**2. Data and methods**

wavelengths below 300 nm is extremely harmful to DNA. The ratio of irradiances in the UVB and UVA spectral ranges is one of the key factors for estimating its possible harmful influence.

Many important consequences can arise from the indirect effects of high UVB radiation through changes in the chemical composition and shape of the plants, or through changes in the abiotic environment. These indirect effects can include changes in the sensitivity of plants to being attacked by insects and pathogenic elements. The effects of UV radiation and those of other environmental factors at the ecosystem level are poorly known, as well as are those

Plants possess a number of defence systems against environmental stress factors in nature. Among such protection mechanisms is the altered synthesis of antioxidant substances as well as other secondary metabolites. Without repairs, the harmful photoproducts ultimately lead to cell death. To avoid this catastrophic effect, all organisms possess DNA repair systems that are able to recognize and remove UV photoproducts. The ultraviolet (UV) index is a standard vehicle for informing the public about the level of UV radiation reaching the Earth's surface and its potential harmful effects on human health. Although the received energy in the UVB band is only a small fraction of the extra-terrestrial solar radiation, it accounts for 80% of the

In most applications of the UVR data quantification of the received spectral doses a under‐ standing of the mechanisms of influence on the cell, organism and ecosystem levels are needed. The variations of total incoming solar radiation as well as spectral composition, especially in the UVB range, beside geographical factors depend strongly on atmospheric factors, such as clouds, total ozone, aerosols and precipitable water vapour. Spectral energy is necessary to

Despite there being different broadband and narrowband UV sensors in use, spectral UV measurements are still considered the irreplaceable, ultimate reference in a variety of appli‐ cations. Spectral measurements allow the data to be applied to any biological process or chemical photoreaction with a known action spectrum. Weather conditions prescribing the availability and spectral composition of ground-reaching UV irradiance in key phenological phases of ecosystem development manifest significant year-to-year and longer-term periodic changes. These changes are reflected in ecosystem health and productivity. For sustainable agriculture and environmental management both the changes in quality of received irradiance and in ecosystem responses need to be investigated on a quantitative level. The present study financed by a programme of research and development of environmental technology is one

In the present chapter, the major features of systematic changes and the variability in groundlevel UVR at subpolar latitude are considered. The work is based on ground-level UVR spectra

together with the auxiliary information on broadband solar radiation and weather conditions recorded by the Tartu-Tõravere meteorological station of the Estonian Environmental Agency at the same site. The homogeneous datasets of the broadband solar radiation and weather conditions for the site extend back to the beginning of 1955 [12-17]. The total number of

.16'N, 26o

.28'E, 70 m a.s.l.) since 2004

estimate over days, parts of days and over longer time intervals.

recorded at a research institute, Tartu Observatory, (58o

at molecular and organism levels.

120 Solar Radiation Applications

harmful effects of sun exposure.

of those attempts.

#### **2.1. Solar radiation research in Estonia**

Attempts at quantitative measurements of solar radiation have been made in Tartu since 1904 and increased in the 1930s after acquiring a modern Ångström pyrheliometer to quite a high level [18]. The instrument was lost during World War II, and new efforts initiated by Juhan Ross began in the early 1950s. The results of regular studies have been published in several publications and in monographs. General climatic features of the broadband solar radiation in Estonia are presented in the *Handbook of Estonian Solar Radiation Climate* [19].

Major results on interannual and intraseasonal variations of broadband solar radiation in Estonia are presented as a chapter in the previous edition of the InTech book *Solar Radiation* [15]. The longest and most complete dataset on solar radiation in Estonia [20, 21] has been collected at a typical Estonian rural site at the Tartu-Tõravere Meteorological Station (58°.16'N, 26°.28'E, 70 m a.s.l.). The site, as well as that used before 1965, being located closer to Tartu town can be considered typical for Northern Europe. The landscape pattern around the location consists of arable fields, grassland areas and patches of mostly coniferous forest.

Between 1950 and 1965 the station was a part of the present research institute of Tartu Observatory and since 1965 it was operated by the Estonian Meteorological and Hydrological Institute, recently reorganized to part of the Estonian Environmental Agency as a State Meteorological Service. In 1996 the station was included to the system of Baseline Surface Radiation Network (BSRN) stations. Until 1996 the Yanishevski AT-50 actinometers and Savinov-Yanishevski M-115 pyranometers were used and then replaced by the Eppley Labor. Inc. pyrheliometers and Kipp & Zonen pyranometers. The absolute accuracy of the ventilated Kipp & Zonen pyranometers is about ±2% and that of the pyrheliometers ±1%.

