluated at Qena, Egypt. The results showed that UVB/G varied from 0.03% to 0.73%, with an average value of 0.27%. This average represents only 17% of the corresponding value outside the atmosphere (1.56% (Adam, 2011b)). This means that atmospheric attenuation notably reduces this ratio at the Earth's surface compared to its value at the top of the atmosphere. On ⁎ Corresponding author. E-mail addresses: [email protected] (M. El-Nouby Adam), [email protected] (E.A. Ahmed).

http://dx.doi.org/10.1016/j.atmosres.2015.09.009 0169-8095/© 2015 Elsevier B.V. All rights reserved.

the other hand, the influences of solar zenith angle (SZA), ozone, aerosols, and water vapor on this ratio were assessed. UVB/G was found to decrease with increasing SZA. Investigation of the relationship between UVB/G and ozone showed that this ratio decreased with increasing ozone. For a restricted change in ozone and almost constant turbidity level, UVB/G became higher with increasing water vapor. This is due to the impact of water vapor on G (Adam, 2014). Although clouds play an important role in atmospheric radiative transfer (Tsay and Stamnes, 1992; Bais et al., 1993), studies of their influences on the UVB/G ratio at Qena are lacking (Adam, 1995, 2011a; El Shazly et al., 1997; El-Noubi, 2006; Adam and El Shazly, 2007). Therefore, the present study has attempted to investigate the effect of clouds on the UVB/G ratio. Understanding the influence of clouds on UVB/G is a constructive contribution to science, for which the effects of clouds on both UVB and G must be illustrated and compared. Clouds reflect, absorb, and transmit incoming solar radiation, modifying in this way the amount and spectral quality of solar radiation reaching the Earth's surface. Cloud particles are responsible for the scattering and absorption of solar radiation. However, the influence of clouds on radiation depends on both the macrophysical cloud features and the microphysical structure of water droplets and ice crystals. The spatial distribution (level of cloudiness) and vertical extension

150

M. El-Nouby Adam, E.A. Ahmed / Atmospheric Research 168 (2016) 149–157

(top–bottom distance) of clouds are basic macrophysical parameters. Cloud microphysical parameters (number of droplets or crystals per volume and droplet size distribution) determine the volume extinction coefficient. Integration of this coefficient over the cloud thickness results in the cloud optical depth, which provides a type of linkage between macro- and microphysical cloud characteristics (Krzyscin et al., 2003). Calbó et al. (2005) noted that clouds dominate any other atmospheric variable as a source of solar radiation variability and that their variability usually masks the effects of other variables. Clouds may affect the absorption of solar radiation by other atmospheric constituents. For example, clouds are the major factor limiting the detectability of ozoneinduced trends in UV radiation (den Outer et al., 2005; Glandorf et al., 2005). Accordingly, clouds enhance the diffusion component of any solar radiative flux, while providing an effective reduction of the direct component if the cloud is in front of the Sun. The sum of both components, called the global component, undergoes a subsequent change in its partitioning into direct and diffuse components and generally an overall reduction over time (Foyo-Moreno et al., 2000). Moreover, UVB is strongly affected by the presence of clouds. The effect of clouds on UVB levels can vary from small enhancements to almost total reduction depending on cloud cover geometry and the relative position of the Sun (Bodeker and McKenzie, 1996; Piedehierro et al., 2014). In this work, fluctuations in global UVB and G radiant fluxes for all cloud conditions have been described. In addition, the dependence of both hourly UVB and G on cloud amount was illustrated (Section 3.1) by examination of the collected data. Before discussing the effect of clouds on the UVB/G ratio, the dependence of both global UVB and G radiant fluxes on cloud amount was compared (Section 3.2). Furthermore, the dependence of this effect on SZA was discussed (Section 3.3). Finally, the net effect of clouds on both UVB and G was clarified by comparing the UVB/G ratios under cloudless and completely cloudy conditions. The main objective of this work was to compare the effect of cloud amount on UVB and G and to discuss the effect of clouds on the UVB/G ratio at Qena. In addition, the dependence of cloud influences on UVB and G on SZA was illustrated.

response exceeds 10% beyond 60° SZAs (Bigelow et al., 1998). In addition, the Precision Spectral Pyranometer (PSP No. 16317IS, Eppley Laboratory Inc.) was used to measure total irradiance from 295 to 2800 nm. The sensitivity of the sensor is 9 μV/Wm− 2 of total irradiance, and its response time is 1 s. The temperature dependence of the PSP is approximately ±1% over the ambient temperature range of − 20 °C to + 40 °C. The cosine response of the instrument is ± 1% from normalization 0°–70° SZA and ±3% 70°–80° SZA. The COMBILOG Datalogger (No. 1020, TH., and Friedrichs & Co.) was used to record hourly integral irradiance (irradiation integrated in one hour) values of UVB and G. Additional details of the instrument specifications and data-gathering procedures at SVU-MRS are available in previous studies such as those of El-Noubi (2006), Ahmed (2008), Adam (2012), El-Noubi (2012), Ahmed and Adam (2013), and Adam (2013; 2014). In addition, the cloud amount (in octas) was used in this work to describe the state of cloudiness of the atmosphere. It is visually observed at the end of each hour in the study station. According to these visually observed (in octas), the datasets of hourly UVB and G under different cloud conditions (5684 cases) were classified to examine the collected data (Section 3.1). 2.2. Definitions 2.2.1. UVB and G under cloudless sky conditions Hourly values of UVB and G under cloudless sky conditions (UVB0 and G0 respectively) were estimated. The empirical approach developed by Adam and El Shazly (2007) was used to estimate UVB values under cloudless sky conditions for each hour throughout the study period. The following equation was used to estimate UVB0 (with coefficient of determination R2 = 0.99 and standard error of the estimate SEE = 0.08): UVB0 ¼ ð0:017  0:001Þm−1:901

ð1Þ

2. Data and methods

where the relative optical air mass (m) is given as a function of Z (the solar zenith angle in degrees) as expressed in the following equation (Kasten, 1966):

