Table of contents

The low-background infrared (LBIR) facility at the National Institute of Standards and Technology (NIST) has continued to develop its facilities and knowledge base to meet the needs of the infrared community. Improvements in refrigeration capability at the LBIR facility have made it possible to perform calibrations of infrared sources and detectors in a stable 17 K background environment as compared to a relatively unstable 25 K environment available until about two years ago. This, combined with improved power measurement instrumentation, allows measurements of 1 nW with a standard uncertainty of 1% due to repeatability and reproducibility effects. A brief overview will be given of the changes to the LBIR facility that led to these improvements. The higher sensitivity in power measurement capability and some of the methods being used to generate low-power beams have highlighted new measurement issues that had previously been relatively unimportant. These issues include aperture quality, background scene temperature stability, beam shuttering, diffraction, and the noise floor of power measurement hardware. Demonstrations of common problems encountered will be shown and guidelines will be given for developing infrared sources that not only meet the needs of the user but can also be well calibrated.

It can be difficult to evaluate Lommel functions U_n(u,v) and V_n(u,v) when u and v are both large. Here, we present an accurate means to evaluate such functions using contour integration. The approach suggests the potential for developing analogous techniques in diffraction calculations for many optical systems.

NIST has established a geometric aperture-area measurement facility for circular apertures. The instrument consists of an interferometrically controlled XY translation stage for high-accuracy positioning and a video microscope for detection of the edge of the aperture. Least-squares fitting of the edge points located along the aperture's inner circumference to the equation of a circle is used to determine the geometric area. In this paper we describe the measurement method, based on the work started and described by Fowler et al (1998 Metrologia35 497–500). Analysis and estimation of various contributions to the overall measurement uncertainty, including the effects of diffraction and partial coherence of light on the edge location, are also discussed.

The desire for high-accuracy infrared sources suitable for low-background seeker/tracker calibrations pushes the limits of absolute cryogenic radiometry and blackbody design. It remains difficult to calibrate a blackbody at irradiance levels below 1 nW cm⁻² using electrical substitution radiometry. Ideally, the blackbody temperature should be chosen so that most of the emitted power lies in the spectral range of interest. This constraint frequently necessitates the use of small apertures (less than 1 mm diameter) to achieve the required reduction in power. However, the dimensions of small apertures are difficult to determine accurately. Also, the effects of diffraction and systematic problems such as aperture heating and light leaks become amplified. To diagnose these problems and to calculate diffraction effects, aperture dimensions must be known accurately. We describe a technique of radiometrically deducing the diameter of small apertures in the blackbody in situ utilizing the cryogenic blackbody calibration.

The design of a portable calibrator for solar ultraviolet spectroradiometers is presented. The calibrator is based on a 1 kW DXW lamp mounted in an aluminium housing. The use of changeable adapter plates enables reliable setting of distance with different spectroradiometers. The lamp irradiance is monitored with two filter radiometers at wavelengths of 313 nm and 368 nm. The calibrator has been thoroughly tested in both the laboratory and the field. The results of the field trials indicate that the device works satisfactorily under a wide variety of environmental conditions. The calibrator can be used to establish the shortest possible traceability chain from the national standards laboratory to field measuring instruments.

The general methodology of the calibration of broadband ultraviolet (UV) radiometers is considered and categorized in this paper, based on the concepts of effective irradiance and effective irradiance responsivity. Practical calibration set-up and examples of uncertainty evaluation are presented as well.

Primary standard detectors based on an ionizing chamber and a pneumatic Golay detector have been developed at VNIIOFI to cover the spectral range 30 nm to 400 nm. The double ionizing chamber was used over the spectral range 30 nm to 126 nm and the Golay detector to cover the range from 105 nm to 400 nm. The main advantage of the Golay detector is its high level, non-selective responsivity of 10³ V W⁻¹. The responsivity threshold is 10⁻¹⁰ W Hz^−1/2. Pulsed synchrotron radiation sources were used for UV beam generation. These small synchrotrons have strong magnetic fields and emit a high intensity, controllable spectrum. A transfer standard multichannel radiometer was used for the measurement of UV-A, UV-B and UV-C irradiance, which permitted the determination of the bactericide, erythemous and generally harmful UV action spectra. The development of primary and transfer standard detectors allowed the establishment of basic national standards for UV radiometry.

The facility for automated spectroradiometric calibrations (FASCAL) is the primary facility for calibration of spectral irradiance and spectral radiance at NIST and has been in continuous use since the early 1970s. Due to the increasing demands for spectroradiometric calibration, especially for supporting the monitoring of global environmental changes, a new facility, FASCAL 2, dedicated to calibrating spectral irradiance has been built. This facility will enable faster responses to calibration requests and, ultimately, result in lower uncertainties in the disseminated spectral irradiances. The FASCAL 2 facility is designed with the objective of achieving a signal-to-noise ratio exceeding 1000 : 1 from 250 nm to 2500 nm in a bandwidth of 4 nm to 8 nm when measuring a 1000 W FEL lamp at a distance of 50 cm with a receiving aperture of 1 cm². The facility will also be capable of calibrating deuterium lamps from 200 nm to 400 nm. The facility has six independent source stations, with four of the stations dedicated to measurements with spectral irradiance lamps and two stations reserved for the realization of spectral irradiance scales and checking the accuracy of automated wavelength. After verifying that the calibrations of spectral irradiance performed in FASCAL and FASCAL 2 agree within their combined uncertainties, FASCAL 2 will become the primary NIST facility for calibration of spectral irradiance.