In the past, inter-calibration of sensors was regularly performed in Voeikov Main Geophysical Observatory (St. Petersburg, Russia), whereas it is now done in the World Radiation Center (Davos, Switzerland). Regular meteorological observations were performed at the site, including the hourly visual cloud detection, at all three basic levels. The long-term record of traditional broadband solar radiation and weather conditions helps the understanding of the regular changes and variations of UVR, which is the main objective of the present chapter.

#### **2.2. UV broadband, narrowband and spectral measurements**

Studies of atmospheric column ozone and UV radiation in Estonia were initiated by the Tartu Observatory in 1993 to consider the possibilities of studying relationships between the UVR and broadband irradiance characteristics. Regular direct sun column ozone measurements have been carried out between 1994-1999 using an especially suited laboratory spectrometer SDL-1 supplied with a mirror system and applying the Dobson retrieval algorithm [22, 23]. Since 2003, direct sun column ozone measurements were performed using a MICROTOPS-II instrument. Mostly, the satellite data were used for atmospheric column ozone [23]. Since summer 2002, a sun photometer of the NASA AERONET network measuring column aerosol optical depth (AOD) and precipitable water vapour operates there, and the aerosol studies group of the University of Tartu performs ground-level atmospheric aerosol-size distribution measurements. For the years before 2002 only pyrheliometer-measured broadband AOD data are available.

Solar UV radiation measurements with filter instruments were performed at the Tartu-Tõravere Meteorological Station under scientific supervision by research scientists of Tartu Observatory [19]. The erythemally-weighted sensors UV-SET had been in operation since January 1998 [12-14]. A Kipp & Zonen narrowband filter instrument CUVB1 with an effective wavelength 306±0.2 nm and bandwidth 2±0.5 nm operates at Tartu-Tõravere meteorological station since 2002 [16]. Similar UVB instruments were installed at two other meteorological stations Tallinn-Harku (59o 26' N, 24o 45' E) and Pärnu (58o 23' N, 24o 38' E). A Kipp & Zonen broadband UVA sensor as well as a YES broadband UVB sensor was installed at the site in 2005.

Spectral measurements of solar UVR were performed at the same site by Tartu Observatory. Since 2004, UVR spectra in the range 280-400 nm were collected using Avantes Inc. simple array spectrometer AvaSpec-256 with a 15-minute interval [24]. In 2009 these measurements were stopped due to the significant drop of array sensitivity. In 2008, the purchase of a spectrometric system based on the Bentham Instruments Ltd. DMc150F-U double monochro‐ mator was realized by funding from the EC REGPOT project EstSpacE. The system was installed in Spring 2009. Spectra in the wavelength range 280-400 nm are recorded by this instrument, also every 15 minutes.

In both systems the radiation-collecting diffuser is placed on the roof and connected with a quartz fibre to the spectrometer installed in the special weather box in the building. Calibration of optical instruments has been performed at Tartu Observatory for several decades. It was based on the 1000 W quartz halogen standard lamp FEL calibrated by Oriel traceable to the USA National Institute for Standards and Technology (NIST) [25-27]. Spectrometer AvaS‐ pec-256 needed recalibration after at least two months. The second disadvantage was a relatively high stray light level within this compact instrument and the necessity of the removal of its contribution in the recorded spectra.

A programme for the compensatory calculation of the stray light influence of the array spectrometer AvaSpec-256 was applied. The slit-scattering function of the spectrometer was measured directly using a 450 W xenon arc source and monochromator at the Metrology Research Institute, Aalto University, Finland. The stray light level of the AvaSpec-256 spec‐ trometer was rather high (0.1-1%), but the slit-scattering function is symmetrical and without noticeable artefacts. The uncertainty estimation of the stray light correction is based on the empirical comparison of the simplified algorithm and deconvolution over the set of measured spectra. For the Bentham double monochromator, a calibrator CL 6 belonging to the set of spectrometer auxiliary instruments was regularly used for checking the instruments' sensi‐ tivity, and the calibrator itself was regularly compared with a certificated FEL lamp.

#### **2.3. Spectra collection**

SDL-1 supplied with a mirror system and applying the Dobson retrieval algorithm [22, 23]. Since 2003, direct sun column ozone measurements were performed using a MICROTOPS-II instrument. Mostly, the satellite data were used for atmospheric column ozone [23]. Since summer 2002, a sun photometer of the NASA AERONET network measuring column aerosol optical depth (AOD) and precipitable water vapour operates there, and the aerosol studies group of the University of Tartu performs ground-level atmospheric aerosol-size distribution measurements. For the years before 2002 only pyrheliometer-measured broadband AOD data