2.1. Measurements

m¼

In this study, hourly values of UVB, G, and cloud amount (CA) at Qena, Egypt (26.2°N, 32.75°E, 96 m above mean sea level) from 2000 to 2009 were used. In previous work by the authors (Adam, 2014), these measurements were used to assess the diurnal variability of the UVB/G ratio under all sky conditions. In addition, the dataset was analyzed under cloudless sky conditions to quantify the effects of SZA, ozone, water vapor, and aerosols on the UVB/G ratio, and datasets under different cloud conditions were separated out. The present study used these separated datasets to investigate the effect of clouds on the UVB/G ratio under all cloud conditions. Adam (2014; 2015) provided descriptions of the study site, the instrumentation, and the measurements. In this earlier study, measurements were collected at the South Valley University Metrological Research Station (SVU-MRS), located at Qena within the subtropical region (Robaa, 2008). The Egyptian Meteorological Authority (EMA) established SVU-MRS in cooperation with South Valley University. Scientific advice and instrument calibrations for SVU-MRS are provided by the EMA. However, the instruments are also calibrated yearly against the World Radiometric Reference (WRR) maintained at Davos, Switzerland (WRC, 1985, 1995). A Model UVB-1 ultraviolet pyranometer (No. 960842, Yankee Environmental Systems Inc.) was used to measure the total irradiance from 280 to 320 nm. The sensitivity of this instrument is 1.97 μV/Wm−2 over the total UVB irradiance, and its response time is approximately 0.1 s. It works over an ambient temperature range of − 40 °C to + 40 °C. The cosine response of this instrument is ±5% for 0°–60° SZA. The departure from the ideal cosine

1 cos Z þ 0:15ð93:885 − Z Þ−1:253

ð2Þ

In addition, the method of Foyo-Moreno et al. (1999) was used to generate clear-sky estimates of G for each hour (G0). The empirical approach developed by Adam (2011a), was used to estimate G values under cloudless sky conditions for each hour throughout the study period. The following equation was established (with coefficient of determination R2 = 0.96 and standard error of the estimate SEE = 0.07): G0 ¼ ð4:002  0:049Þm−0:790:01

ð3Þ

2.2.2. Cloud modification factors for UVB and G Cloud modification factors for UVB and G (CMFUVB and CMFG respectively) were determined in this study. Adam (2011a) and Adam and El Shazly (2007) stated that cloud effects on UVB and G could be described using the cloud modification factor (the ratio of the irradiation occurring in the presence of clouds to the irradiation that would occur under the same atmospheric and surface conditions in the absence of clouds). This is a simple and operational approach, and several earlier studies used similar methods to analyze cloud effects on G, UV, and UVB. For example, effects on G were studied by (Kasten and Czeplak, 1980; Blumthaler et al., 1994; Davies, 1995; Josefsson and Landelius, 2000; Adam, 2011a), effects on UV and the photosynthetically active spectral range were studied by (Alados et al., 2000; Foyo-Moreno et al., 2001, 2003), and effects on UVB were studied by (El-Noubi, 2006; Adam and El Shazly, 2007). The relation between solar global

M. El-Nouby Adam, E.A. Ahmed / Atmospheric Research 168 (2016) 149–157

collected data. Table 1 summarizes the primary statistical parameters (average “ave.”, standard deviation “std.”, standard error of the mean “SEM” and number of observations in each CA category “n”) for hourly G, G0, and CMFG. These primary statistical parameters for UVB, UVB0, and CMFUVB, are shown in Table 2, as well as the UVB/G ratio (%). In addition, the average of each variable for all cloud categories is included in both tables. Moreover, the average of hourly values of both G and UVB for each CA category is graphically shown in Fig. 1 (a and b), as well as CMFG and CMFUVB (Fig. 1 (c and d)). From Tables 1 and 2 and Fig. 1 (a and b), the average hourly UVB and G for each cloud category decreased with increasing CA. The results showed that the average of all cases (under all cloud conditions) of hourly G was 1.9 ± 0.9 MJ m−2, whereas the corresponding value for cloudless sky conditions (G0) was 2.8 ± 0.7 MJ m−2. The averages for hourly UVB and UVB0 were 0.005 ± 0.004 MJ m−2 and 0.009 ± 0.005 MJ m− 2 respectively. Accordingly, the average UVB/G ratio under all cloud conditions was 0.25 ± 0.10%. This average represents 76% of the corresponding value for cloudless sky conditions (0.33 ± 0.11%). Accordingly, it can be concluded that clouds in the atmosphere decrease the UVB/G ratio by 0.08%. This suggests that the contribution of clouds reduces UVB more than G. In addition, it is clear that clouds do not transmit UVB and G equally at Qena. The earlier studies such as; Nann and Riordan (1991), Blumthaler et al. (1994), Mims and Frederick, 1994; Kylling et al. (1997), and Schwander et al. (2002), found that higher cloud effects for different spectral ranges of the solar spectrum occurred at shorter wavelengths. This wavelength dependence was due to photons that were backscattered into the atmosphere at the top of the cloud and then experienced Rayleigh scattering, which was larger at shorter wavelengths (Seckmeyer et al., 1996). Fig. 1 (a and b) includes bars representing the standard deviation, the size of which indicates the wide spread of hourly UVB and G for a given CA category. This scatter is partly associated with the variety of cloud types included in each category of this rather simple classification. The position of clouds with respect to the Sun, which was not taken into account here, can also be responsible for this scatter. In addition, the effects of other atmospheric components contributed in the wide spread of hourly UVB and G for a given CA category. Previous studies have indicated that the atmospheric compounds with the strongest influence on UVB radiation are ozone, aerosols, and clouds (Alados-Arboledas et al., 2003; Antón et al., 2012). Moreover, other gasses such as water vapor control irradiance in the solar infrared bands reaching the Earth's surface (Marín et al., 2007; Jacovides et al., 2009; Bilbao et al., 2011; de Miguel et al., 2011). In previous papers (Adam, 2010, 2013, 2014), the effects of ozone, water vapor, and aerosols on hourly UVB and G were examined in detail at Qena, as well as the effect of SZA. These effects are not the subject of this study. The following sections describe an attempt to compare the effect of cloud cover categories on UVB and G (Section 3.2), as well as the dependence of these effects on SZA (Section 3.3). Finally, a general conclusion of the net effect of cloud amount on the UVB/G ratio is presented.