At the Physikalisch-Technische Bundesanstalt (PTB) silicon diode-based narrow-band interference filter radiometers (FR) are used for the radiometric determination of thermodynamic temperatures. To extend the temperature range down to 400 °C, the centre wavelength of the FRs was shifted from the visible to the near infrared (900 nm, 1000 nm and 1595 nm). An improved calibration procedure is presented which achieves relative uncertainties for the spectral responsivity of the FRs of the order of 10⁻⁴, by taking into consideration the increased temperature coefficient of the spectral responsivity and the nonlinearity of the silicon detectors when operated in the near infrared wavelength region.

An intercomparison of radiation temperature measurements was performed at VNIIOFI during October 2000 using a pyrolytic graphite blackbody operating over the temperature range from 1600 K to 3300 K. A pyrometer and two photometers from VNIIOFI, a pyrometer and four broadband glass filter detectors from PTB, and two narrow-band interference filter based radiometers and a broadband glass filter radiometer from NPL were used to perform the temperature measurements in either radiance or irradiance mode. Across almost the entire temperature range the VNIIOFI, NPL and PTB instruments showed results within the combined standard measurement uncertainties.

This intercomparison of spectral responsivity between the National Institute of Standards and Technology (NIST) in the United States and the National Research Council (NRC) of Canada was carried out to test the level of agreement between the two laboratories for routine detector calibrations and also to support the uncertainties quoted by NIST and NRC in appendix C of the Mutual Recognition Arrangement (MRA) drawn up by the International Committee of Weights and Measures (CIPM). The comparison was carried out in two stages: 250 nm to 1100 nm using three silicon (Si) photodiodes, then 700 nm to 1800 nm using two germanium (Ge) and two indium gallium arsenide (InGaAs) detectors. This paper describes briefly the derivation of the spectral responsivity scales, measurement procedures and apparatus used for each laboratory. The results of the intercomparison are presented. In the spectral ranges from 250 nm to 1000 nm for silicon and 750 nm to 1700 nm for germanium and InGaAs, the agreement between the two laboratories is well within the combined uncertainties. In particular, in the 450 nm to 1000 nm spectral range, the agreement is better than ±0.2%. In the 900 nm to 1600 nm spectral range, the agreement is typically ±0.5%. Consequently, the results of this intercomparison are shown to support the uncertainties claimed by NIST and NRC in their routine calibrations, and in their appendix C entries of the MRA.

Narrow-band filter radiometers at 248 nm, 313 nm, 330 nm and 368 nm wavelengths were used to compare calibration facilities of spectral (irradiance) responsivity at HUT, NPL and BNM–INM. The results are partly in agreement within the stated uncertainties. Use of demanding artefacts in the intercomparison revealed that the wavelength scales of the participating institutes deviate more than expected. Such effects cannot be seen in typical intercomparisons of spectral responsivity or spectral transmittance, where spectrally neutral samples are used.

The National Institute of Standards and Technology (NIST) and the National Physical Laboratory (NPL) realize independent scales for regular transmittance and near-normal regular reflectance in the mid-infrared part of the spectrum. Comparisons of these scales have recently been completed and the results are reported here. The agreement was excellent, lying within the quadrature combined uncertainties for the great majority of values measured and within the simple sum of uncertainties in all cases, demonstrating the level of equivalence of the NIST and NPL scales.

Experience with the two types of radiometers in the experiment VIRGO on the ESA/NASA mission SOHO launched in December 1995 has revealed important information about the long-term behaviour of such radiometers in space. Exposure-dependent changes can be determined from comparison of radiometers with different periods of exposure to the sun, as was done for, for example, the ACRIM series. The physical mechanisms for the long-term changes are not yet well understood. However, with the available data a model is developed which describes the temporal behaviour quite well and allows some inference of the mechanisms involved. The direct comparison of the two types of radiometers, after they have been individually corrected for the exposure-dependent part, reveals changes that are independent of the exposure to solar radiation. From this comparison an upper limit for the uncertainty of the long-term changes can be deduced. The uncertainty of the composite TSI depends mainly on the uncertainty of the tracing of ACRIM II to I, which amounts, together with the uncertainty of the HF correction during the gap between the two ACRIM experiments, to ±62 ppm (i.e. 62×10⁻⁶). Together with the result from a comparison with the independent record of ERBE the long-term uncertainty for the whole record from 1978 to present is estimated to ±85 ppm.

More quality assurance activities are needed to minimize uncertainties and improve equivalence of the UV radiation data obtained at different sites worldwide in order to monitor trends and to determine changes due to the depletion of the stratospheric ozone layer. Aspects of traceability based on a calibration chain from a cryogenic radiometer to the spectroradiometers within a UV monitoring network, as well as equivalence and conformity based on intercomparisons of standard lamps and instruments, are discussed. Some recent developments to be used for quality assurance/control in UV monitoring networks are presented, covering portable monitor radiators, monitor systems containing photodiodes and/or LEDs as well as the calibration transfer from the horizontal to the vertical optical axis.

