1. Industrial UV Applications and Measurements

1. Measurement of Intense UV-C radiation

2. Proposal for Industrial UV-Radiometry

3. The SPECTRO320D-A Fast Scanning Spectroradiometer Based on a Novel Double Monochromator Design

4. UV Radiometry and Calibration in Non-Destructive Testing

5. Cosine Correction of UV Measurements

6. A Stabilised Transfer Standard System for Spectral Irradiance

7. Developments in Deuterium Discharge Lamps

2. Measurement Standards and Calibration Methods

1. New Developments in Ultraviolet Measurement at NPL

2. Radiometric Standards of the PTB

3. Monochromator-based cryogenic radiometry at NMi-VSL

4. Establishment of a facility for the absolute calibration of broadband detectors

5. Stability and quantum efficiency of a novel type of a-Si:H/a-SiC:H Based UV Detector

6. A Portable Lamp Unit for UV-Instrument Inter-comparisons

7. Detector Based Calibration of Solar UV Radiometers

8. UV-Calibrations for Spectral Irradiance and Spectral Responsivity and Their Uncertainties

9. Generation of Blue and UV Radiation by Frequency Doubling of Diode Lasers

10. Filter Radiometry Based on Direct Utilisation of trap detectors

11. UV-Biological irradiance measurements in BNM-LNE

3. Solar UV Measurements

1. Current Needs for Solar UV Spectroradiometry

2. Solar UV Monitoring by BfS/UBA, Improved QA/QC With Newly Developed Calibration Units

3. A Multi-channel UV-B Spectrometer

4. An Improved Diffuser for Global UV Irradiance Measurements

5. Sensors for Measurements of Atmospheric UV Radiation

6. Solar UV Metrology: Activity Report of the PTB

7. Artificial UV for Biological Experiments

8. Cosine Correction of Brewer Global UV spectra

9. Measurements of Spectral and Broadband Ultraviolet Radiation Performed by the Norwegian Polar Institute in Ny-Ålesund, Svalbard, Norway. Instruments and Applications

10. Solar UV Measurements at the Agricultural University of Norway

11. The Nordic Intercomparison of Ultraviolet and Total Ozone Instruments at Izaña, October 1996

12. Monitoring UV irradiances; quality control and cosine correction

13. A Review of Solar UV Measurements and Modelling at SMHI (Swedish Meteorological and Hydrological Institute)

4. UV Measurements Related to Health and Safety

1. Ultraviolet Radiation Measurement in Medicine

2. Type-testing of sun tanning devices

3. Measurement of the erythemal effective dose caused by UV-B and UV-A with a detector film (biochip VioSporâ).

4. UV-health Hazard Assessment- Guidelines, Measuring Methods and Equipment

5. An Electrical Dosimeter (ELUV-14) for Personal-Related Outdoor UV-B-Dosimetry

6. Determination of the erythemal solar radiation and spectral characterisations of various UV radiation sources using the biochip VioSporâ - a new UV-detection film system

Industrial UV Applications and Measurements

Measurement of Intense UV-C radiation

There are important industrial UV applications as for instance photochemical curing of lacquers, colours and plastics, UV oxidation of pollutions or UV synthesis of chemicals where rather high UV irradiances up to several kW per m² are applied to the objects. For optimum process control, such irradiances, mostly in the UV-C, have to be measured quantitatively, mostly continuously over an integral range in the air UV-C, UV-B and/or UV-A. So, detectors should be stable over thousands of hours against intense and hard UV radiation, solar-blind, calibrable and linear also in the high-intensity range and sufficiently responsive over a broad wavelength range.

It is rather difficult to meet such requirements. For stability and linear response, all UV sensors, except for the MI sensors developed by us, need a signal-attenuating coating that itself is affected by hard radiation, humidity and contamination. In order to fit the wavelength range to be measured, optical filters and/or phosphor layers are positioned in front of detectors. However, solarisation, enhanced temperature and humidity strongly influence spectral transmission. Diffusers put in front of sensors do not perform cosine correction equally for all incident wavelengths and sufficiently for angles larger than 60°. With respect to ageing by intense radiant exposure, diffusion type Si pn-photodiodes become unstable under irradiation below 250 nm. Diffusion type np-junction diodes with nitride oxide passivation layer exhibit an increased radiation hardness in the air UV. Best results are obtained with Schottky-barrier diodes and the spatially more uniform PtSi - n-Si diodes, because there is no oxide passivation layer in which traps can be formed by exposure to hard UV. All such sufficiently stable sensors, except for SiC-photodiodes, are not solar-blind, but have highest response in the VIS or IR. SiC-diodes have the disadvantage of rather small, not very uniform sensitive area.

We have tried to develop some alternatives with respect to the state of art. Phototubes were made with bulb of glass, for instance Vycor, that has the right cut-off wavelength and has been pre-aged under strong UV. For the air UV, Sn and Au have proved as rather stable coatings of the photocathode. Under very intense UV-C, they need an attenuating PTFE diffuser in front.

Photoelectric MI sensors have been made of metal-coated insulators. The semitransparent front layer of Al, Cu, Cr, Au or Pt forms the photocathode. The thin insulating semiconductor, for instance sapphire or magnesium oxide, determines the wavelength onset of responsivity at about . The opaque metal back layer gives the anode. A voltage of a few hundred volts is applied across the insulator. In contrast to Schottky diodes, the metal coating generates an accumulation layer at the MI interface. An UV-C exposure of 750 kJ/m² (by xenon short arc) produced a responsivity decrease of less than 3 % of the initial value.

Another problem associated with the measurement of intense UV-C radiation is the calibration of spectral detector responsivity, because there are no transfer standards, neither lamp nor detector standards, available for such high UV-C irradiances.

Proposal for Industrial UV-Radiometry

Scope of the proposal

It is well known that measurements of UV-radiation can be performed with spectral and integral methods. With the spectroradiometry, highly accurate and resoluble measuring results can be achieved, which are necessary for scientific documentation. However, spectroradiometers with acceptable quality are cost- and service-intensive and require qualified technicians.

Measuring equipment according to the integral method generally are portable (hand held), battery-powered, mostly payable and usable for high-speed signals, if they are equipped with suitable electronics. Beside this, there are some problems with such measuring equipment:

• Up to last two years, existing integral UV-meters are not accurate,
• There is not a common internationally accepted standard characterising and classifying UV-meters into different classes corresponding to different UV-measuring tasks.

These problems are to be overcome in the next future.

Existing guidelines for characterising radiometers and photometers of integral method

Up to now, following guidelines are worked out:

1. CIE Publication No. 53, 1982: "Methods of characterising the performance of radiometers and photometers,"
2. German standard 5032 - part 6 (only for V()-photometers): "Photometry, photometers, concept characteristics and their designation,"
3. German standard 5032 - part 7 (only for V()-photometers): "Photometry, classification of illuminance and luminance meters."

Papers 1. and 2. defined all uncertainties associated with radiometer heads and display units, among them the spectral matching uncertainty, the directional response according to cosine-law and the modulated radiation. Based on German standard 5032-6, an European Standard for photometers is under development.

Recommendations for future works in the industrial UV-radiometry

It is recommended to carry out developments of an European Standard classifying UV-radiometers with participation of instrument manufacturers and national standard institutes. Instrument manufacturers should indicate in their product documents the uncertainty limits of UV-meters for different types of uncertainties. Furthermore, from practical point of view, organisation of some workshops comparing the uncertainties of UV-meters of different manufacturers with discussions on suitable calibration methods and standards should be recommended.