Table 1 Statistical parameters (average “ave.”, standard deviation “std.”, standard error of the mean “SEM”, and number of observations “n”) of G0, G, CMFG for each cloud cover category under all cloud conditions at Qena from 2000 to 2009. CA (octas)

1 2 3 4 5 6 7 8 All cases

G0 (MJ m−2)

G (MJ m−2)

CMFG

n

ave.

std.

SEM

ave.

std.

SEM

ave.

std.

SEM

2.70 2.73 2.67 2.80 2.68 2.89 2.94 2.87 2.78

0.67 0.69 0.70 0.72 0.70 0.69 0.66 0.69 0.75

0.03 0.02 0.03 0.02 0.03 0.03 0.04 0.07 0.01

2.08 2.11 2.03 2.08 1.82 1.94 1.85 1.24 1.89

0.84 0.86 0.88 0.89 0.82 0.83 0.86 0.69 0.93

0.04 0.02 0.03 0.03 0.04 0.03 0.05 0.07 0.01

0.74 0.74 0.73 0.71 0.65 0.65 0.60 0.42 0.65

0.17 0.17 0.19 0.19 0.18 0.20 0.20 0.18 0.18

0.01 0.00 0.01 0.01 0.01 0.01 0.01 0.02 0.00

466 1896 750 940 456 748 316 112 5684

radiation under all sky conditions (G) and cloudless sky radiation (G0) can be expressed as (Foyo-Moreno et al., 2001) G ¼ CMFG G0 ;

ð4Þ

where CMFG is a function of cloud features (amount, type, and height). After Adam and El Shazly (2007), this function can be adapted to UVB solar radiation, as expressed in the following equation: UVB ¼ CMFUVB UVB0 ;

ð5Þ

2.2.3. Clearness index (Kt) Hourly value of clearness index (Kt) was defined as the ratio of G on a horizontal surface at ground level to its corresponding value outside the Earth's atmosphere (El-Nashar, 1991; Foyo-Moreno et al., 1998; Kudish et al., 2005). Therefore, Kt can be expressed as follows: Kt ¼

G ; Gext

151

ð6Þ

where the Gext are extraterrestrial hourly values in spectral bands of G. 3. Results and discussion 3.1. Variability of global UVB and G radiant fluxes for all cloud conditions Adam et al. (2013) and Adam (2014) provided a summary of descriptive statistics for collected measurements of both hourly UVB and G at Qena from 2000 to 2009. However, these earlier studies described the general behavior of hourly UVB and G under all sky conditions. In addition, they developed a statistical relationship between hourly UVB and G. In this section, the dataset used includes all cases of hourly UVB and G under all cloud conditions from 2000 to 2009. A simple classification of hourly UVB and G was carried out according to the cloud cover categories described above (in octas) to examine the

Table 2 Statistical parameters (average “ave.”, standard deviation “std.”, standard error of the mean “SEM”, and number of observations “n”) of UVB0, UVB, CMFUVB and UVB/G ratio for each cloud cover category under all cloud conditions at Qena from 2000 to 2009. CA (octas)

1 2 3 4 5 6 7 8 All cases

UVB0 (MJ m−2)

UVB (MJ m−2)

CMFUVB

UVB/G (%)

n

ave.

std.

SEM

ave.

std.

SEM

ave.

std.

SEM

ave.

std.

SEM

0.0086 0.0088 0.0085 0.0094 0.0085 0.0100 0.0103 0.0099 0.0091

0.0045 0.0048 0.0048 0.0051 0.0048 0.0051 0.0049 0.0049 0.0051

0.0002 0.0001 0.0002 0.0002 0.0002 0.0002 0.0003 0.0005 0.0001

0.0056 0.0055 0.0049 0.0056 0.0047 0.0051 0.0051 0.0028 0.0051

0.0034 0.0036 0.0034 0.0036 0.0030 0.0032 0.0033 0.0023 0.0041

0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0002 0.0002 0.0001

0.64 0.60 0.56 0.57 0.56 0.50 0.48 0.29 0.53

0.19 0.17 0.16 0.16 0.16 0.15 0.17 0.16 0.17

0.01 0.00 0.01 0.01 0.01 0.01 0.01 0.02 0.00

0.256 0.245 0.229 0.256 0.250 0.262 0.262 0.226 0.251

0.092 0.096 0.097 0.102 0.084 0.113 0.090 0.088 0.104

0.004 0.002 0.004 0.003 0.004 0.004 0.005 0.009 0.001

466 1896 750 940 456 748 316 112 5684

M. El-Nouby Adam, E.A. Ahmed / Atmospheric Research 168 (2016) 149–157

CA (octas) 0

G (MJ m-2)

3.0

1

2

3

4

CA (octas) 5

6

7

8 8

7

6

5

4

3

2

1

(a)

0 0.010

(b)

0.008

2.0

0.005 1.0

UVB (MJ m-2)

152

0.003 0.0

0.000 1.000

(d)

(c)

0.8

0.800

0.6

0.600

0.4

0.400

0.2

0.200

0.0

CMFUVB

CMFG

1.0

0.000 0

1

2

3

4

5

6

7

8

8

7

6

CA (octas)

5

4

3

2

1

0

CA (octas)

Fig. 1. Dependence of hourly G (a) and hourly UVB (b) on cloud amount (CA), as well as CMFG (c) and CMFUVB (d), at Qena under all cloud conditions from 2000 to 2009.

3.2. Comparison between cloud effects on UVB and G Fig. 2 displays the fitting functions obtained by linear regression of the dependence of average hourly CMFUVB and CMFG (Fig. 1 (c and d)) on CA category. Both CMFUVB and CMFG have a fitting equation of the following form: CM F x ¼ a3 CA3 þ a2 CA2 þ a1 CA þ a0 ðthe sub‐index x represents UVB or GÞ:

ð7Þ

50

1.00 (CMFG - CMFUVB ) / CMFG

CMFUVB

CMFG

40

0.80

30

0.60

20

0.40

10

0.20

0

1

2

3

4

5

6

7

8

CMFG& CMFUVB

(CMFG - CMFUVB ) / CMFG (%)

For CMFUVB, the values of a0, a1, a2, and a3 in Eq. (7) were 0.76 ± 0.05, − 0.15 ± 0.05, 0.039 ± 0.010, and − 0.004 ± 0.001 (with correlation coefficient R = 0.98). However, for CMFG, these values were 0.8 ± 0.1, − 0.07 ± 0.05, 0.02 ± 0.01, and − 0.002 ± 0.001 (with correlation coefficient R = 0.95). These fitted lines reveal the efficacy of the partly covered category. As mentioned above, similar nonlinear behavior has been reported by other authors for the whole solar spectrum (Kasten and Czeplak, 1980; Davies, 1995; Adam (2011a), for photosynthetically active radiation (Alados et al., 2000), for the thermal infrared emissions of the atmosphere (Alados-Arboledas et al., 1995), for UVB erythemal irradiance (Kuchinke and Nunez, 1999; Foyo-Moreno et al., 2001), and for UVB solar radiation (Adam and El Shazly, 2007).

