**3. Main features of annual and daily regular changes in UV irradiance in cloudless conditions**

Environmental effects of UV irradiance at any site depend strongly on the availability of direct sunshine and on solar elevation during the sunshine episodes. Solar elevation modulates the UV irradiance and its spectral composition in the absence of sunshine. Solar elevation changes regularly from sunrise to sunset, reaching the highest value at noon in real solar time. Noon solar elevation manifests a regular annual cycle, being the smallest at winter solstice and the largest at summer solstice. Usually SZA is used instead of solar elevation for characterization of solar position on the celestial sphere. At the Tartu Observatory site the extreme values of noon SZA are 81.7˚ (solar elevation 8.3˚) and 34.8˚ (solar elevation 55.2˚), respectively. The most influential part of the UVR ground-level spectrum is the UVB range.

In the days around summer solstice, the threshold of the Bentham double monochromator's sensitivity is 294 nm at noon in sunshine conditions. In early mornings and late evenings, the threshold of sensitivity drops to approximately 310 nm. In cloudy weather, the irradiance levels may be somewhat lower and the threshold of sensitivity can be several nanometres larger than in clear conditions. At an SZA above 80˚, the accuracy of spectral irradiance measurements is significantly lower than at higher sun conditions and the influence of column ozone and AOD on its spectral composition is less clear. In Figure 1 the diurnal cycles of threshold UVB wavelength and SZA during a nearly clear midsummer day are illustrated.

**Figure 1.** Diurnal cycles of SZA and shortwave threshold for UVB irradiance in midsummer cloudless conditions

The recorded UVB range in recorded spectrum is much shorter than the UVA range. The width of the first varies during the day from 5 to 20 nm. In the UVA range, it is constant and much larger, 85 nm. Due to the small and varying contribution of the UVB range, the ratio of the received irradiance energy or power also varies across a wide range. To avoid very small numbers, it is better to use the version UVA/UVB for presenting this ratio instead of UVB/ UVA.

**3. Main features of annual and daily regular changes in UV irradiance in**

Environmental effects of UV irradiance at any site depend strongly on the availability of direct sunshine and on solar elevation during the sunshine episodes. Solar elevation modulates the UV irradiance and its spectral composition in the absence of sunshine. Solar elevation changes regularly from sunrise to sunset, reaching the highest value at noon in real solar time. Noon solar elevation manifests a regular annual cycle, being the smallest at winter solstice and the largest at summer solstice. Usually SZA is used instead of solar elevation for characterization of solar position on the celestial sphere. At the Tartu Observatory site the extreme values of noon SZA are 81.7˚ (solar elevation 8.3˚) and 34.8˚ (solar elevation 55.2˚), respectively. The most

In the days around summer solstice, the threshold of the Bentham double monochromator's sensitivity is 294 nm at noon in sunshine conditions. In early mornings and late evenings, the threshold of sensitivity drops to approximately 310 nm. In cloudy weather, the irradiance levels may be somewhat lower and the threshold of sensitivity can be several nanometres larger than in clear conditions. At an SZA above 80˚, the accuracy of spectral irradiance measurements is significantly lower than at higher sun conditions and the influence of column ozone and AOD on its spectral composition is less clear. In Figure 1 the diurnal cycles of threshold UVB wavelength and SZA during a nearly clear midsummer day are illustrated.

**Figure 1.** Diurnal cycles of SZA and shortwave threshold for UVB irradiance in midsummer cloudless conditions

influential part of the UVR ground-level spectrum is the UVB range.

**cloudless conditions**

124 Solar Radiation Applications

It is commonly known that the spectral composition of the UVB irradiance depends on stratospheric column ozone. At a large SZA the optical path of incoming solar rays is long, and diffuse radiation dominates. In those conditions much of the UVB radiation is absorbed by tropospheric ozone and attenuation reaches longer wavelengths than at high sun conditions. As a result the UVB day is seemingly shorter than the UVA day. In the UVA spectral range the level of irradiance is higher, and radiation is not absorbed by ozone. Beside the atmospheric column ozone another modulating factor of UVR is AOD. Its influence also tends be larger in the UVB range.