The SPECTRO320D-A Fast Scanning Spectroradiometer Based on a Novel Double Monochromator Design

The SPECTR0320D measures UV­B Irradiance 20 to 50 times faster and with an order­of­magnitude better wavelength accuracy than existing commercial units.

The measurement time is just 5 seconds for the spectral range 280 to 320 nm and the wavelength accuracy is 0.03 nm. At 1 nm band-bass the minimum detectable irradiance is 1 W/m²nm.

Meteorologists and atmospheric scientists can now make accurate measurements of small­scale temporal fluctuations in UV­B solar radiation levels.

UV Radiometry and Calibration in Non-Destructive Testing

Levy Hill Laboratories and its predecessor, Levy West, have been supplying radiometers for the non-destructive testing (NDT) market for over 25 years. The latest product, a low-cost hand-held combined UV/VIS radiometer/photometer will be described in its design, operation and its usage for a range of applications, e.g. testing of lamps used in fluorescent penetrant techniques. The instrument is intended for laboratory or field use, is hand held, robust and resistant to attack by a wide variety of chemicals.

In addition, Levy Hill Laboratories is a designated calibration authority for leading aerospace and defence companies in the UK, including Roll Royce, British Aerospace, Westland Helicopters and the MOD. The techniques used for UV calibration will be discussed .

Cosine Correction of UV Measurements

The correct measurement of irradiance requires an entrance optics which makes detection of irradiance proportional to the entrance angle. This is necessary not only for exact solar measurements but also in order to quantify photobiological or photochemical effects produced by extended UV sources near to the treated surfaces. The usual correcting methods as the Ulbricht sphere or the transmitting diffuser have a rather strong dependence of cosine error on wavelength in the shorter UV. We have experienced better results by using a reflective optics with a specially treated PTFE plate.

A Stabilised Transfer Standard System for Spectral Irradiance

High-power quartz-halogen lamps are in use as transfer standards for a long time. However, even the best lamps available are not stable enough, to match the increasing requirements due to the progress of the measuring technique. Especially the long term stability in the UV-B region has to be improved. In order to reach higher long-time stability, it is necessary to change the operating conditions from constant current to a detector controlled system.

The new transfer standard system that consist of a controlling unit including power supply and a lamp-detector-unit is described. The main features are the long time stability of more than 500 hours without re-calibration and the indication of the actual data and their deviation from the calibration values on an integrated display, that allows a permanent supervision of the stability of the lamp.

Typical results of the characterisation of the new transfer standard system are presented in order to demonstrate the performance of the new system. The extension of the spectral region by using a system based on deuterium lamps is discussed.

Developments in Deuterium Discharge Lamps

This poster deals with recent developments in Deuterium discharge lamps in the following areas of application.

• Deuterium Lamps used as Calibration Standards.
• Higher Power Deuterium Lamps and Applications.
• Deuterium Lamps for industrial Applications.
• Deuterium Lamps for Analytical Instrumentation,

The usage of the Deuterium discharge lamp continues to grow. New applications are beginning to develop away from the conventional applications in standards laboratories and in laboratory analytical instrumentation. This has created demand for improvements in lamp performance.

The poster highlights some of the areas in which developments have been made in order to bring about improvements.

Measurement Standards and Calibration Methods

New Developments in Ultraviolet Measurement at NPL

An overview of developments will be given on the following subjects: primary UV emission scales based on high temperature black bodies; use of lasers in UV measurement; novel detectors for the UV; detector stabilised sources as UV transfer standards; measurement of diffuse UV transmittance by materials such as textiles.

Radiometric Standards of the PTB

The PTB offers calibration services of radiation units both of detectors in the spectral range between 248 nm and 2500 nm and of radiators in the spectral range between 200 nm and 2500 nm where detectors are compared to detector standards and sources to source standards.

The absolute spectral responsivity S() of detectors is based on several electrically calibrated thermal detectors realised with the lowest uncertainty (<10^-4) in the PTB's cryogenic radiometer facility used as a primary standard. In the visible range, calibration values are determined at the laser wavelength of an Ar+ laser (488 nm, 514 nm) and of He-Ne laser (633 nm) using the cryogenic radiometer. In the UV range the spectral responsivity scale could be transferred to detectors at discrete Hg lines (248 nm, 265 nm, 280 nm, 289 nm, 302 nm, 313 nm, 334 nm, 366 nm) in a standard calibration facility. A broad-band cryogenic radiometer covering the range above 200 nm without gaps is being installed in the clean-room centre allowing a continuous calibration at an uncertainty level of cryogenic radiometers.

The spectral irradiance scale of radiators is based on the thermal radiation of high- (<3000 K) and very high-temperature (3300 K) black-bodies as primary standards used with well-defined and precisely determined geometries. The temperature of the radiation is measured using absolutely calibrated broad-band filter detectors. Applying this method, the achieved uncertainty of the temperature measurement is better than 0.5 K. In a direct comparison against the black-body radiator, an uncertainty of the spectral irradiance scale of 1% could be reached at 300 nm using lamps of the FEL type. In a new facility, it was realised for the first time to extend the use of thermal radiators like FEL lamps instead of deuterium lamps and very high-temperature black-bodies as standards of spectral irradiance down to 200 nm covering the whole air-UV range. Both the standard calibration facility of the spectral irradiance will be improved and the reliability and an optimised use of working standards will be examined. Thus, a more stable irradiance working standard is proposed.

The reliability of scales are demonstrated in several international inter-comparison (spectral responsivity: CCPR inter-comparison 1990 and 1994; PTB/BIPM Cryogenic Radiometer Intercomparison 1995; spectral irradiance: CCPR inter-comparisons 1989 and 1994). The offer and specifications of the calibration services of radiometric units are published in the world wide web pages of PTB (www.ptb.de).

Monochromator-based cryogenic radiometry at NMi-VSL

A new facility for spectral calibrations of transfer radiometers is presented. The facility includes a cryogenic radiometer and a double monochromator. The present operating range is between 250 and 2500 nm. In the near future this range will be extended to 200 nm - 15 m.

Establishment of a facility for the absolute calibration of broadband detectors

A facility for the absolute calibration of erythemal broadband detectors is being developed at the Laboratory of Atmospheric Physics to serve the quality control of its monitoring network. It comprises two units, one for the determination of the spectral response and a second for the cosine response of the broadband detectors.

The spectral response is determined by measuring the quasi-monochromatic light, provided by a xenon source through a monochromator at different wavelengths, alternatively by the detector and by a spectroradiometer, and by taking their ratio. This cycle is repeated until the whole region of the detector's sensitivity is covered. Due to the reduced sensitivity of these detectors in the UVA region, it is necessary to increase accordingly the bandwidth of the monochromator. Having determined their spectral response, the absolute radiometric calibration of the detectors can be achieved by comparison with collocated spectroradiometric measurements under the sky. To decrease the uncertainty of the calibration, these measurements are performed under high-sun and clear-skies conditions.

The second phase of the calibration is the conversion of a detector's measurements to CIE weighted erythemal irradiance. This is achieved by applying a model-derived conversion factor, which is function of total ozone and solar zenith angle.

A third step, which is still under investigation, will comprise the application of cosine correction to the measurements, to compensate for the imperfect cosine response of the detectors. This correction factor will be calculated following existing procedures and it will based on the assumption of isotropic directional distribution of diffuse radiation, and on estimates of the direct-to-global ratio from the absolute irradiance level for the given solar zenith angle and total ozone.