The quantitative study of climate change over decades requires successive generations of satellite, airborne and ground-based instrumentation carefully calibrated against a common radiance scale. In NASA's Earth Observing System (EOS) program, the most common sources used in the laboratory radiance calibration of satellite, ground-based and airborne instruments operating in the reflective solar wavelength region of 400–2500 nm are integrating spheres and diffuse reflectance panels illuminated by irradiance standard lamps. Since 1995, the EOS calibration program operating within NASA's EOS Project Science Office (PSO) has enlisted the expertise of the National Institute of Standards and Technology (NIST), the University of Arizona Optical Sciences Center's Remote Sensing Group (UA), Japan's National Research Laboratory of Metrology (NRLM) and NASA's Goddard Space Flight Center (GSFC) in an effort to validate the radiance scales assigned to sources used in the prelaunch calibration of EOS instruments and to critically examine the operation, repeatability and stability of those sources. Radiance scale validation is accomplished using stable transfer radiometers operating at visible to shortwave infrared wavelengths and calibrated and characterized by each institution using a variety of techniques. In the 10 comparisons performed since February 1995, the agreement between the radiance measurements of these transfer radiometers is ±1.80% at 411 nm, ±1.31% at 552.5 nm, ±1.32% at 868.0 nm, ±2.54% at 1622 nm, and ±2.81% at 2200 nm (σ = 1).

Launching on the SORCE (SOlar Radiation and Climate Experiment), the total irradiance monitor (TIM) will measure the total solar irradiance to a relative standard uncertainty of 10⁻⁴ and a noise level of 2×10⁻⁶ each 400 s. Because of the outstanding brightness, uniformity, stability and collimation of the Sun, this ambient temperature primary radiometer achieves an absolute accuracy comparable to that of a cryogenic radiometer. The major paradigm shift from previous solar radiometers is the use of phase-sensitive detection at the shutter fundamental. We describe the equivalence between replacement power and radiant power as a complex number, the ratio of the (complex) thermal impedances. Our aperture measurements are provided by NIST, with added corrections for diffraction and scattering. We measure the cavity reflection losses versus wavelength using laser scans. We have verified methods for measuring, in flight, the servo gain, cavity reflection changes and shutter-modulated infrared offset (dark signal).

In this paper, we describe an algorithm to correct a spectrograph's response for stray light. Two recursion relations are developed: one to correct the system response when measuring broad-band calibration sources, and a second to correct the response when measuring sources of unknown radiance. The algorithm requires a detailed understanding of the effect of stray light in the spectrograph on the instrument's response. Using tunable laser sources, a dual spectrograph instrument designed to measure the up-welling radiance in the ocean was characterized for stray light. A stray-light correction algorithm was developed, based on the results of these measurements. The instrument's response was corrected for stray light, and the effects on measured up-welling in-water radiance were evaluated.

Because of its stable surface reflectance and its stable irradiance in the visible (VIS) and near-infrared (NIR) spectral regions, the Moon can be used as a reference source for monitoring the response stability of Earth-orbiting sensors. MODerate Resolution Imaging Spectroradiometer (MODIS) Protoflight Model (PFM), launched on-board the NASA Terra spacecraft on 18 December 1999, has been in operation for more than two years. Its reflective solar bands (RSB) are calibrated on-orbit by a solar diffuser (SD). In addition, MODIS on-orbit lunar views are used to characterize the stability of the RSB radiometric calibration and the response change at an angle of incidence (AOI) different from that of the SD. In this report, we describe an approach that properly selects the viewing conditions to minimize the effects that need to be corrected and its application to the MODIS lunar observations. Results, derived from MODIS lunar views, are reported and compared with the on-board SD observations.

The MODIS Protoflight Model, launched on-board the NASA EOS Terra spacecraft on 18 December 1999, has been in operation for more than two years, providing the science community with calibrated data sets for global studies of the Earth's land, oceans and atmosphere. MODIS has 36 spectral bands, covering the spectral range from 412 nm to 14 200 nm, and provides spatial resolutions of 0.25 km (two bands), 0.5 km (five bands) and 1 km (29 bands) at nadir. The key on-board calibrators (OBCs) include a solar diffuser (SD) and a solar diffuser stability monitor (SDSM) system for calibration of the reflective solar bands, and a V-grooved flat-panel blackbody (BB) for calibration of the thermal emissive bands. In this paper, we describe the use of OBCs for the sensor's radiometric calibration and characterization and discuss on-orbit performance. In addition, we provide an assessment of the on-orbit degradation of the SD and MODIS optics.

This report describes the first Sensor Intercomparison and Merger for Biological and Interdisciplinary Oceanic Studies (SIMBIOS) Radiometric Intercomparison (SIMRIC-1). The purpose of the SIMRIC-1 is to ensure a common radiometric scale among the calibration facilities that are engaged in calibrating in situ radiometers used for ocean colour-related research and to document the calibration procedures and protocols. SIMBIOS staff visited the seven participating laboratories for at least two days each. The SeaWiFS Transfer Radiometer (SXR-II) measured the calibration radiances produced in the laboratories. The measured radiances were compared with the radiances expected by the laboratories. Typically, the measured radiances were higher than the expected radiances by 0 to 2%. This level of agreement is satisfactory. Several issues were identified where the calibration protocols need to be improved, especially the reflectance calibration of the reference plaques and the distance correction when using the irradiance standards at distances greater than 50 cm. The responsivity of the SXR-II changed from 0.3% (channel 6) to 1.6% (channel 2) from December 2000 to December 2001. Monitoring the SXR-II with a portable light source showed a linear drift of the calibration, except for channel 1, where a 2% drop occurred in summer.