Fig. 2 shows a comparison between CMFUVB and CMFG. This figure reflects the average attenuation of G and UVB solar radiation relative to their values under cloudless sky conditions as a function of CA. It was found that clouds can reduce G and UVB radiation significantly. The effect of CA on UVB radiation is greater than its effect on global solar radiation. However, the average hourly CMFUVB values were lower than the corresponding values of CMFG . In addition, Fig. 2 includes the variation in the percentage increase of CMF G over CMFUVB ((CMFG–CMFUVB)/CMFG (%)) for each cloud amount category. On average, this percentage increase was equal to 20 ± 5%. The maximum value was equal to 29.52% for CA = 8 octas, whereas the minimum value was equal to 13.41% for CA = 1 octa. Furthermore, the relation between both dimensionless factors, hourly CMFUVB and hourly CMFG , has been discussed for all cloud conditions throughout the period of this study (Fig. 3). However, the effect of clouds on both UVB and G is obvious in the hourly values of CMFUVB and CMFG . Fig. 3 shows an evident departure from the 1:1 line, which represents coincident factors, i.e., the fact that clouds do not transmit UVB and G equally. Observing the scatter of point in this figure, three cases were distinguished in the analyses: points with CMFUVB equal or close to CMFG (group I); points with CMFUVB greater than CMF G (group II); and points with CMF UVB less than CMFG (group III).

0.00

CA (octas) Fig. 2. Comparison of the average values of CMFUVB and CMFG for each cloud amount category at Qena under all cloud conditions from 2000 to 2009. In addition, the variation of the percentage ((CMFG–CMFUVB)/CMFG (%)) for each cloud amount category is included.

M. El-Nouby Adam, E.A. Ahmed / Atmospheric Research 168 (2016) 149–157

153

Fig. 3. Relationship between dimensionless factors (hourly CMFG and hourly CMFUVB) for each group (I, II, and III) under all cloud conditions at Qena from 2000 to 2009.

The first group represented the similar effect of clouds on UVB and G. The average value of Kt for all cases in group I was calculated. However, average Kt was determined as 0.41 ± 0.11 (with standard error of the mean SEM = 0.004). This may mean that most cases in this group represented a partly cloudy condition (Escobedo et al., 2009). As mentioned above, Bodeker and McKenzie (1996) stated that the effect of clouds on the solar radiation spectrum depends not only on the micro- and macrophysical characteristics of clouds, but also on cloud cover geometry and the relative position of the Sun. They explained that partly cloudy skies can either enhance or reduce UVB. For these cases (group I), transmission of UVB and G through clouds was equal; however, the average UVB/G ratio for this group (0.27 ± 0.09%, with standard error of the mean SEM = 0.003) was equal to the UVB/G ratio under all sky conditions (0.27%, after Adam, 2014). Frederick et al. (1993) supported this result; however, they found that G attenuation by clouds was the same as UVB attenuation. The relation between CMFUVB and CMFG can be represented as a first-order linear relation of the following form: CMFUVB ≅ CM FG

  with coefficient of determination R2 ¼ 0:98 :

ð8Þ

In the second group, the effect of clouds on UVB and G was different. The second group contained cases where UVB through clouds was Table 3 Frequency distribution (%) of SZA values for the various SZA classes under all cloud conditions at Qena from 2000 to 2009. SZA classes (rad.)

0.06–0.10 0.11–0.20 0.21–0.30 0.31–0.40 0.41–0.50 0.51–0.60 0.61–0.70 0.71–0.80 0.81–0.90 0.91–1.00 1.01–1.10 1.11–1.20 1.21–1.30 1.31–1.40 1.41–1.50

Frequency (%) Group I

Group II

Group III

CMFUVB ≌ CMFG

CMFUVB N CMFG

CMFUVB b CMFG

0.0 0.0 0.0 1.7 0.0 3.9 5.6 6.0 8.2 10.3 10.8 12.1 13.4 13.8 14.2

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 11.7 11.7 11.7 16.3 16.3 16.3 16.3

0.5 1.3 2.2 2.9 3.6 5.0 5.9 6.4 6.6 7.2 7.3 11.9 12.1 12.6 15.2

enhanced relative to its value under cloudless conditions. Bodeker and McKenzie (1996) explained that UVB is strongly affected by the presence of clouds. In addition, the effect of clouds on UVB can vary from slight enhancement to almost total reduction. The average value of Kt for all cases in group II was estimated as 0.35 ± 0.12 (with standard error of the mean SEM = 0.004). According to Escobedo et al. (2009), this may mean that most cases in this group represented cloudy sky conditions. According to earlier studies, the enhancement of the solar radiation spectrum under particular sky conditions is wavelengthdependent. Calbó et al. (2005) supported this explanation; they stated that, although the cloud modification factor showed no significant wavelength dependence, more recent analysis has shown some evidence to the contrary. They noted that Seckmeyer et al. (1996) had carried out a study of transmissivity through stratified cloud, with measurements above the top and below the bottom of the cloud. They found that transmittance of UVB (60%) was greater than that of UVA (45%). In addition, they reproduced these results using a radiative-transfer model and explained that the wavelength dependence was due to photons that were backscattered into the atmosphere at the top of the cloud and then experienced Rayleigh scattering. This scattering was larger at shorter wavelengths. Furthermore, Nann and Riordan (1991), Blumthaler et al. (1994), Mims and Frederick, 1994; Kylling et al. (1997), and Schwander et al. (2002) found a relative increase in radiation at shorter wavelengths and a higher cloud modification factor at shorter wavelengths. In these cases (group II), clouds transmit UVB more than G; however, the average UVB/G ratio for this group (0.33 ± 0.12, with standard error of the mean SEM = 0.004) was greater than the corresponding ratio under cloudless sky conditions. Blumthaler and Ambach (1988) supported these results. They found that G was attenuated more effectively than UVB by a factor of 1.2 in the Swiss Alps. The relation between CMFUVB and CMFG was described in first-order linear form as: CM F UVB ¼ ð1:44  0:01ÞCM F G

  with coefficient of determination R 2 ¼ 0:30 :