The basic equipment that are used in the facility are:

• A Light source (ORIEL 66024) comprising a 1000-W Xenon lamp, the lamp housing with a rear reflector, the power supply, a light-beam condenser with an infrared absorbing water filter and a focussing lens.
• A double monochromator (SPEX 1680) with five fixed slits assemblies of effective band passes between 0,45 - 9 nm and two 1200 l/mm ruled gratings blazed at 330 nm.
• A Brewer MKIII double monochromator spectroradiometer with 286-366 nm spectral range, 0,55 nm resolution, and 0,5 nm scanning increment. Its measurements are regularly calibrated with a NIST traceable source and are corrected for its cosine response.
• A removable detector mount allowing X-Y-Z positioning and X-Z tilting, with the ability of repeatable re-positioning.

Under consideration is the use of a photo-feedback system for controlling the lamp's stability and the use of a calibrated silicon detector instead of the Brewer spectroradiometer. The immediate plans, before starting the operation of the facility, are the examination of the stability of the lamp system within several minutes (that are requited to complete a full cycle of measurements) and the optimization of the monochromator's band-pass at the various spectral regions between 280 and 400 nm.

Stability and quantum efficiency of a novel type of a-Si:H/a-SiC:H Based UV Detector

The request of large area and low cost UV detectors with rejection of the visible spectrum is very wide-spread in commercial and scientific applications. Photodetectors commercially available are crystalline silicon photodiodes and CCDs. They can be fabricated only over small areas and show a response in the visible spectrum equal or even greater than the response in the UV range.

The utilisation of amorphous photoconductors, such as a-Si:H and a-SiC:H, allows the production of large area UV-sensitive devices at a low cost. An efficient rejection of the visible spectrum is possible with the deposition of very thin active layers. This allows the production of p-i-n junctions in which visible radiation is transmitted, while UV radiation is uniformly absorbed in the whole thickness, due to the greater absorption coefficient of the amorphous silicon-based photoconductors in the UV band.

In this work, we have investigated the possibility to realise by p-i-n a-Si:H/a-SiC:H based junctions a sensitive detector with high selectivity and high performance stability. A set of devices has been produced on corning glass substrates of 10cmx10cm area covered with a SnO2 layer. The i-type and n-type layers were hight electronic quality a-Si:H films, while p-type layers were a-SiC:H wide band-gap materials. Thickness of doped layers has been fixed to 20 nm, while thickness of the intrinsic layers has been varied from 20 to 100 nm.

The quantum efficiency of the devices (i.e. the number of electrons collected per incident photon) has been measured: values of 0.6 for UVA radiation at 365 nm and of 0.15 to 0.03 for visible radiation in the range 400 nm to 650 nm have been obtained for the better devices, showing a good selectivity.

The stability of the devices has also been investigated, by irradiating the photodetectors with 2000 Joules of laser UV radiation at 353 and 361 nm. A variation in responsivity lower than 10% has been observed.

A Portable Lamp Unit for UV-Instrument Inter-comparisons

A Portable Lamp System (PLS) universally applicable for point­source calibration intercomparison of spectroradiometers has been designed at the Swedish Radiation Protection Institute. The PLS will be described and also experiences from circulations of it between Nordic solar UV laboratories. Some results from its use at two Nordic Solar UV Instrument inter-comparisons on Tenerife 1993­1996, and how far it improves inter-comparability of UV-instruments will also be presented. The PLS's present limitations as a point-source to make measurements inter-comparable may be further reduced, but as other errors then may dominate problems with regard to solar UV­instrument inter-comparability they need to be addressed.

Detector Based Calibration of Solar UV Radiometers

The assessment of health and ecological impacts of the ozone depletion is difficult since no firm experimental evidence exists on the persistent increase in solar UV in the global scale. The trend estimates are commonly based on theoretical radiation transfer calculations and ozone measurements, which do not take into account other atmospheric variables such as changing cloudiness and increased amount of aerosol particles having considerable influence into the terrestrial UV. Accurate measurements are also needed for establishing local UV climatologies, for verifying radiation transfer models, for verifying UV forecasts, and for informing the general public.

The main obstacle in improving the stability and global comparability of solar UV measurements is the variability of the measurement scales traced to different primary standards and the lack of stable transfer standards used to disseminate the scales to the users of UV radiometers. Intercomparisons of 1 kW quartz-halogen standard lamps, most commonly used for the calibration of spectroradiometers, show differences up to 8 % , and sudden changes of several per cents are not uncommon. These and other radiometric uncertainties are a formidable obstacle for the verification of present trends of global solar UV, typically of the order of 10 % .

A promising solution for realisation and dissemination of accurate spectral irradiance scales is the use of stable narrow-band standard detectors instead of unstable standard lamps. This is the main objective of the present study. The general aim of the project is to improve the calibrations of solar UV monitoring spectroradiometers to the level sufficient for detecting the present trends in solar UV radiation. The uncertainty of standard lamps will be reduced from ± 3 % to ± 2 % and the calibration will be transferred to the solar UV monitoring spectroradiometers with an absolute uncertainty less than ± 3 % and long term stability less than ± 2 %.

The new UV-scale is based on stable trap detectors equipped with narrow-band filters (10 nm) at six wavelengths from 280 to 380 nm. The absolute spectral responsivity of the detectors are determined very precisely with accurate transmittance measurements and the absolute calibration is traced to a cryogenic absolute radiometer, presently the most accurate primary standard in optical radiometry. The calibration is transferred to 1 kW quartz-halogen standard lamps by measuring the irradiance with each filter radiometer and by accurately matching the spectrum with the irradiances at each band. As a result, the number of calibration steps reduces significantly, the absolute uncertainty decreases along the whole calibration chain and the stability of the calibrations improves.

The detector-based absolute scale has already been established for the wavelengths 380 - 900 nm where the relative uncertainty (2 sigma) of approximately 0,4 % throughout most wavelengths. At 380 nm the uncertainty increases to 1,5 % since the absolute scale must be transferred with a pyroelectric radiometer from the cryogenic radiometer calibrated only at visible wavelengths. A preliminary experiment was also carried out at 312 nm where a promising agreement of 1 % was obtained with a 1 kW quartz-halogen standard lamp supplied and calibrated with NIST.

The problem of transferring the calibration from the laboratory based standards to the spectroradiometers monitoring solar UV will be solved by developing a portable calibrator where a quartz-halogen standard lamp will be directly calibrated with the filter radiometers and stabilised with filter detectors. Since at least two 10 nm bands at UV-B and UV-A ranges will be monitored, the change of the level and slope of the spectral irradiance can be compensated by adjusting the current of the lamp and/or by correcting numerically the spectrum in the exit aperture.

UV-Calibrations for Spectral Irradiance and Spectral Responsivity and Their Uncertainties

Introduction

Today, Gigahertz-Optik Company has the only German calibration laboratory for radiant quantities in a wide wavelength range from UV up to IR in the frame of the German calibration Service (Deutscher Kalibrierdienst = German Calibration Service). DKD-laboratories are accredited and supervised in Germany by the Physikalisch-Technische Bundesanstalt (PTB). The German Calibration Service is the member of the EAL (European Cooperation for Accreditation of Laboratories). The aim of the EAL is the establishment of an European base of mutual acceptance of calibration certificates issued by the laboratories accredited by EAL-members.