Highly polished plane gloss standards are indispensable when using a glossmeter. The gloss value of a primary reference standard is defined by its refractive index at 589.3 nm. This definition allows ease of use, but the calibration accuracy is limited to about ±0.5% due to inconsistency between the monochromatic definition and the polychromatic use in the glossmeter. Detector-based radiometry has improved a lot in the last 20 years and it is now time to think about changing the primary gloss standard calibration to absolute specular reflectance-based measurements. This paper is a step into this direction. It shows a refractive index fitting method for gloss standards based on gonioreflectometric specular-reflection measurements, Fresnel's equation and least-squares fitting. The refractive index fitted data showed ±0.23% agreement on average with the gonioreflectometric data measured in 1° steps and ±0.07% agreement on average for the angular corrected data. The requirement for an uncertainty of ±0.1% in the gonioreflectometric calibration was also analysed.

Electron storage rings are nearly ideal radiation sources for radiometry over a broad spectral range from the infrared to the x-ray region. Many national metrology institutes make extensive use of synchrotron radiation for radiometry. The Synchrotron Ultraviolet Radiation Facility SURF III, which has recently been upgraded, is used by the National Institute of Standards and Technology (NIST), and the Physikalisch-Technische Bundesanstalt (PTB) operates a European radiometry laboratory at the third-generation storage ring BESSY II. The storage rings are used as calculable broadband radiation sources (primary source standards) and in combination with primary detector standards as sources of continuously tunable monochromatic radiation for detector calibration. The use of cryogenic electrical-substitution radiometers as primary detector standards has made it possible to considerably extend high-accuracy calibrations from the ultraviolet (UV) to the x-ray regime. Radiometry in the extreme UV (EUV) range, and in particular reflectometry, has become increasingly important for the development of EUV lithography. An overview of synchrotron radiation based radiometry will be given.

After the shutdown of the BESSY I electron storage ring, PTB's two normal-incidence monochromator beamlines for the calibration of radiation sources and detectors in the wavelength range from 40 nm to 400 nm were transferred to PTB's new radiometry laboratory at BESSY II. In this paper, the beamlines and their properties are briefly described. First results for the calibration of secondary source standards and detector standards at BESSY II are shown and demonstrate the successful re-establishment of the measurement capabilities. In the ultraviolet, calibration of deuterium lamps was reproduced within 1% and of silicon trap photodetectors within 0.2%.

Subcomparison CCPR-K1.b of the CCPR key comparison of spectral irradiance covering 200 nm to 400 nm will be carried out in 2002/03 piloted by the Physikalisch-Technische Bundesanstalt (PTB) in Braunschweig, Germany. Deuterium arc lamps will be used as transfer standards of spectral irradiance, although they are known to have, in general, poor stability and reproducibility. PTB investigated a variety of deuterium lamps to test their capability (stability and reproducibility) for use as transfer standards. A measurement facility capable of characterizing up to three lamps at a time has been set up. Investigations of the spectral ageing pattern will help to interpolate between different reference measurements to markedly reduce the influence of ageing effects. The application of an external SiC photodiode as a monitor detector was tested. Spatial characterization of the irradiance of each lamp was also carried out. All this all helped us to find suitable types of lamp, producing reasonable irradiance homogeneously enough to be non-sensitive to small misalignments.

The electron storage ring TERAS was used to calibrate the spectral radiance of transfer standard light sources in the ultraviolet (UV) and vacuum ultraviolet (VUV) regions. This calibration system, constructed at the beamline, is an improvement on a prototype system we constructed previously. The new system is simpler and does not require breaking the vacuum of the calibration system. Also, the signal from the transfer standard light source is strong, the conditions of the light sources (synchrotron radiation and transfer standard light source) can be observed directly (it is possible to evaluate the profiles or some parameters of light sources without breaking the vacuum), the polarimeter is improved and the uncertainty due to the polarization effect is expected to be improved. In addition to describing these improvements, we re-examine some uncertainties.

DXW-lamps are used to calibrate spectroradiometers measuring solar UV radiation. To study the ageing properties of DXW-lamps two pre-aged 1 kW lamps were burned for over 80 h, simulating the normal use of the lamps. The lamps were burned about 1.5 h at a time with a 30 min period between the consecutive burns to allow the lamps to cool down. Usually there were four burns in a day. The irradiance measured by a spectroradiometer decreased by about 0.055% h⁻¹ and the difference between 280 nm and 400 nm wavelengths was 1% at the end of the test. The measured spectra were analysed to find out the emissivities and the temperatures of the filaments. The emissivities were compared to the emissivity at the start of the test burns. The emissivity seemed to remain stable during the whole burning time. The spectral changes can be fully explained by the changes of the filament temperatures, which decreased by about 8 K during the 80 h burns. No absorption or emission lines were found in the spectra during the test burns.

A new source-based radiometry beamline was recently constructed at the Synchrotron Ultraviolet Radiation Facility (SURF III). The goal of this beamline is to establish a national source standard with a wide spectral range from the far ultraviolet to the infrared by using the calculability of SURF III. The new beamline is a straight-through white-light beamline with few, if any, optical components involved in delivering the radiation to the user end-station for applications such as source and irradiance intercomparisons. Due to the low tunable operating electron energy of SURF III (from 380 MeV to less than 100 MeV), the beamline is uniquely suited to work in the ultraviolet to infrared spectral range. In this paper we discuss the design of the beamline and present first results on the vertical angular distribution of the radiation determined with filtered radiometers in the ultraviolet, visible and near-infrared spectral ranges.