Solar UV radiation measurements with filter instruments were performed at the Tartu-Tõravere Meteorological Station under scientific supervision by research scientists of Tartu Observatory [19]. The erythemally-weighted sensors UV-SET had been in operation since January 1998 [12-14]. A Kipp & Zonen narrowband filter instrument CUVB1 with an effective wavelength 306±0.2 nm and bandwidth 2±0.5 nm operates at Tartu-Tõravere meteorological station since 2002 [16]. Similar UVB instruments were installed at two other meteorological

45' E) and Pärnu (58o

broadband UVA sensor as well as a YES broadband UVB sensor was installed at the site in

Spectral measurements of solar UVR were performed at the same site by Tartu Observatory. Since 2004, UVR spectra in the range 280-400 nm were collected using Avantes Inc. simple array spectrometer AvaSpec-256 with a 15-minute interval [24]. In 2009 these measurements were stopped due to the significant drop of array sensitivity. In 2008, the purchase of a spectrometric system based on the Bentham Instruments Ltd. DMc150F-U double monochro‐ mator was realized by funding from the EC REGPOT project EstSpacE. The system was installed in Spring 2009. Spectra in the wavelength range 280-400 nm are recorded by this

In both systems the radiation-collecting diffuser is placed on the roof and connected with a quartz fibre to the spectrometer installed in the special weather box in the building. Calibration of optical instruments has been performed at Tartu Observatory for several decades. It was based on the 1000 W quartz halogen standard lamp FEL calibrated by Oriel traceable to the USA National Institute for Standards and Technology (NIST) [25-27]. Spectrometer AvaS‐ pec-256 needed recalibration after at least two months. The second disadvantage was a relatively high stray light level within this compact instrument and the necessity of the removal

A programme for the compensatory calculation of the stray light influence of the array spectrometer AvaSpec-256 was applied. The slit-scattering function of the spectrometer was measured directly using a 450 W xenon arc source and monochromator at the Metrology Research Institute, Aalto University, Finland. The stray light level of the AvaSpec-256 spec‐ trometer was rather high (0.1-1%), but the slit-scattering function is symmetrical and without noticeable artefacts. The uncertainty estimation of the stray light correction is based on the empirical comparison of the simplified algorithm and deconvolution over the set of measured spectra. For the Bentham double monochromator, a calibrator CL 6 belonging to the set of

23' N, 24o

38' E). A Kipp & Zonen

26' N, 24o

are available.

122 Solar Radiation Applications

2005.

stations Tallinn-Harku (59o

instrument, also every 15 minutes.

of its contribution in the recorded spectra.

The spectral data collection system including software for the AvaSpec-256 spectrometer was designed at the Tartu Observatory. UVR spectra were recorded as separate files but were also grouped automatically on the calendar day level. This allows relatively easy selection of all spectra recorded during each day for further treatment and analysis. One of the main proce‐ dures in data treatment is calculating the received energy within different wavelength ranges and time intervals. The ratios of UVA/UVB irradiances were calculated automatically for each spectrum, allowing the ease of obtaining the same ratios in daily or part-day doses. In the case of the Bentham DMc150F-U spectrometer, the producers' software BenWin+ was used for instrument control and data recording.

The measured spectra were recorded in the memory of the control computer as separate files. The name of the file contains information on the time of recording (year, month, date, hour and minute of the beginning of spectrum record). Later the files were transformed to the Excel environment and organized as workbooks for each month and sheets for each day. Around the summer solstice, the amount of informative spectra per day was about 75 and around winter solstice it was about 25.

The database of spectra was supported by two other databases useful for grouping spectra on the bases of different seasonal and weather conditions. One of them contains all supporting daily data, such as the daily doses of direct, diffuse and global broadband radiation, daily sunshine duration, daily erythemally-weighted and UVB narrowband doses, atmospheric column ozone, AOD, cloud and snow cover data. Direct and global irradiance relative to assumed normal clear conditions for each day are included. Normal clear conditions mean those for a typical column ozone and AOD for that calendar day.

Smoothed annual cycles of daily normal clear daily sums of global and direct irradiance have been composed empirically using the respective data since 1955 and have been used as the reference in the reconstruction of erythemally-weighted daily doses back to 1955 [16]. The other auxiliary database contains hourly sums of broadband direct, diffuse and global irradiance useful for studying relationships between broadband and UV radiation in a day. The hours are accounted from local noon in real solar time to both sides.

The dataset of UV spectral irradiance allows performing of integration of spectral energy by wavelength ranges and time. For example, it allows easy calculation of UVB and UVA spectral energy for different time intervals. Often, there arose a necessity to integrate different action spectra-weighted energies over days, parts of days and over longer intervals like weeks or months. Those products are useful in making comparison of received spectral energy in various conditions. One possible application is a study of relationships between UV doses in different wavelength ranges and the hourly sums of the pyranometer-measured broadband irradiance.