DKD-Calibration for the spectral irradiance

Generally, the calibration of the spectral irradiance can be carried out by comparing the output signals of a suitable double-monochromator setup delivered by the lamp to be calibrated and the standard lamp under the same irradiation conditions. The calibration arrangement in the DKD-laboratory consists of a movable table containing the standard lamps and the lamp to be calibrated, input optics putting the radiation of standard and calibrated lamps onto input slit of the double-monochromator, a double-monochromator strongly reducing the stray light, different detectors for calibrations between 250 nm and 2500 nm, a data acquisition system with measuring and system controlling software. With this software, photosignals of the detector can be evaluated for determining the absolute spectral irradiances.

The DKD-laboratory of Gigahertz-Optik Company uses as standard lamps for spectral irradiance in UV-range 1000 W halogen incandescent lamps of FEL-type. The FEL-lamps of OSRAM/ SYLVANIA were carefully tested by PTB and are suitable for calibration purposes in UV-ranges from 250 nm. For the PTB-standard lamps, used for DKD-calibration purposes, an uncertainty of 0,001 µW cm^­2 nm^­1 for =250 and 260 nm and 3 % for 270 nm <  < 400 nm are estimated. The transfer uncertainty in the DKD-laboratory can be estimated to be 1% resulting into a DKD-uncertainty of 4 % between 270 nm and 400 nm.

The calibration of the spectral responsivity

The spectral setup for the calibration of the spectral responsivity of detectors is generally the same for the calibration of the spectral irradiance. The DKD-standard detectors, together with the photocurrent meter, are calibrated by PTB. They are silicon detectors (HAMAMATSU S1337-1010 BQ) with quartz window. Their inside temperature is measured with PT100-resistance. According to investigations of PTB and NPL (UK), these detectors have a shunt impedance of about 2 G yielding a linearity of better than 0,05 % over 6 decades and an ageing rate of better than 0,3 % per annum between 250 nm and 1000 nm. PTB gives a relative uncertainty of 0,7 % at 248,3 nm and 0,5 % for all other UV-wavelengths and an additional absolute uncertainty of about 0,4 %. At the DKD-wavelengths up to 366 nm, an uncertainty of 3 % and for calibrations at 380 nm and 400 nm, an uncertainty of 2 % can be estimated.

References:

[1] PTB- Calibration certificate No. 41352-PTB-95 ( 03.08.1995)

[2] PTB- Calibration certificate No. 9111-PTB-97 ( 28.05.1997)

Generation of Blue and UV Radiation by Frequency Doubling of Diode Lasers

Compact, reliable and efficient blue and ultraviolet (UV) lasers are needed for many applications e.g. laser printing, optical data storage and spectroscopic analysis. In optical metrology, UV lasers are required for radiometric characterisation of detectors in the biologically interesting solar UV band 300-320 nm. Unfortunately, commercially available diode lasers cover only the red and infrared part of the spectrum. However, their useful band can be expanded by frequency doubling in a nonlinear crystal. The conversion efficiency can significantly be improved by placing the crystal in a resonant build-up cavity. We have constructed two resonantly enhanced diode laser systems to generate single-mode, continuous-wave radiation at 317 nm and 492 nm.

Frequency-doubled diode laser at 317 nm is accomplished by frequency doubling a 635 nm diode laser which has an output power of 15 mW. The beam is coupled into a build-up ring cavity that increases the optical power by a factor of 55. Rubidium dihydrogen phosphate (RDP) is used as the nonlinear crystal because of its noncritical phase matching around 630 nm. UV radiation of 34 µW was generated when 10 mW of the fundamental power was coupled into the cavity.

The second frequency doubling system consists of a 984 nm, 180 mW laser diode and a potassium niobate (KNbO3) crystal inside a build-up ring cavity. Potassium niobate is used because it has large nonlinear coefficients and noncritical phase matching can be achieved at the used wavelength. With this configuration, we are able to generate 4 mW of coherent light at 492 nm.

Filter Radiometry Based on Direct Utilisation of trap detectors

We have developed a new type of filter radiometer, comprising of a reflection trap detector, radiometric aperture and a set of temperature-controlled band-pass filters. The construction of the filter radiometer is presented in Figure 1. The temperature controller has been realised in a way that allows fast and reliable changing of the filters, without altering the alignment or contaminating the filters.

Figure 1. Construction of the new filter radiometer of HUT.

Due to the low reflectance of the trap detector, all the components forming the filter radiometer may be characterised separately. This is a significant advantage, as it allows the use of the most suitable instruments for all characterisations. The spectral responsivity of the trap detector is determined with the aid of a cryogenic radiometer. A high-accuracy reference spectrometer is used in the characterisation of the filters. The area of the limiting aperture is measured by using a novel laser-based method.

The filter radiometer has been successfully used in several applications, including a high-accuracy luminous intensity scale and a spectral irradiance scale in the 300-900 nm range. The spectral irradiance scale has been recently compared with a 1-kW spectral irradiance standard lamp that has a calibration traceable to NIST. This intercomparison indicates that in the UV region between 300 and 400 nm, the scales of NIST and HUT agree within 1 %.

UV-Biological irradiance measurements in BNM-LNE

BNM-LNE is a part of the Bureau National de Métrologie (BNM), which is responsible for the National Metrology activity in France.

At BNM-LNE, Radiometry and Photometry are performed in the optical division "Optical Radiation Metrology."

BNM-LNE participates to the organisation of the metrological approach needed for the application of a decret which has been taken by the French Health Department. According to this text, the UV-Biological irradiance levels delivered by UV systems used in suntan Institutes have to be measured. This measurements have to be carried out by control organisations.

In a first step, BNM-LNE will test the specifications of the available UV-Radiometers and spectroradiometers (geometrical diffused configuration, spectral characteristics, linearity). This step is started.

In a second step, BNM-LNE will calibrate radiometers for the control organisations and set up inter comparisons.

At this time, the traceability of the measurements is carried out by BNM-INM.

Solar UV Measurements

Current Needs for Solar UV Spectroradiometry

IFU performs high accuracy measurements of spectral solar UV irradiance. IFU operates three high quality UV­spectroradiometers: A stationary system in Garmisch­Partenkirchen (730 m), a stationary system at the top of the Zugspitze (3000 m) and a mobile spectroradiometer. The instruments are part of the Global atmosphere watch (GAW) programme established by the World Meteorological Organisation (WMO). WMO has established a scientific steering committee that should co-ordinate UV activities and advice WMO in the formation of a global UV monitoring network. The aims of solar UV-measurements are: 1. to understand the spectral effects in the UV region of changing atmospheric composition (e.g., ozone, aerosols, clouds) 2. to understand geographical differences in UV 3. to monitor long term changes in UV 4. to make properly calibrated UV data available to the community

With the help of the GAW activities the co-ordination of solar UV has been improved considerably in recent years. However, the UV measuring community has a lot of technical difficulties. To overcome these difficulties a close cooperation with state laboratories and industry is required. Some examples for further developments are:

­ Improvement of calibration accuracy

­ Development of a detector based calibration

­ Development of faster spectroradiometers

Solar UV Monitoring by BfS/UBA, Improved QA/QC With Newly Developed Calibration Units

Since 1993 the German Federal Office for Radiation Protection (BfS) in cooperation with the German Federal Environmental Agency (UBA) is operating a network for continuous monitoring the spectral resolved solar UV radiation. The present network consists of four monitoring sites at different locations in Germany. The site located in Munich-Neuherberg is the reference station where the QA/QC responsibilities are concentrated. The network is augmented by the UV measurement activities of four other federal state agencies and scientific research institutes. The technical specifications of the measurement systems (sensitivity 10^-6 W/m², spectral resolution 1nm, wavelength range 290-450nm, scan time 90s, scan frequency 6min from sunrise to sunset) are in accordance with radiation hygienic requirements.