High-temperature fixed-point blackbodies based on Re–C and TiC–C metal–carbon eutectic alloys are being investigated for use as radiance and irradiance sources for precise measurements in radiometry, photometry and radiation thermometry above the conventionally assigned values of temperatures on the ITS-90 scale. Graphite crucibles having inner diameters varying between 4 mm and 10 mm were used to prepare the metal–carbon and metal carbide–carbon eutectics; the cells were designed and manufactured at VNIIOFI, Russia using high-purity materials.

The melting and solidification temperatures of the cells were measured. Their reproducibility was investigated. The radiance reproducibility of the Re–C and TiC–C fixed points was found to be from 0.01% to 0.03% at 650 nm wavelength depending on the cell. Preliminary investigations of ZrC–C fixed-point reproducibility have been carried out. The radiance of all measured cells agreed at the solidification point within 0.02%.

This paper considers the possibility of improving the accuracy of the predictable quantum efficiency method (self-calibration) by using custom photodiodes at cryogenic temperatures. The photodiode quantum deficiency is partitioned into nine terms associated with different phenomenological loss mechanisms. The size of each term is estimated for operation of the photodiode at 72 K and 16 V reverse bias. Requirements for high-accuracy self-calibration of the value and uncertainty of each of the quantum deficiency terms are discussed.

Infrared (IR) radiometric applications require the use of IR detectors having characteristics which are stable with time. Little has appeared in the scientific literature on the ageing effects of IR detectors. The purpose of this paper is to describe some effects that can be attributed to ageing and that were observed in some of the IR detectors studied, and to identify their origins. It was observed that some IR detectors exhibit stronger ageing effects than others and, furthermore, that the effects of some of the observed ageing processes can be reversed. Some ageing effects designated apparent ageing effects are not true manifestations of ageing but originate from the presence of atmospheric absorption bands. The measured apparent structure in the response of an IR detector arises due to changes in the conditions which prevail in the measurement facility, differences in path lengths during the characterization of the detectors, errors in the wavelength scale or radiation bandwidth used and, in some cases, due to nonlinearities in the response of some IR detection systems.

A difficulty in diverse metrological applications is the measurement of relatively low intensity ultraviolet radiation signals in the presence of an intense visible radiation background. The development of a detector technology that combines the high efficiency and ease of use found in Si photodiodes with the solar-blind characteristics of many photoemissive devices would offer an ideal solution to this problem. Wide bandgap semiconductors (e.g. SiC, GaN and AlGaN) should combine sensitivity comparable to Si photodiodes with solar-blind response. We present ultraviolet quantum efficiency measurements of AlGaN Schottky and GaN p–i–n photodiodes demonstrating responsivity in the far ultraviolet to wavelengths as short as 116 nm (AlGaN Schottky devices) or 180 nm (GaN p–i–n devices) and several orders of magnitude reduction in efficiency for photon energies below the bandgap. The bandgap for GaN corresponds to 365 nm wavelength, and AlGaN devices exhibit cut-off wavelengths as short as 260 nm.

Evaluation of five types of silicon photodiode was undertaken to verify their suitability for absolute radiometry and also for their use as transfer standards in the spectral region from 1 nm to 1100 nm. Four types of photodiode were fabricated for this study; these were the p-on-n photodiode, n-on-p photodiodes with silicon dioxide front windows and n-on-p photodiodes with a metal-silicide front window. Fabrication of photodiodes with 100% internal quantum efficiency is demonstrated and their necessity for making absolute radiometric measurements with the lowest possible uncertainty is pointed out. The linearity characteristics of these devices, as measured by the ac/dc method, are far superior to those of the p-on-n diodes especially fabricated for this work and also to those exhibited by p-on-n diodes widely used at present by the radiometric community. Results on the stability of the quantum efficiency of the fabricated diodes after exposure to intense radiation of 13 nm, 120 nm, 157 nm, 193 nm and 254 nm radiation will also be presented. Photodiodes with a metal-silicide front window were the only devices stable when exposed to the intense beams of third-generation synchrotrons and UV excimer lasers.

A device composed of commercial optical fibre components and based on a flux-addition method has been designed for the measurement of the linearity of near-infrared detectors over six decades with uncertainties of less than ±0.04%.

Non-linear behaviour of commercial InGaAs photodiodes is presented and discussed. Detector non-linearity figures are shown with overfilled illumination, while in underfilled illumination all the photodiodes tested show a very good linearity figure.

In radiometry or pyrometry, radiometers are often used to assign the spectral radiance or radiance temperatures of sources using ratios of signals which can differ by several decades. For performing ratios between such sources with low uncertainties, the linearity of the detectors used in the transfer radiometers needs to be characterized. The linearity of InGaAs photodiodes has been studied using the flux-addition method using a broadband infrared source with a visible-blocking filter. Using this technique, 18 InGaAs photodiodes from four different vendors were studied without spectral filtering in a broad wavelength region from 900 nm to 1700 nm with the diodes underfilled by the incident flux. The linearity of InGaAs photodiodes was determined within the range of photocurrents from 10⁻⁸ A to 10⁻⁴ A. All the InGaAs photodiodes demonstrated linearity from 10⁻⁷ A to 10⁻⁴ A within the expanded uncertainties of 0.08% (k = 2). The uncertainty in the linearity measurement below 10⁻⁸ A is increased due to the increased noise in the photocurrent arising from the feedback resistance of the transimpedance amplifier being greater than the shunt resistance of the photodiode.