ð9Þ The third group included points for which the value of CMFG was more than on CMFUVB. The average value of Kt for all cases in group III was estimated as 0.52 ± 0.09 (with standard error of the mean SEM = 0.001). According to Escobedo et al. (2009), this may mean that most cases in this group represented partly cloudy sky conditions. This result may have been due to clouds affecting UVB more than G. However, Mayer et al. (1998) illustrated that in the presence of thick clouds, coupling of cloud scattering and molecular or particulate

154

M. El-Nouby Adam, E.A. Ahmed / Atmospheric Research 168 (2016) 149–157

Fig. 4. Frequency distribution (%) of SZA for the various SZA classes under all cloud conditions at Qena from 2000 to 2009.

CM F UVB ¼ ð0:69  0:01ÞCM F G

  with coefficient of determination R2 ¼ 0:40 :

ð10Þ Because of the strong dependence of UVB and G on SZA, the dependence of both CMFUVB and CMFG on SZA will be investigated (Section 3.3), and it is constructive to separate this relationship for each group (I, II, and III). 3.3. Dependence of CMFUVB and CMFG on SZA To see the effect of SZA on CMFUVB and CMFG more clearly, further analysis was done. An analysis for each separated group (I, II, and III, see Section 3.2) was performed. Hourly CMFUVB and CMFG values were classified according to the corresponding SZA “in radius”

CMFUVB

1.0

(a)

(0.06 ≤ SZA ≤ 1.50). For each SZA class, the range of SZA was equal to 0.09, except for the first class, which included SZA values from 0.06 to 0.10. Then the average hourly values of CMFUVB and CMFG for each class were estimated. Before analyzing the dependence of CMFUVB and CMFG on SZA, a clear picture of the frequencies of SZA values for the different SZA classes was obtained. These frequencies of SZA values (%) in these classes are reported in Table 3 for groups I, II, and III. Moreover, the distribution of the frequencies of SZA values over the different SZA classes is presented graphically in Fig. 4. From Table 3 and Fig. 4, it is evident that the frequency (%) for group I, II, and III has a peak at approximately 14.2% (SZA class 1.41–1.50), 16.3% (SZA classes 1.11–1.20, 1.21–1.30, 1.31–1.40, and 1.41–1.50), and 15.2% (SZA class 1.41–1.50) respectively. For groups I and III, the cases were distributed over a wide range of SZA (0.31–1.50 and 0.06–1.50 respectively). For group II, the cases were distributed over a range of SZA N 0.80. For the three groups (I, II, and III), the relationship between CMFUVB, as well as CMFG, and the average values of SZA for each SZA class is shown in Figs. 5 (a and b), 6 and 7, respectively. Generally, these figures contain long bars representing the standard deviation. These indicate a wide spread of hourly CMFUVB and CMFG for a given SZA class. As mentioned above (Section 3.1), this scatter is partly associated with the variety of cloud types included under each SZA class in this rather simple classification. In addition, the analysis did not consider unique values of SZA, but rather an average value for each SZA class. Moreover, the

1.0

(b)

0.8

0.8

0.6

0.6

0.4

0.4 CMF UVB = -0.27±0.03 SZA + 0.84±0.03 R² = 0.91

0.2

CMF G = -0.27±0.03 SZA + 0.85±0.03 R² = 0.90

0.2

0.0

0.0 0.0

0.5

1.0

SZA (rad.)

1.5

0.0

0.5

1.0

SZA (rad.)

Fig. 5. Dependence of CMFUVB (a) and CMFG (b) on SZA for group I (the bars represent the standard deviation).

1.5

CMFG

absorption can result in a strong enhancement of absorption, which may appear as a wavelength dependence of cloud attenuation. As mentioned above, this may have been due to the effect of ozone on UVB radiation. In these cases (group III), clouds transmitted UVB less than G; however, the average UVB/G ratio for this group (0.23 ± 0.09%, with standard error of the mean SEM = 0.001) was less than the corresponding ratio under cloudless sky conditions. The relation between CMFUVB and CMFG can be described in first-order linear form as:

M. El-Nouby Adam, E.A. Ahmed / Atmospheric Research 168 (2016) 149–157

155

position of the clouds with respect to the Sun, which was not taken into account here, could also be responsible for this scatter. For the three groups of datasets (I, II, and III), the fitted linear equations between both variables (CMFUVB and CMFG) and SZA (in radius) have the following form (Eq. (11)): CMFx ¼ ‐a SZA þ b

ðthe sub‐indexxrepresents UVB or GÞ:

ð11Þ

The fitting parameters and the coefficient of determination, R2, for Eq. (11) are summarized in Table 4, as well as the number of cases for each group. For the three groups of datasets, CMFUVB and CMFG decreased with increasing SZA. The following discussion addresses the sensitivity of CMFUVB to SZA, ΔCMFUVB/ΔSZA, and CMFG to SZA, ΔCMFG/ΔSZA, for each group: • For group I (Fig. 5 (a and b)), the results can be summarized by saying that the value of the correlation coefficient, R, between CMFUVB and SZA is approximately equal to the value of R between CMFG and SZA (≌0.95). Moreover, ΔCMFUVB/ΔSZA is approximately equal to ΔCMFG /ΔSZA (≌0.27 ± 0.03). This occurs because the values of hourly CMFG ≌ CMFUVB, i.e., the clouds transmit UVB and G equally. The results showed that the coefficients of determination of the fitted linear equations for both cases were ≌0.90. This implies that 90% of the variance in CMF UVB and CMFG was explained by SZA. Accordingly, 10% of the variance in CMFUVB and CMFG may be due to other factors such as the type of clouds and the position of clouds relative to the Sun. Despite the importance of the effects of cloud types, they were not considered in this analysis. Rather, the dataset used was divided into three groups (I, II, and III) and then classified according to SZA range. • For group II (Fig. 6), as mentioned above, the cases in this dataset had SZA N 0.80. Moreover, 48.9% of these cases had SZA N 1.10. According to earlier studies, when SZA increases, both UVB and G cross a greater portion of the atmosphere and therefore are affected by more clouds. Román et al. (2012) stated that the relationship of both UVB and G to SZA is due to the geometrical Sun position. In addition, as reported in Adam (2014), the modification of the atmosphere through which the solar radiation must pass is a function of the radiation path length through the atmosphere and the amount of each attenuator (such as clouds) along that path length. Bilbao et al. (2014) reports that the analytic relation between the irradiances with SZA cosine and, as expected, UV fits as a power function and solar global fits as a linear behavior. The reason for this behavior is caused by the ratio between diffuse and direct components. Similar fits were obtained by Román et al. (2012). The results for this dataset showed that the CMFUVB values were greater than the corresponding CMFG values for each SZA class and that the value of R between CMFUVB and SZA was lower than its value between CMFG and SZA (0.71 and 0.96 respectively). In addition, ΔCMFUVB/ΔSZA was approximately half of ΔCMFG/ΔSZA (0.32 ± 0.06 and 0.65 ± 0.05, respectively). Although SZA described 92% of the variance in CMFG, only 51% of the variance in CMFUVB was explained by SZA. As mentioned in Section 3.2, this may have been due to UVB enhancement rather than the value under cloudless sky conditions. • For group III (Fig. 7), as mentioned above, the dataset included the cases where CMFUVB was less than CMFG. According to Fig. 4 and Table 3, these cases were distributed over a wide range of SZA (0.06–1.50). When SZA varies over a wide range, the interaction between both G and UVB and clouds is also changed. The results reflect that the value of R between CMFUVB and SZA was lower than its value between CMFG and SZA (0.89 and 0.92 respectively). The coefficients of determination of the fitted linear equations for both cases were 0.77 and 0.85. This implies that 77% of the variance in CMFUVB was explained by SZA, but that 85% of the variance in CMFG was explained by SZA. Moreover, ΔCMFUVB/ΔSZA (0.18 ± 0.03) was approximately equal to ΔCMFG/ΔSZA (0.18 ± 0.02).

Fig. 6. Dependence of CMFUVB (the bars represent the positive values of standard deviation) and CMFG (the bars represent the negative values of standard deviation) on SZA for group (II).

4. Conclusions Over a 10-year period, for all cloud conditions, the overall findings of this study can be briefly summarized as follows: • The analysis of hourly values for both CMFUVB and CMFG illustrated that the effect of CA on UVB radiation was (I) equal to its effect on global solar radiation (9% of all cases), (II) less than its effect on global solar radiation (11% of all cases), or (III) greater than its effect on global solar radiation (80% of all cases). On average, the effect of CA on UVB radiation is greater than its effect on global solar radiation. The average values of CMFG for each cloud category were greater than their corresponding values of CMFUVB (by 20 ± 5% on average). • The net effect of clouds on both UVB and G was clarified by comparing the values of the UVB/G ratio under cloudless sky conditions and its corresponding value under all cloud conditions. On average, the UVB/G ratio at the Earth's surface under all cloud conditions was 0.25 ± 0.10%, with standard error of the mean SEM = 0.001. This average represents 76% of the corresponding value for cloudless sky conditions (0.33 ± 0.11%, with standard error of the mean SEM = 0.001). Accordingly, clouds in the atmosphere decrease this ratio (UVB/G) by 0.08%, which suggests that the contribution of clouds reduces UVB more than G. In addition, it is clear that clouds do not transmit UVB and G equally at Qena.

Acknowledgments The associate editor of the journal, Dr. L. Alados-Arboledas, and the anonymous reviewers are highly acknowledged for their constructive

Fig. 7. Dependence of CMFUVB (the bars represent the positive values of standard deviation) and CMFG (the bars represent the negative values of standard deviation) on SZA for group (III).

156

M. El-Nouby Adam, E.A. Ahmed / Atmospheric Research 168 (2016) 149–157

Table 4 The fitting parameters and the coefficient of determination, R2 of Eq. (11), as well as the number of cases for each group. CMFUVB

Group I Group II Group III

CMFUVB ≌ CMFG CMFUVB N CMFG CMFUVB b CMFG

CMFG

n

−a

b

R2

−a

b

R2

0.27 ± 0.03 0.32 ± 0.06 0.18 ± 0.03

0.84 ± 0.03 1.04 ± 0.05 0.67 ± 0.02

0.91 0.51 0.77

0.27 ± 0.03 0.65 ± 0.06 0.18 ± 0.02

0.85 ± 0.03 1.18 ± 0.05 0.85 ± 0.01

0.90 0.92 0.85

comments and suggestions. The authors would like to thank the Program Research Center at the Arabic Linguistics Institute, Deanship of Scientific Research, Vice Rectorate for Graduate Studies and Scientific Research, King Saud University, Riyadh, Kingdom of Saudi Arabia, for funding and supporting this research.