The systems are operating at stable laboratory conditions. Only the entrance optics are placed on the roof outside, connected by a 4m lightfibre. So, mechanical and thermal problems connected with the outdoor operation of spectral radiometers are avoided. Therefore QC is focused predominantly on wavelength shift and absolute sensitivity. A possible wavelength shift is traced by highly resolved monitoring of the Fraunhofer absorption line at 393.36nm. Until now all systems show a very good wavelength stability ( 0.1nm). The system sensitivity is calibrated with a 1000W halogen lamp unit developed at BfS. To avoid changes in the optical path which can influence the system sensitivity the calibration unit has to be mounted directly on the entrance optics. Therefore a soffite type lamp was built into a sheet metal tube which fits to the console of the entrance optics. The black coloured inner part of the tube is equipped with two apertures to avoid reflections. The top of the tube is covered by a quartz glass dome as an efficient path for thermal radiation to prevent overheating. However, this construction does not protect the entrance optics from interfering daylight and therefore calibration can be carried out only during night time.

Because of the expense in time and personal this calibration procedure is only carried out in periods of about three months. To make a monitoring of short term changes in system sensitivity feasible an automatically working auxiliary lamp unit was developed. The mechanics of this unit is based on a modified wet-only precipitation sampler regularly used in air pollution monitoring. UV radiation is produced by a 35W halogen lamp. The lamp current is computer controlled with an accuracy of 10^-5A. During night the lamp unit is swivelled automatically over the entrance optics and after a half hour warm-up time 5 spectra are measured. Additionally the lamp current, voltage, socket temperature and total UV radiation provided by a SiC photo diode are monitored as auxiliary parameters. The spectral and auxiliary data are automatically transferred to the reference station where they are evaluated. If a long term drift is indicated a new absolute calibration is demanded. Until now the test of this system shows an overall variability of about +5% .

A problem with this auxiliary lamp unit is the commercial availability of appropriate lamps: the used lamps are the 12V types well known from room illumination. These lamps show a good long-term stability and a sufficient UV radiation strength. But as a consequence of health protection efforts lamp manufacturers are more and more using glass for the bulbs which is not transparent for UV radiation. Therefore the manufacturers are claimed to produce low power halogen lamps with high UV emission and optical stability.

A Multi-channel UV-B Spectrometer

One of the major objects of UV measurements is to detect long­term trends in UV irradiance. The most interesting areas corresponding to ozone depletion are Antarctica and of course the region around the northern pole. Long­term field measurements under these extreme climatical and logistical conditions could not be realised simply by using a commercial spectrophotometric laboratory system. It seemed to be necessary to develop a specialised instrument for this intended use. One of the most important features of the instrument should be a stable operation in field for a period of about one year without intensive manpower for service necessary.

For this purpose we built a non­scanning spectrometer based on a Bentham DM 150 double monochromator with a multi-channel detector system. The detector is based on a low­resistance microchannel plate with 32 channels working in a photon counting mode. We run our double monochromator with a broad centre slit to enable the detection of the whole UV­B range simultaneously. To handle the stray-light problem in the detector chamber due to the broad centre slit, we cut 3 magnitudes of dynamic in the UV­B range using a steep interference filter with transmission maximum at 289 nm. By this we enhanced the dynamic range of our system up to 10⁶.

In consequence of the fixed gratings, our spectrometer is less sensitive toward transport by ship and plane, and shows a better wavelength stability as scanning systems. The latter means that there is no further calibration necessary for about one year.

Because we get the whole UV­B spectrum at one time we can also measure under fast changing clouds. Particularly this is important in comparison of the spectra to model calculations of the radiative transfer including cloud processes.

Since March 1996 our system is installed in Ny-Ålesund and at Neumayer Station (Antarctica). In August '97 we participated in the German spectroradiometer intercomparison IC97 in Garmisch-Partenkirchen.

An Improved Diffuser for Global UV Irradiance Measurements

The angular response of diffusers used with spectroradiometers for measurements of solar UV irradiance is generally not in good agreement with the ideal cosine response. Although a correction procedure may reduce the misleading effect of the cosine error, this can not be done satisfactorily for all atmospheric conditions, i.e. broken cloudiness. Therefore, it is necessary to strive for a diffuser with an improved angular response. Based on a recent development of a new type of Teflon diffusers in New Zealand, we improved this new type furthermore by optimising the dimensions of the diffuser to give best response when an additional quartz dome is used. The manufacturing of the new diffuser is carried out with a computer controlled milling machine, so that the optimal shape can be reproduced exactly, The deviation of the new diffuser, combined with a standard quartz dome, from the ideal cosine response is smaller than ±3% for zenith angles up to 75. The corresponding diffuse cosine error, calculated for isotropic diffuse sky radiance, amounts to 1.56%. Therefore, the usage of this diffuser will allow global irradiance measurements on a level good enough to avoid any additional cosine error correction procedure with its inherent uncertainties.

Sensors for Measurements of Atmospheric UV Radiation

Scintec manufactures a line of broadband UV radiation sensors for atmospheric applications. Different models are available for UVA, UV­B and erythemal active UV radiation (CIE 1987). The spectral and angular responses have been designed and optimised for atmospheric radiation conditions. The sensors are thermostatted and weather proof allowing continuous outdoor operation.

Measurements of Spectral and Broadband Ultraviolet Radiation performed by the Norwegian Polar Institute in Ny­Ålesund, Svalbard, Norway. Instruments and Applications.

Solar UV Metrology: Activity Report of the PTB

In the field of solar UV measurements, the PTB is not involved in the measurement or monitoring of solar UV-B radiation but in the development of standards, instrumentation and methods that are needed in this field. The following recent projects in the field of UV radiometry were given top priorities: Improvement of the stability and reproducibility of selected quartz halogen lamps for use as a standard of spectral irradiance needed for the calibration of UV spectroradiometers and UV sources; development of a highly accurate, fast scanning spectroradiometer for UV monitoring; development of methods, working standard and a laser-based instrument for the calibration and characterisation of UV spectroradiometers in the laboratory and in the field.

A short status report is presented and future activities are summarised and discussed.

Artificial UV for Biological Experiments

Several researchers have pointed out that a realistic risk assessment of UV­B induced damages in organisms can only be obtained if the experiments are performed under natural light and radiation conditions. This applies particularly to the balance between the UV­B, UV­A and the photosynthetic active, radiation (PAR). Natural global radiation exhibits marked seasonal and diurnal variations of its spectral composition and a realistic simulation of solar radiation has, therefore, to account for such variations,

This contribution shows the present state of illumination for biological experiments and demonstrates that by using the state­of­the­art techniques, i.e. selected modern lamp types, appropriate spectral shaping methods [2,3] and a flexible electronic lamp control a large variety of natural radiation climates can be simulated. The spectral measurements demonstrate that both a steep realistic UV­B absorption edge and a naturally balanced UV-B:UV-A:PAR ratio can be obtained. In addition the simulation of enhanced UV-B scenarios is also possible.

REFERENCES:

[1] M.M. Caldwell, S.D. Flint and P.S. Scarle, Plant Cell and Environment 17, 267­276 (1994).

[2] S. Thiel, T. Döhring, M. Köfferlein, A. Kosak, P, Martin and H.K. Seidlitz, J. Plant Physiol. 148, 456-463 (1996).