When using semiconductor photodiodes in the VUV, photoelectrons emitted from the photodiode surface form an additional current circuit when electrons reach a vacuum chamber wall surrounding the photodiode. In the case where the rear electrode of the photodiode is grounded, and hence the sensing terminal of an electrometer is connected to the front electrode, the electrometer measures the superimposed current, that is, the photocurrent internally generated in the p–n junction plus the photoemission current. Experiments done for diffusion-type silicon photodiodes and Schottky-type GaAsP photodiodes in the wavelength range from 50 nm to 200 nm show that the photoemission contribution has peaks of about 10% and 25%, respectively, of the internal photocurrent at around 100 nm for both without an extraction voltage. Therefore, for precise measurements of semiconductor photodiode photocurrents, the sensing terminal of an electrometer should be connected to the rear electrode of a photodiode to avoid any contribution from the photoemission current.

This paper presents results of investigations of the radiant power spectral responsivity within the near-infrared spectral range. Large-area windowless Ge photodiodes 10 mm in diameter used as single photodiodes, as well as in a trap configuration, were studied within a wide spectral range. The spectral responsivity of the reflection trap detector equipped with three photodiodes is compared with that of a single Ge photodiode from 600 nm to 2000 nm (all photodiodes of the same type). The temperature effect on the spectral responsivity is shown within the temperature range from 10 °C to 30 °C. The results are discussed in terms of spectral quantum efficiency. Effects of non-linearity are also discussed.

Correlations in spectral standards between values at different wavelengths should be considered when estimating uncertainties of spectral integrals. Simple correlations are introduced through fixing an absolute value to a relative distribution over wavelengths, or through systematic errors. Such correlations are easily treated when calculating effects that depend only on the relative distribution. A much more complex situation arises when a fitting process is used as part of the primary derivation. Two examples are given: that of spectral responsivity when the quantum defect of silicon photodiodes is fitted to measured data, and that of spectral irradiance where filter radiometers are used to calibrate a lamp directly and a fitted function is used for interpolation. In both cases, correlations between spectral points are significant and varied, and while these are diluted in the transfer from the primary scales, the effects may be significant in calculating correction factors in subsequent measurement where values at different wavelengths are combined.

The realization of a detector-based spectral irradiance scale at the National Institute of Standards and Technology is described. The new scale is established using filter radiometers calibrated for absolute spectral power responsivity traceable to the NIST High-Accuracy Cryogenic Radiometer unlike the previous NIST spectral irradiance scales based upon the gold freezing-point blackbody. The radiance temperatures of a high-temperature blackbody (HTBB) operating near 3000 K are found using calibrated filter radiometers. The spectral irradiances of a group of 1000 W FEL lamps are assigned using the spectral irradiance of the HTBB determined using the knowledge of the geometric factors and the detector-based radiance temperatures. The detector-based spectral irradiance scale leads to a reduction in the uncertainties from the previous, source-based, spectral irradiance scales by at least a factor of two in the ultraviolet and visible wavelength regions, and also leads to a reduction in the uncertainties in the short-wave infrared wavelength region by at least a factor of two to ten, depending on the wavelength. Following the establishment of the spectral irradiance scale in the early 1960s, the detector-based spectral irradiance scale represents a fundamental change in the way that the NIST spectral irradiance scale is realized, and beginning in the calendar year 2001, all spectral irradiance sources are issued using the detector-based scale.

Here we report on an alternative approach for making spectral irradiance measurements of radiation sources traceable, employing our monochromator-based absolute cryogenic radiometer (ACR) facility to its full extent. The method makes use of the continuously tunable absolute radiant flux emerging from the ACR facility to characterize and calibrate the spectral irradiance responsivity of a second adjacent double-monochromator system. This system will be part of a new facility at NMi-VSL, called SIR, that will be used to measure spectral irradiance distributions of radiation sources from the ultraviolet to the far infrared. Following this traceability route, a fully characterized high-temperature Planckian radiator is not needed. Proof-of-principle measurements in the visible part of the spectrum show very encouraging results. Although the new facility is under construction, the theoretical background and a schematic description will be presented in this paper.

A new spectral responsivity scale has been derived at Instituto de Física Aplicada (CSIC) in the range 400 nm to 800 nm. The scale is based on three single silicon photodiodes whose spectral responsivity was calibrated at several laser wavelengths against a trap silicon detector traceable to a cryogenic radiometer. Spectral responsivity values are obtained by using a physical model for the diode's reflectance and internal quantum efficiency. The uncertainty of the new responsivity scale is half the previous one. The new scale agrees with the CCPR-K2.b-PREV world mean.

The design and concept of a new integrating sphere for absolute diffuse reflectance measurement are presented, together with an evaluation of the reflectance non-uniformity. The new integrating sphere was designed with emphasis on uniform reflectance over the sphere wall. Spectralon^TM was employed as the inner material for the sphere wall. The integrating sphere with geodesic design was developed for the most appropriate specifications to realize the high reflectance uniformity. The evaluation of the reflectance non-uniformity of the sphere wall in the visible region revealed that the geodesic integrating sphere had a quite uniform reflectance with non-uniformity of about 0.16% on the average.