References Adam, M.E.-N., 1995. Studies of Solar Radiation Components Related to Aerosols in the Atmosphere at Qena, Upper Egypt. South Valley University, Qena, Egypt (M.Sc. thesis). Adam, M.E.-N., 2010. Effect of stratospheric ozone on UVB solar radiation reaching the Earth's surface at Qena. Egypt. Atmos. Poll. Res. 1, 155–160. Adam, M.E.-N., 2011a. Effect of macrophysical parameters of clouds on broadband solar radiation (295–2800 nm) at a subtropical location. Atmos. Ocean. Sci. Let. 4, 180–185. Adam, M.E.-N., 2011b. Effect of the atmosphere on UVB radiation reaching the Earth's surface: dependence on solar zenith angle. Atmos. Ocean. Sci. Let. 4, 139–145. Adam, M.E.-N., 2012. Using sunshine duration to estimate UVB solar radiation under sky cover conditions at Qena (Egypt). Can. J. Comp. Math. Nat. Sci. Eng. and Med. 3, 7–24. Adam, M.E.-N., 2013. Sensitivity analysis of aerosols' effect on UVB transmission to solar zenith angle at subtropical location (Qena, Egypt). Atmos. Environ. 71, 311–318. Adam, M.E.-N., 2014. Atmospheric modulations of the ratio of UVB to broadband solar radiation: effect of ozone, water vapor and aerosols at Qena. Egypt. Int. J. Climatol. 34, 2477–2488. Adam, M.E.-N., 2015. Determination of daily total ultraviolet-B in a subtropical region (Upper Egypt): an empirical approach. Atmos. Res. 153, 1–9. Adam, M.E.-N., El Shazly, M., 2007. Attenuation of UVB radiation in the atmosphere: cloud effects at Qena (Egypt). Atmos. Environ. 41, 4856–4864. Adam, M.E.-N., Kassem, Kh.O., Ahmed, E.A., 2013. Statistical relationship between UVB (280–320 nm) and broadband solar radiation (295–2800 nm) at a subtropical location (Qena, Egypt). Atmos. Ocean. Sci. Lett. 6, 173–178. Ahmed, Z.M., 2008. Empirical Models for Calculating Some Solar Radiation Components at Qena/Upper Egypt Using Sunshine Period Measurements. South Valley University, Qena, Egypt (M.Sc. thesis). Ahmed, E.A.A., Adam, M.E.-N., 2013. Estimate of global solar radiation by using artificial neural network in Qena, Upper Egypt. J. Clean En. Tech. 1, 148–150. Alados, I., Olmo, F.J., Foyo-Moreno, I., Alados-Arboledas, L., 2000. Estimation of photosynthetically active radiation under cloudy conditions. Agri. For. Meteor. 102, 39–50. Alados-Arboledas, L., Vida, J., Olmo, F.J., 1995. The estimation of thermal atmospheric radiation under cloudy conditions. Int. J. Climatol. 15, 107–116. Alados-Arboledas, L., Alados, I., Foyo-Moreno, I., Olmo, F.J., Alcántara, A., 2003. The influence of clouds on surface UV erythemal irradiance. Atmos. Res. 66, 273–290. Antón, M., Alados-Arboledas, L., Guerrero-Rascado, J.L., Costa, M.J., Chiu, J.C., Olmo, F.J., 2012. Experimental and modeled UV erythemal irradiance under overcast conditions: the role of cloud optical depth. Atmos. Chem. Phys. 12, 11723–11732. Bais, A.F., Zerefos, C.S., Meletie, C., Ziomas, I.C., Tourpali, K., 1993. Spectral measurements of solar UV-B radiation and its relations to total ozone, SO2, and clouds. J. Geophys. Res. 98, 5199–5204. Bigelow, D.S., Slusser, J.R., Beaubien, A.F., Gibson, J.H., 1998. The USDA Ultraviolet radiation monitoring program. Bull. Am. Meteorol. Soc. 79, 601–615. http://dx.doi. org/10.1175/1520-0477. Bilbao, J., Mateos, D., de Miguel, A., 2011. Analysis and cloudiness influence on UV total irradiation. Int. J. Climatol. 31, 451–460. Bilbao, J., Román, R., Yousif, C., Mateos, D., De Miguel, A., 2014. Total ozone column, water vapour and aerosols effects on erythemal and global solar irradiance in Marsaxlokk. Malta. Atmos. Environ. 99, 508–518. Blumthaler, M., Ambach, W., 1988. Human solar ultraviolet radiance: exposure in high mountains. Atmos. Environ. 22, 749–753. Blumthaler, M., Ambach, W., Salzgeberger, M., 1994. Effects of cloudiness on global and diffuse UV irradiance in a high-mountain area. Theor. Appl. Climatol. 50, 23–30. Bodeker, G.E., McKenzie, R.L., 1996. An algorithm for inferring surface UV irradiance including cloud effects. J. Appl. Mete. 35, 1860–1877. Calbó, J., Pagès, D., González, J.A., 2005. Empirical studies of cloud effects on UV radiation: a review. Rev. Geophys. 43, 1–28. Davies, J.A., 1995. Comparison of modeled and observed global irradiance. J. Appl. Meteorol. 35, 192–201. de Miguel, A., Roman, R., Bilbao, J., Mateos, D., 2011. Evolution of erythemal and total shortwave solar radiation in Valladolid, Spain: effects of atmospheric factors. Atmos. Terr. Phys. 73, 578–586.