[3] T. Döhring, M, Köfferlein, S. Thiel, and H.K. Seidlitz, J. Plant Physiol. 148, 115­119 (1996).

Cosine Correction of Brewer Global UV spectra

A methodology is proposed, which enables the cosine correction of the global irradiance spectra obtained by the Brewer spectroradiometers. To determine the cosine correction factor, it is required to know the ratio between the global and the direct component of solar radiation at each wavelength and the cosine response of the instrument, which is determined in the laboratory. In the calculation it is assumed that the diffuse radiation is isotropic.

If I denotes the actual and I' the measured solar irradiance, the cosine correction factor f_g of the global irradiance measured by a spectroradiometer is defined as

f_g = I'_g / I_g = (I'_d + I'_b) / I_g ,

where the indices g, d, b correspond to global irradiance and its diffuse and direct components respectively.

The Brewer software was modified to enable simultaneous recording of the direct irradiance during the global spectral irradiance measurements. The direct-to-global ratio is a smooth function of wavelength, so the direct irradiance is sampled at 10 nm intervals, and polynomial interpolation of the calculated direct-to-global ratio is used to match the 0,5 nm regular sampling resolution of the Brewer. To simplify the procedure, sampling of the direct component is done without the use of attenuation filters. In more sensitive spectroradiometers the attenuation filters must be used to avoid possible damages from PMT overexposure.

The Cosine Correction of global UV scans is applied regularly in Brewer #086 since mid 1996. The effectiveness of the methodology was tested by comparison with model calculations (being in principle free of cosine errors) and with simultaneous measurements of the University of Innsbruck Bentham DT150 spectroradiometer, which has a different cosine response and was also corrected for its cosine response. In both cases the diurnal variation of their ratio, caused mostly by the cosine error, was reduced significantly, down to about 2 %.

The errors arising from the non-ideal angular response of Brewer UV spectroradiometers can affect significantly the accuracy of their measurements, usually by 2-7 %. Therefore, appropriate correction methods are necessary for improving the accuracy of their measurements.

The proposed methodology is applicable to all existing Brewer spectroradiometers. However knowledge of the angular response for each particular instrument is required. In addition it is necessary that the direct solar irradiance measurements be calibrated in absolute scale.

Small uncertainties in the determination of the angular response and the direct to global ratio do not affect significantly the resultant correction factor, ranging from about 0,2 % to 2 %. This enables the calculation of reliable cosine correction factors even with less accurate knowledge of these parameters.

If we make the reasonable assumption that the angular responses of all Brewers are similar, the error in the determination of the cosine correction factor due to the assumption of isotropic diffuse skylight is expected to be less than 2 %.

Measurements of Spectral and Broadband Ultraviolet Radiation Performed by the Norwegian Polar Institute in Ny-Ålesund, Svalbard, Norway. Instruments and Applications

Introduction

At the Norwegian Polar Institute (NP) research station in Ny­Ålesund, Svalbard a number of permanent and campaign based measurements of solar UV­radiation are performed. The activity supports the increased interest from the international society to quantify the changing UV-environment in the Arctic and to assess possible biological effects of increased levels of UV/UVB-radiation as a result of antropogenic ozone­depletion. The research and monitoring programs run by NP and co­operating Norwegian and foreign research institutions are complemented by foreign biological UV­programmes at the other research stations in Ny­Ålesund. The programs benefit from the advanced research facilities within the framework of the "European Large Scale Facility for Arctic Environmental Research" in Ny­Ålesund. This contribution intend to present the instruments and facilities used for spectral and broadband measurements of UV­radiation by NP as well as some results from applications within biological effect studies and UV­monitoring.

Monitoring of Broadband and Spectral UV­radiation

Broadband measurements of solar UV­radiation (Eppley TUVR 295­385 nm) has been performed in Ny­Ålesund since 1981 by NP (Hisdal et al., 1992). Broadband UVB measurements have been performed since 1992 by means of RB­meters (Solar Light Company model SL501), measuring biologically effective erythemal UV­doses. A high resolution UV­spectroradiometer type Bentham DM150 is set into operation from summer 1997. In addition to the permanent instruments a portable Licor 1800UW Underwater spectroradiometer measuring low res. spectra from 300­800 nm is used for biological related field campaigns.

NP runs side by side at the research station in Ny­Ålesund the Bentham UV-spectroradiometer, a DOBSON­spectrophotometer (for Univ. of Oslo), a broadband UV­Biometer and a multi-channel UV filter instrument (for NILU). This extensive instrument park gives good opportunities for quality check and to meet the different biological and geophysical needs.

An optical calibration laboratory is being built up and instrumented within the framework of the Large Scale Facility. The lab is used by the permanent optical programmes and is offered to external institutions and co­operating institutions as well.

Applications

The main objective of the activities is to investigate the natural intensity and variability of arctic UV­radiation and biological effective UV­doses at different time­scales, including surface and underwater measurements. The measurements support biological effect studies on marine, terrestrial and freshwater ecosystems, with main goals to:

• Determine the daily and seasonal surface UV/UVB­climatology by means of spectral and broadband measurements in Ny­Ålesund, including the investigation of the influence of atmospheric ozone, clouds and albedo on the spectral distribution of the surface UV radiation in the Arctic, as well as the relation to short­ and long-wave radiation and meteorological parameters such as humidity and temperature.
• Determine the physical parameters controlling the attenuation of UV/UVB­radiation in arctic water columns such as DOC, sediments, chlorophyll and other oceanographic parameters connected to studies of marine primary productivity.
• Provide high time resolution biological effective UV­doses with different action spectra such as erythema dose rates, DNA­effective radiation, etc., on different time-scales and sites around Ny-Ålesund, as well as evaluate the relative intensity of biological effective UV­doses to other parts of the radiation spectrum such as UVB/UVA, UVB/PAR, etc., in support to biological effect studies on arctic terrestrial vegetation and freshwater plankton.

Solar UV Measurements at the Agricultural University of Norway

A meteorological station was established at the Agricultural University of Norway at Ås (35 km south of Oslo) in 1870. The purpose of the station is to study the interaction between the weather/climate and agricultural production. During the early years, measurements were restricted to parameters such as air and soil temperature, humidity and wind velocity. Since approximately 1950, these measurements have been extended to include solar radiation. Broadband ultraviolet measurements were commenced in 1977. Eppley TUVR radiometers have since been used to measure UV radiation in the wavelength band 290­385 nm. The data has been used to study UV effects on plant growth and the mechanical properties of materials as well as to assess ­atmospheric models of radiative transfer.

Calibration of the Eppley TUVR radiometers remains a challenge. We are working on establishing proper calibration routines such that the radiometers can be frequently and easily calibrated. Both laboratory and in the field methods are of interest. Such routines will ensure the reliability of the data. Future plans include extending the measurements by the addition of newer multi-band radiometers.

The Nordic Intercomparison of Ultraviolet and Total Ozone Instruments at Izaña, October 1996

The presentation and the related publication [1] summarise the experiences and new knowledge relating to UV measurements gained by the following activities organised, mainly in 1996, by the Nordic Ozone Group:

• a laboratory campaign for comparing devices used for measuring the calibration lamp current,
• another laboratory campaign for comparing the secondary standard lamps of home laboratories,
• the circulation of a portable lamp unit between monitoring sites in the Nordic countries,
• an extensive instrument intercomparison campaign together with a UV workshop.