Absolute irradiance (and radiance) measurements are normally carried out with a radiometric bench consisting of two field apertures and a photodetector. Irradiance measurements are taken considering rectilinear propagation and are normally corrected from diffraction effects. In this work we treat irradiance measurements considering wave propagation and calculating the double diffraction. Within our proposal, the effects of diffraction, typically referred to as diffraction corrections, are handled naturally in the propagation of the cross-spectral density. This wave propagation allows us to calculate an instrument constant of the optical system, which for this case can be treated as a constant correction in irradiance as a function of distance. We show experimental results that verify our proposal.

The National Reference Reflectometer has been developed by the National Physical Laboratory (NPL) to realize scales for radiance factor and total diffuse reflectance, the latter scale being obtained through the spatial integration of goniometric measurements of reflectance. Analysis of its performance in the spectral region from 400 nm to 1000 nm shows that for spectrally neutral white materials such as matte white tiles and Spectralon^TM plaques the instrument is capable of realizing 0/45 radiance factor measurements with an uncertainty of 0.2% (k = 2) and 0/d total diffuse reflectance measurements with an uncertainty of approximately 0.25% (k = 2). The uncertainties depend on the variation of reflectance with angle, and there will therefore be some dependence on the material being measured. This instrument is now used to establish UK diffuse reflectance scales in the visible part of the spectrum.

The KRISS spectral radiance standards have been realized from 250 nm to 2.5 µm with reference to the radiance temperature scale. Based on the definition of the ITS-90, the temperature scale has been realized using a copper and silver point blackbody. This paper briefly describes the calibration facility, which consists of the variable temperature blackbody, the standard radiation pyrometer and the spectroradiometer. All of the radiating sources and the measurement system were fully controlled automatically and operated by two personal computers. The stabilities, noise characteristics, linearity and size of source effect of the spectroradiometer have been measured. Using two sheets of polarizer, the degree of linear polarization induced by the spectroradiometer and the Mueller matrix elements of the polarizer have also been simultaneously measured. For a tungsten strip lamp to be used as a spectral radiance transfer source, the current coefficient of spectral radiance, the stability in conjunction with the power supply and the degree of polarization have been measured to yield correction values at the reference condition after calibration. A mathematical model for the correction procedure and uncertainty analysis of the calibration is proposed. The relative standard uncertainty of the KRISS spectral radiance scale is determined to be 0.83% at 250 nm, 0.49% at 300 nm and less than 0.3% from 650 nm to 2.5 µm.

The importance of reliable spectral transmittance measurements for reliable thin-film characterization is demonstrated. By using a model of the spectral transmittance of thin-film samples, the effect of various uncertainty components in transmittance measurements on the determined thin-film parameters is analysed. The experimental results for aluminium dioxide and tantalum pentoxide thin-film samples show good agreement with the model.

This paper reports on the interference and diffraction effects produced by a double slit illuminated with parametric fluorescence light. We have evaluated, in a one-dimensional frame, the single and joint photon detection probability per unit of area, with respect to the position of a couple of detectors, far from the slits, in the so-called Fraunhofer region.

The realization of a number of radiometric and photometric units requires one to measure accurately the area of precision apertures. The BIPM employs the laser-beam scanning technique which is based on the generation of a uniform irradiance by a two-dimensional superposition of many regularly spaced laser beams.

The existing literature describes the formation of the uniform irradiance as a superposition of Gaussian laser beams and derives a condition for the largest allowable lattice spacing to obtain the necessary uniformity. In this paper we analyse the effect of the use of real laser beams which are almost, but not completely, Gaussian by applying Fourier analysis. Experimental results of aperture measurements with different spacings of the laser beams are also shown. The results of the calculations and the measurements show a characteristic dependence of the measured area on the grid spacing, even for spacings sufficiently small according to the condition based on Gaussian theory. This can be explained by very small deviations of the actual beam shape from a perfect Gaussian profile.

Spectralon^TM is a well known diffuse reflectance standard. Although ageing effects of Spectralon under high-level VUV and air UV irradiation have been reported, Spectralon is assumed to be stable under the normal radiation conditions used for calibration in standard facilities. In this paper, ageing effects of Spectralon under low-level irradiation and in the dark are presented. It has been found that ageing is mainly affected by four parameters: wavelength of radiation, irradiance level, radiant exposure (dose) and storage time of the samples. Therefore, the use of Spectralon^TM as a transfer standard in international comparisons and maintenance of reference values is discussed.

Comparison of independent measurement results allows an assessment of the accuracy of the underlying methods. In this work, four multichannel filter radiometers were calibrated using tunable laser-illuminated and lamp-illuminated integrating sphere sources (ISSs). The determination of the radiance of the laser-illuminated ISS is based on electrical substitution radiometry at cryogenic temperatures and dimensional metrology of circular apertures. The determination of the spectral radiance of the lamp-illuminated ISS is based on blackbody physics, referenced to the freezing temperature of gold. By calibrating the filter radiometers using both methods, we compare the `detector-based' to the `source-based' radiance scale at the National Institute of Standards and Technology (NIST).

Results of filter radiometer characterization with a wavelength-tunable Ti : sapphire laser in the wavelength band around 900 nm are presented. The effect of interference between the reflections from filter surfaces in the case of coherent laser light was studied and reduced with a special filter design with antireflection coatings. Measuring the responsivity as a function of wavelength over a very narrow band was used to reveal the remaining interference effects. Uncertainty analysis and test results indicate that filter radiometers can be characterized with a relative standard uncertainty of 9×10⁻⁴ using the scanning method. The results agree well with more conventional monochromator-based measurements.