In Figure 2 the diurnal cycles of irradiance at some UVB and UVA wavelengths are presented. The selected day, 12 July 2010, was almost clear. Only a very small amount of *cumulus (Cu)* clouds between 13 and 17 in local time were met. Total ozone was 299 DU and the AOD at the UV wavelength 340 nm was relatively large, at 0.428. In cases of good atmospheric transpar‐ ency it is around 0.1 and its median and trimean values in 2002-2012 have been close to 0.2. The conventional mean is not a relevant characteristic of the AOD due to a strong asymmetry of distribution.

**Figure 2.** Diurnal cycles of spectral irradiance at selected UVB (left) and UVA (right) wavelengths on an almost clear summer day 12 July 2010

Finding of relevant data is restricted by the deficit of cloudless days. The ratio UVA/UVB in the received daily spectral energy accumulates from the recorded spectra. The contribution from the spectra recorded at smaller SZA is larger in it. Immediately after sunrise and before sunset the UVB irradiance is strongly suppressed and the value of the UVA/UVB ratio often reaches 500 to 600 at SZA around 87˚. At an SZA above that value the UVA/UVB ratio manifests strong instability due to the low reliability of recording UVB irradiance and is not presented in the figures. With the decreasing SZA the relative contribution of UVB irradiance increases.

In Figure 3 the average dependence of UVA/UVB ratio on SZA is presented in almost cloudless weather conditions for the abovementioned full day. In Figure 4 the same is presented for another almost clear day on 13 April when noon SZA reaches only 62˚ and column ozone, 376 DU, is close to its normal spring level. The value of AOD at 340 nm is once again quite large, at 0.493.

**Figure 3.** Diurnal cycles of UVA/UVB irradiance ratio in almost clear summer day 12 July 2010

The smallest of UVA/UVB values are met around the daily smallest SZA. Around summer solstice the range of diurnal changes in the UVA/UVB ratio reaches 15 times, but in some cases is only eight to ten times. With the decrease of SZA to 70˚ the ratio UVA/UVB drops to about 100 and after reaching SZA 50˚ to about 50. In the Northern European summer conditions most of the UV radiation is received during six hours around noon.

An example of relative contribution from these six hours in the full day dose is presented in Figure 5 at wavelengths 300 to 400 nm for clear day in July.

One can see that at wavelengths around 300 nm the contribution from these six hours is 85 to 90% of the daily total, and decreases with increasing wavelength due to decrease of ozone absorption. Around wavelength 330 nm the noon six hours contribution reaches a 60% level and remains at an almost constant level.

Instrumentation and Measurement of Ground-Level Ultraviolet Irradiance and Spectral Composition in Estonia http://dx.doi.org/10.5772/59615 127

**Figure 4.** Diurnal cycles of UVA/UVB irradiance ratio in almost clear spring day 13 April 2010

Finding of relevant data is restricted by the deficit of cloudless days. The ratio UVA/UVB in the received daily spectral energy accumulates from the recorded spectra. The contribution from the spectra recorded at smaller SZA is larger in it. Immediately after sunrise and before sunset the UVB irradiance is strongly suppressed and the value of the UVA/UVB ratio often reaches 500 to 600 at SZA around 87˚. At an SZA above that value the UVA/UVB ratio manifests strong instability due to the low reliability of recording UVB irradiance and is not presented in the figures. With the decreasing SZA the relative contribution of UVB irradiance increases.