The instrument intercomparison campaign was carried out at the Izaña, Global Atmospheric Watch Observatory, Tenerife, Spain, from 8 to 20 October, 1996, and was participated in by seventeen UV spectroradiometers and nine UV filter-radiometers. The purpose of the experiment was to gain information on the comparability of ozone and ultraviolet measurements and to verify if the measurements had improved during recent years. The main results of the intercomparison are the following:

• the agreement between spectral global UV measurements has improved during recent years,
• the use of lamp measurements in correcting for differences in the irradiance scale could noticeably improve the agreement,
• data-analysis methods, taking into account different instrumental properties, facilitated the comparison of spectral data. The reference for global sky measurements was determined by an objective algorithm,
• several instruments show potential for routine measurements of the direct solar beam irradiance,
• two of the filter-radiometers demonstrated a good relative stability over a long period of time. It is also shown that in comparison with the manufacturers' calibrations, centralised calibrations significantly improve the comparability. The general agreement was as expected from previous inter-comparisons,
• the agreement between the ozone instruments was acceptable and had the same accuracy as previous campaigns have shown.

The results are presented in the ten independent papers of the publication. Suggestions and some of the key results are collected as an Executive Summary at the beginning. A general outline of the experiments is given in the first paper.

REFERENCES:

[1] Berit Kjeldstad, Bjørn Johnsen and Tapani Koskela (Eds.), 1997: THE NORDIC INTERCOMPARISON OF ULTRAVIOLET AND TOTAL OZONE INSTRUMENTS AT IZAÑA, OCTOBER 1996. FINAL REPORT. Finnish Meteorological Institute. Meteorological Publications No. 36, FMI-MET-36. ISBN 951-697-475-9. 185 pages.

Monitoring UV irradiances; quality control and cosine correction

The main goal of the RIVM UV-monitoring program is to determine long-term trends in the UV climate. An UV monitoring system has been designed and built and is operational since '93. The system consists of a spectroradiometer, Robertson-Berger SL 501 biometers and pyranometers. The system is built in a light-tight and temperature stabilised mobile container. In the monitoring routine measurements are performed every 12 minutes from sunrise to sunset. The measurements are used for monitoring and validation of atmospheric UV transfer models, to obtain a tool for estimating past and future trends.

We have developed techniques to use the Fraunhofer structure in the solar spectrum to provide a high accuracy wavelength alignment, and to correct for slit function differences by means of deconvolution. Application provides a tool for the Quality Assurance of wavelength alignment in routine monitoring. The wavelength alignment technique has been extended, improved and extensively tested to enable wavelength dependent shift determinations in the UVB region of the spectrum and to increase the efficiency of the calculations to facilitate application in routine monitoring. The methods were thusfar applied on over 30 different instruments during large scale international intercomparison campaigns. The accuracy of the alignment method is 0,01-0,03 nm for instruments with slit functions with a FWHM around or below 1 nm. The repeatability is around 0,01 nm. Larger uncertainties can occur if the FWHM of the instruments is considerably higher, and for very low solar angles.

We also have developed a cosine correction method for UV spectral measurements which is applicable not only under clear sky or solid overcast conditions, but also under a variable cloud cover, i.e. for routine monitoring data. Many UV spectrometers are equipped with a diffuser plate to measure solar UV irradiances with non-ideal angular responses. A 'cosine correction' should be made to account for this error. In order to perform this correction, the angular distribution of the irradiation on the diffuser plate should be known, i.e. the ratio of direct to diffuse irradiation in first approximation. Our method provides a protocol for clear sky and all cloudy situations and is based on three steps:

1. The measured global irradiance, determines together with the expected level the reduction of the global irradiation by clouds or aerosols. This measured reduction of the global irradiation is translated to an optical thickness of the clouds.
2. A modelled direct to diffuse ratio for clear sky conditions is modified for the overcast condition using the derived optical thickness in step 1.
3. Using these modified values of direct and diffuse irradiance, the cosine correction factor is calculated.

The knowledge necessary for evaluating step 1 is based on Monte Carlo simulations of the total transmission as a function of optical thickness and angle of incidence. In step 2, a standard atmosphere is used with a fixed value for the ozone column and a low aerosol concentration to estimate the clear sky direct to diffuse ratio. This is sufficient for our purpose. We have investigated also the impact on the cosine correction of an anisotropic distribution of diffuse light emerging from the sky or clouds. Generally, it reduces the cosine correction.

A Review of Solar UV Measurements and Modelling at SMHI (Swedish Meteorological and Hydrological Institute)

There have been a number of projects on different aspects of measuring and modelling solar UV at SMHI. Most of them in co-operation with the Swedish Radiation and Protection Institute (SSI). It started in the early 1983 with measurements using a Robertson-Berger (RB) meter and a Brewer spectroradiometer. A simple empirical model based on the Brewer measurements was used to map the climatological distribution of ACGIH-weighted irradiance in Sweden [Josefsson (1986)]. This model was also used for computing clear sky values for lower latitudes to get an idea of the potential UV at typical resorts outside Sweden.

In the early 1990-ties a small network was established to study the UV-climatology in Sweden using Solar Light Model 500-instruments, which were available at that time [Josefsson (1996) and Josefsson (1997)]. Along with the UV-monitoring there has also been measurements of the total ozone (two sites). Our institute has participated in a number of inter-comparisons Norrköping 1991, NOGIC-93, NOGIC-96 and SUSPEN 1997. These has been very valuable for improving the quality of our data. In cooperation with SSI we have been producing forecasts of UV-index since May 1992. During the first years we had our own index (ACGIH, scale 0-100). But, in 1996 we adopted the internationally recommended scale. The UV-index is distributed daily during Spring and Summer and can be found on our web-site where also all our values of total ozone are available (www.smhi.se).

As sites for measurements are sparse we have started a project for modelling of radiation parameters with a meso-scale resolution, in this case 22 km, using available meteorological information. The initial phase includes UV-irradiance, global and direct solar irradiance and photosynthetic active radiation and sunshine duration.

SMHI is participating in the SUVDAMA, UVRAPPF and the Thematic Network projects sponsored by the European Community. Within the SUVDAMA we have contributed with five years of spectral measurements recorded by the Brewer spectroradiometer to the data base. In the UVRAPPF we are studying our long-term RB-series starting in 1983. Other small contributions will be on the influence of clouds, aerosols and directional characteristics of the instruments on the measured UV-irradiance.

REFERENCES:

Josefsson W. (1986), Solar Ultraviolet Radiation in Sweden, RMK. No.53, SMHI, October 1986.

Josefsson W. (1996), Five years of solar UV-radiation monitoring in Sweden, SMHI Reports Meteorology and Climatology, RMK No.71, Oct 1996, ISSN 0347-2116.

Josefsson W. (1997), Solar UV-radiation monitoring 1996, SMHI Reports Meteorology and Climatology, RMK No.74, Feb 1997, ISSN 0347-2116.

UV Measurements Related to Health and Safety

Ultraviolet Radiation Measurement in Medicine

This presentation will examine the situation regarding UV measurement in medicine. This will be analysed under three headings. First consider what is being done at present. Measurements are conducted within the medical field in several disciplines for different reasons, for example physiotherapy, dermatology, neonatal care, dentistry, laboratory, safety, cancer prevention. Examples will be given of the types of measurements performed. Next, the radiometric requirements for the various applications will be discussed. This encompasses UV-sensitive film badges, hand-held filtered photodiode instruments, and sophisticated spectroradiometers. Finally, the problems will be addressed. These relate to lamps with different spectra, patients with varied sensitivities, and calibration methods of uncertain significance.