504 606 4574

den Outer, P.N., Slaper, H., Tax, R.B., 2005. UV radiation in the Netherlands: assessing long-term variability and trends in relation to ozone and clouds. J. Geophys. Res. 110 (D02203), 2005. http://dx.doi.org/10.1029/-2004JD004824. El Shazly, S.M., Abdelmageed, A.M., Adam, M.E.-N., 1997. Solar radiation characteristics at Qena, Egypt. Időjárás, Quart. J. Hung. Meteorol. Serv. 101, 215–231. El-Nashar, A.M., 1991. Solar radiation characteristics in Abu Dhabi. Sol. Energy 47, 49–55. El-Noubi, E.F.E., 2006. Surface Ultraviolet Radiation Measurements in Qena, Egypt M.Sc. thesis, Faculty of Science, South Valley University, Qena, Egypt (150 pp.). El-Noubi, E.F.E., 2012. Distribution of UV Index in Some Upper Regions Ph.D., Faculty of Science, South Valley University, Qena, Egypt (249 pp.). Escobedo, J.F., Gomes, E.N., Oliveira, A.P., Soares, J., 2009. Modeling hourly and daily fractions of UV, PAR, and NIR to global solar radiation under various sky conditions at Botucatu, Brazil. Appl. Energy 86, 299–309. Foyo-Moreno, I., Vida, J., Alados-Arboledas, L., 1998. Ground-based ultraviolet (290–385 nm) and broadband solar radiation measurements in southeastern Spain. Int. J. Climatol. 18, 1389–1400. Foyo-Moreno, I., Vida, J., Alados-Arboledas, L., 1999. A simple all-weather model to estimate ultraviolet solar radiation (290–385 nm). J. Appl. Meteorol. 38, 1020–1026. Foyo-Moreno, I., Vida, J., Olmo, F.J., Alados-Arboledas, L., 2000. Estimating solar ultraviolet irradiance (290–385 nm) by means of the spectral parametric models SPCTRAL2 and SMARTS2. Ann. Geophys. 18, 11382–11389. Foyo-Moreno, I., Vida, J., Alados-Arboledas, L., 2001. On the use of a cloud modification factor for solar (290–385 nm) spectral range. Theor. App. Meteorol. 38, 1020–1026. Foyo-Moreno, I., Alados, I., Olmo, F.J., Alados-Arboledas, L., 2003. The influence of cloudiness on UV global irradiance (295–380 nm). Agric. For. Meteorol. 120, 101–111. Frederick, J.E., Koop, E.K., Alberts, A.D., Weatherhead, E.C., 1993. Empirical studies of tropospheric transmission in ultraviolet: broadband measurements. J. Appl. Meteorol. 32, 1883–1892. Glandorf, M., Arola, A., Bais, A., Seckmeyer, G., 2005. Possibilities to detect trends in spectral UV irradiance. Theor. Appl. Climatol. 81 (1-2), 33–44. Iqbal, M., 1983. An Introduction to Solar Radiation. Academic Press, Toronto. Jacovides, C.P., Tymvios, F.S., Asimakopoulos, D.N., Kaltsounides, N.A., Theoharatos, G.A., Tsitouri, H., 2009. Solar global UVB (280–315 nm) and UVA (315–380 nm) radiant fluxes and their relationships with broadband global radiant flux at an eastern Mediterranean site. Agric. For. Meteorol. 149, 1188–1200. Josefsson, W., Landelius, T., 2000. Effect of clouds on UV irradiance as estimated from cloud amount, cloud type, precipitation, global radiation, and sunshine duration. J. Geophys. Res. 105, 4927–4935. Juzeniene, A., Brekke, P., Dahlback, A., Andersson-Engels, S., Reichrath, J., Moan, K., Holick, M., Grant, W., Moan, J., 2011. Solar radiation and human health. Rep. Prog. Phys. 74, 66701–66757. Kasten, F., 1966. A new table and approximate formula for relative optical air mass. Arch. Meteorol. Geophys. Bioklimatol. B. 14, 206–223. Kasten, F., Czeplak, G., 1980. Solar and terrestrial radiation dependent on the amount and type of cloud. Sol. Energy 24, 177–189. Kripke, M., 1992. Effect on human health. UVB Monitoring Workshop: A Review of the Science and Status of Measuring and Monitoring Programs. AFEAS, Science and Policy Associates Inc., Washington, pp. C58–C63. Krzyscin, J.W., Jaroslawski, J., Sobolewski, P.S., 2003. Effects of clouds on the surface erythemal UV-B irradiance at northern mid-latitudes: estimation from the observations taken at Belsk, Poland (1999–2001). J. Atmos. Sol. Terr. Phys. 65 (4), 457–467. Kuchinke, C., Nunez, M., 1999. Cloud transmission estimates of UV-B erythemal irradiance. Theor. Appl. Climatol. 63, 149–161. Kudish, A.I., Lyubansky, V., Evseev, E.G., 2005. Inter-comparison of the solar UVB, UVA, and global radiation clearness and UV indices for Beer Sheva and Neve Zohar (Dead Sea), Israel. Energy 30, 1623–1641. Kylling, A., Albold, A., Seckmeyer, G., 1997. Transmittance of a cloud is wavelengthdependent in the UV range: physical interpretation. Geophys. Res. Lett. 24 (4), 397–400. Marín, M.J., Esteve, A.R., Tena, F., Utrillas, M.P., Martinez-Lozano, J.A., 2007. UVI dependence on ozone amount and turbidity in Valencia. Opt. Pura Apl. 40, 25–30. Mayer, B., Kylling, A., Madronich, S., Seckmeyer, G., 1998. Enhanced absorption of UV radiation due to multiple scattering in clouds: experimental evidence and theoretical explanation. J. Geophys. Res. 103 (31) (241–231, 254). McKinlay, A.F., Diffey, B.L., 1987. A reference action spectrum for ultraviolet-induced erythematic in human skin. Int. Congr. Ser. 6, 17–22. Mims, F.M., Frederick, J.E., 1994. Cumulus clouds and UVB. Nature 371, 291. Nann, S., Riordan, C., 1991. Solar spectral irradiance under clear and cloudy skies: measurements and semi-empirical model. J. Appl. Meteorol. 30, 447–462. Piedehierro, A.A., Antón, M., Cazorla, A., Alados-Arboledas, L., Olmo, F.J., 2014. Evaluation of enhancement events of total solar irradiance during cloudy conditions at Granada (Southeastern Spain). Atmos. Res. 135–136, 1–7. Robaa, S.M., 2008. On the estimation of UV-B radiation over Egypt. Időjárás, Quart. J. Hung. Meteorol. Serv. 112, 45–60.

M. El-Nouby Adam, E.A. Ahmed / Atmospheric Research 168 (2016) 149–157 Román, R., Mateos, D., de Miguel, A., Bilbao, J., Perez-Burgos, A., Rodrigo, R., Cachorro, V.E., 2012. Atmospheric effects on the ultraviolet erythemal and total shortwave solar radiation in Valladolid, Spain. Optica Pura Apl. 45, 17–20. Schwander, H., Koepke, P., Kaifel, A., Seckmeyer, G., 2002. Modification of spectral UV irradiance by clouds. J. Geophys. Res. 107 (D16), 4296. http://dx.doi.org/10.1029/2001JD001297. Seckmeyer, G., Erb, R., Albold, A., 1996. Transmittance of a cloud is wavelength-dependent in the UV range. Geophys. Res. Lett. 23, 2753–2755. Tsay, S.C., Stamnes, K., 1992. Ultraviolet radiation in the Arctic: the impact of potential ozone depletion and cloud effects. J. Geophys. Res. 97, 7829–7840.

157

UNEP (United Nations Environmental Program), 1989. Environmental Effects Panel Report. United Nations Environmental Program, Nairobi, Kenya. Webb, A.R., 1998. Techniques for measuring solar ultraviolet radiation. In: Auliciems, A. (Ed.), Advances in Bioclimatology: Human Bioclimatology. Springer-Verlag, Berlin, Heidelberg, Germany. WRC (World Radiation Centre), 1985. International Pyrheliometer Comparison (IPC VI). Working Report No. 188, September 25-October 13 (Davos, Switzerland). WRC (World Radiation Centre), 1995. Sixth International Pyrheliometer Comparison (IPC VII). Working Report No. 137, October 1–18 (Davos, Switzerland).