In Figure 3 the average dependence of UVA/UVB ratio on SZA is presented in almost cloudless weather conditions for the abovementioned full day. In Figure 4 the same is presented for another almost clear day on 13 April when noon SZA reaches only 62˚ and column ozone, 376 DU, is close to its normal spring level. The value of AOD at 340 nm is once again quite large,

**Figure 3.** Diurnal cycles of UVA/UVB irradiance ratio in almost clear summer day 12 July 2010

of the UV radiation is received during six hours around noon.

Figure 5 at wavelengths 300 to 400 nm for clear day in July.

and remains at an almost constant level.

The smallest of UVA/UVB values are met around the daily smallest SZA. Around summer solstice the range of diurnal changes in the UVA/UVB ratio reaches 15 times, but in some cases is only eight to ten times. With the decrease of SZA to 70˚ the ratio UVA/UVB drops to about 100 and after reaching SZA 50˚ to about 50. In the Northern European summer conditions most

An example of relative contribution from these six hours in the full day dose is presented in

One can see that at wavelengths around 300 nm the contribution from these six hours is 85 to 90% of the daily total, and decreases with increasing wavelength due to decrease of ozone absorption. Around wavelength 330 nm the noon six hours contribution reaches a 60% level

at 0.493.

126 Solar Radiation Applications

**Figure 5.** Dependence of the ratio of UV irradiances from six noon hours and full day on wavelength

On cloudy days the ratio manifests variations due to attenuation and enhancement of irradi‐ ance by clouds. The contribution from six hours around noon is a useful indicator only from May to August when the outdoor activities of the population take place frequently in sunshine.

During sunshine episodes the ground-level UVR and its spectral composition are modulated by atmospheric column ozone and aerosols. Atmospheric column ozone manifests an annual cycle with the maximum around 390 Dobson Units (DU) in March-April and a minimum of around 270-280 DU in October-November [23]. From the first half of May until the middle of September the column ozone decreases quite linearly and the variations around this linear decrease are mostly moderate, within ± 20 DU. Larger variations were recorded in spring, from February until the middle of May, and also in October-November. Within that period, prolonged episodes reaching a week or even more in length were recorded when column ozone was even more than 50 DU lower of its normal seasonal level.

Aerosol-size distribution is characterized by the fine mode fraction, e.g., how much the submicronic particles contribute to the AOD at 500 nm. Smoke often contributes more than 90% of small particles' influence in AOD, reducing radiation more strongly at shorter wave‐ lengths [28-31]. Smoke was a major reason for a large AOD in years when there were prolonged dry periods in summer. The major season of forest fires in the region is in July-August and that of landscape burnings in late April to early May.

In most of the years, the dryness and frequency of fire outbreak are moderate. In 2002-2012 the AERONET system recorded AOD values above the threefold and twofold median were met at Tartu-Tõravere meteorological station in about 6% and 17%, respectively, out of all 1500 days of the AERONET measurement data. The influence of the landscape fire episode in April-May 2006 on the UV spectrum is described in [32]. In conditions of seasonal normal ozone between 379 and 391 DU three almost clear days with AOD values at 340 nm between 0.979 and 1.299 were found.

**Figure 6.** Comparison of mean daily spectral doses for the group of days manifesting smoke induced large AOD and a clear day with almost equal noon SZA, similar column ozone 384 DU and moderate AOD (0.16 at 340 nm)

The cloud influence is often described by the cloud modification factor (CMF). CMF is defined as a ratio of measured cloudy irradiance to that of normal conditions of clear sky. Here we use a similar aerosol modification factor (AMF) to compare total influence on the UVR. The AMF in daily irradiance totals was found to be 0.84 in UVA and 0.75 in UVB ranges. It is comparable to the CMF of middle-level clouds. The major difference is that the AMF decreases with wavelength decrease in the whole UV range while CMF manifests in the UVA range decrease with the increase of wavelength. Comparison of mean spectral daily doses for the mentioned days manifesting smoke-induced large AOD and a clear day with almost equal noon SZA, similar column ozone 384 DU and moderate AOD (0.16 at 340 nm) is presented in Figure 6.