Type-testing of sun tanning devices

The product directives of the European Union demand CE marking of products marketed within the EU. The CE mark is the manufacturers affirmation that all safety standards are fulfilled. However, it is no certification based on objective testing. A problem is that the manufacturers and national representatives often are not aware of the safety directives and EN standards, resulting in products entering the market without complying with them. That applies also to sun-tanning devices. The CE documentation for these devices often lacks important information, such as the UV type and specific type of UV emitting sources. The UV protocols, either made by the manufacturer, commercial testing laboratories or health authorities may diverge by more than 50 percent, due to

1. different interpretation of the EN 60 335-2-27 "Safety of household and similar electrical appliances Part 2: Particular requirements for ultra-violet and infra-red radiation skin treatment appliances for household and similar use,"

• the most important difference is whether the limits is a maximum or mean-value
• which UV type for cosmetic use

1. different measurement procedures,
   • what about uncertainty level, calibration routines, temperature effects, input optics, orientation of the monochromator in the actual measurement situation,
2. different calibration standards.

For the national authorities this results in additional product control and complications for the acceptance of tanning devices. There is a need for :

• Harmonised interpretation of the standard,
• Harmonised measurement procedures,
• Common scale of irradiance calibration,
• Ring test evaluation of measurement results for type-testing laboratories, and
• Criteria for single-lamp testing and evaluation.

Measurement of the erythemal effective dose caused by UV-B and UV-A with a detector film (biochip VioSpor^â).

For the first time, a detector film has been manufactured which can be used for dose measurements of biologically weighted solar radiation (CIE-MED) as well as for artificial lamps (solaria). Two examples are given to illustrate the effectiveness of the UV-B as well as of the UV-A spectral region in solar radiation and artificial lamps. These data underline the importance of correct measurements in the UV-B and also in the UV-A spectral region.

For the measurement of solar radiation, the detector system has to be optimised in the UVB spectral region. At high solar elevation angles, the UVA contributes only a minor part to the total erythemal effectiveness of solar radiation. In the given example UVA (315-400) contributes 23% of the total erythemal irradiation.

At lower solar elevation angles and for many artificial lamps, the UVA part of the spectrum contributes to a high extent to the total erythemal effectiveness. Under these conditions it is very important that a broadband detector gives reliable results not only in the UVB but also in the UVA. The given example of a medical lamp illustrates that in this case UVA contributes 47% of the total erythemal radiation.

UV-health Hazard Assessment - Guidelines, Measuring Methods and Equipment

With the strong depletion of the stratospherical ozone layer and consequently overproportional increasing of UV-B-radiation on the earth surface and with new application fields of ultraviolet radiation in medicine, biology, chemistry and industrial manufacturing processes, there is an urgent need to recognise and assess the potential radiate health hazards as well on the indoor as on the outdoor workplaces.

Generally, actinic effects can be divided into effects on the human skin (erythema, DNA, non-melanoma-skin cancer) and effects on the human eye (photoconjunctivitis, photoceratitis) and the general health damage on work places (ACGIH-actinic spectrum).

In the last 15 years, a number of international and national institutions have recommended relevant actinic functions being able to describe the UV-health damages on work places and Maximum Permissible Exposures (MPE) for occupational and public exposure to UV-radiation. All guidelines have recommendation character. Among them, the following two guidelines are mostly used:

• American Conference of Governmental Industrial Hygienists, ACGIH: "Threshold Limit Values for Chemical Substances and Physical Agents and Biological Exposures Indices," Cincinnati, 1995
• IRPA/ICNIRC: "Guidelines on limits of exposure to ultraviolet radiation of wavelengths between 180 nm and 400 nm" (1985 & 1989)

For the assessments of UV-health damages, a general health damages function was defined. This ACGIH-function describes the spectral effectiveness of UV-radiation on human skin and eye. The well-known UV-erythema function was defined for the skin reddening caused by UV-radiation and the threshold dose for sensitive skin type should be 250 J/m². Today, these functions are objects of intensive discussions today on long-term effects of UV-radiation (person-dosimetry).

For health damage assessments, two measuring methods can be used:

• Spectral methods: measurements of the spectral quantities to be needed [e.g. spectral irradiance E_e() or spectral radiance L_e()], evaluation with the action spectrum s()_biol,rel to be observed and integrating this product.
• Integral methods: measurements of actinic radiant quantities can be done with "radiometers for the measurements of actinic radiant quantities," using a radiometer head with a relative spectral responsivity s()_rel, which is spectrally matched to the action spectrum of the considered effect.

In the last time, Gigahertz-Optik has developed the radiometerheads spectrally matching the ACGIH- and erythema-functions from 220 nm up to 400 nm. The uncertainty of cosine-response is better than ±5% for incidence angles up to 60°. With the laboratory equipment, it is possible to measure fast changing radiation intensity (welding arcs, flash lamps, pulsed lasers…) up to 10 MHz.

For hand-held and laboratory measuring devices, the dose function is integrated. The devices are calibrated in the laboratory of the German Calibration Services.

An Electrical Dosimeter (ELUV-14) for Personal-Related Outdoor UV-B-Dosimetry

An electronic UV-B-dosimeter (ELUV-14) was developed to carry out dosimetry measurements at personal-related level. The ELUV-14 is a small robust waterproof 3-channel data-logger equipped with a special UV-B-sensor, Pt 100 temperature sensor and a silicon photodiode as brightness sensor. It was designed to carry it as personal outdoor dosimeter with dimensions of 80 X 40 x 20 mm and weight of 95 g including the Lithium battery. The UV-B-sensor was especially developed to have a relative spectral response similar to that of the standard erythemal response after McKinlay-Diffey ( CIE-87-standard). Those allowed the use of the ELUV-14-dosimeter for erythemal weighted dosimetry directly. The basic sample rate is one minute and could be set up to 255 minutes. The storage capacity by sampling rate of 1 minute is about 30 days using a Flash EPROM Memory.

System setting and data read out are controlled by a PC programme running under MS Windows and using an RS-232 serial ASCII-spreadsheet data transfer.

A test series under real conditions was carried out at the German station "Neumayer", Antarctica (70°39'S, 8°15'W). Several persons carried the dosimeter during the summer time at Neumayer-Station under real working conditions. Initial data analysis shows a high variability of UV-B-dose among the individuals due to different activities.

Properties, functions and results of the first use of ELUV-14 will be presented.

Determination of the erythemal solar radiation and spectral characterisations of various UV radiation sources using the biochip VioSpor^â - a new UV-detection film system

The detector film VioSpor® (spore film; biochip) was tested in several field campaigns. The spore film data (given as daily dose in CIE-MED) were compared with data (given as daily MED dose) deduced from spectroradiometric measurements with a Brewer instrument.

Date	Time interval	VioSpor [MED]	Brewer [MED]
		integrated measurements	hourly measurements plus interpolation
28.06.1996	sunrise - sunset	9,33	9,64
	sunrise - 12.00 12.00 - sunset	8,34	9,64
	1hour sequential 9.00 - 16.00	7,87	8,14

1 MED (CIE) = 250 J/m²

The VioSpor system includes a filter system. Various spectral measurements are done during one exposure experiment. One of the determined values gives the erythemal dose (MED) and the other are necessary for the spectral characterisation of the UV source. The ratio of the various values enable the determination of l_ER. This value is defined as the wavelength where 50% of the total erythemal effectiveness is caused by longer and by shorter wavelengths respectively. Three examples shall underline the potential of spectral characterisations with VioSpor.