During the past decade, high-accuracy soft x-ray radiometry has increasingly gained in importance. In particular, its routine availability is a prerequisite for the development of extreme UV lithography (EUVL). The PTB has been providing EUV calibration services for many years at a dedicated radiometry beamline at the electron storage ring BESSY I and from the year 2000 at the new PTB synchrotron radiation laboratory at BESSY II. Reflectometry, especially the measurement of the reflectance of Mo/Si multilayer mirrors for EUVL, is a major activity at the beamline. As the performance of the third-generation storage ring BESSY II is better than that of BESSY I, the measurement uncertainties could be significantly reduced. Efforts were made to accurately determine the wavelength scale for the monochromator and to suppress diffusely scattered light. For the spectral reflectance of a mirror in the EUV spectral region, a relative uncertainty of u = 0.14% is achieved. A new reflectometer will enable PTB to characterize EUVL optics up to 550 mm in diameter and 50 kg in mass. Detector calibration is based on a primary detector standard, a cryogenic electrical substitution radiometer. Photodiodes with a nearly constant spectral responsitivity in the soft x-ray range are calibrated as transfer detector standards with a relative uncertainty of u = 0.26% by direct comparison to the radiometer.

One of the most demanding applications of optics and optical coatings is in extreme ultraviolet lithography. Optical substrates must be polished to an accuracy of less than a nanometre, both the substrate and multilayer coatings must be atomically smooth, and the coatings must have a period that is controlled to a few tens of picometres. The metrology must have even greater capabilities to assure that the strict tolerances are met. Our extreme ultraviolet reflectometry facility has recently been furnished with the addition of a sample chamber that is capable of scanning the entire surface of a mirror up to 35 cm in diameter. The capabilities of this chamber will be described, as will the performance of the monochromator. Finally a plan will be presented to increase the stability of our monochromator in order to meet the needs of future generations of lithography tools.

Abel P: S93 Abu-Kassem I: S50 Andor G: S97 Anevsky S: S25 Arp U: S124

Barnes P Y: S172 Barnes R: S93 Barnes R A: S70 Barnes W: S85, S89 Boivin L P: S45 Bos H C D: S177 Brida G: S132, S204 Brown S W: S81, S216 Bruce S S: S45 Butler J J: S70, S154, S216

Campos J: S150, S181 Carter A C: S1, S13 Chang E K: S5 Chiang K: S89 Chow P: S141 Chunnilall C J: S55, S192 Clark D K: S81 Clarke F J J: S55 Cooper J: S93 Corredera P: S150, S181 Crane L: S192 Cunningham D: S145

Datla R: S1 Datla R V: S13 Davis C: S93 Deadman A J: S192

Eppeldauer G P: S154 Esposito J: S89

Fargion G: S93, S216 Feinholz M E: S81 Fischer J: S39 Flora S J: S81 Fowler J: S9 Fox N P: S39, S136 Friedrich R: S35, S39 Fröhlich C: S60 Frouin R: S93

Gardner J L: S167 Geist J: S132 Gibson C E: S30, S172, S216 Godin M: S93 Goebel R: S208 González-Herráez M: S150 Grantham S: S229 Guenther B: S85, S89 Gullikson E: S145

Hanssen L M: S55 Harder J: S78 Harrison N J: S39, S50 Hartmann J: S35 Hartree W S: S50 Heine R: S163 Hernanz M L: S150 Hertog B: S141 Hofer H: S163 Hollandt J: S35, S107 Hooker S B: S216 Höpe A: S212 Huang X: S21

Ikonen E: S17, S200, S220 Ivanov V: S25

Johnson B C: S81, S216 Jokela K: S17, S120 Jung T M: S1, S13

Kaplan S G: S55 Kärhä P: S17, S50, S120 Khlevnoy B B: S39, S128 Khromchenko V B: S39, S128 Kopp G: S78 Korde R: S145 Korwan D: S93 Koskela T: S17 Kroth U: S107 Kuznetsov V: S25

Larason T C: S154 Lawrence G M: S78 Lee I W: S196 Litorja M: S9 Lorentz S R: S1, S13 Loui H: S78 Lucatorto T B: S229 Lykke K R: S81, S124, S216

Maffione R: S93 Manoocheri F: S200, S220 Markham B: S216 McClain C: S93 McLean S: S93 Meister G: S93, S216 Menzies D: S93 Metzdorf J: S39, S66, S111 Minaeva O: S25 Mishima Y: S185 Möller W: S212 Morozov O: S25

Nawo B: S111 Nevas S: S50, S200 Nikolaus K-P: S212 Noorma M: S220 Novero C: S204

Ogarev S A: S39, S128

Park S N: S196 Parr A C: S124 Paustian W: S107 Pavlovitch M: S25 Pollard D F: S39 Pons A: S181 Poteau A: S93 Prince C: S145 Proctor J E: S30

Raatz K-H: S111 Rabus H: S107 Rastello M L: S132 Richter M: S107 Robertson J: S93 Rogers L J: S39 Rottman G: S78

Saito I: S115, S185 Saito T: S159 Samoylov M L: S39, S128 Sapritsky V: S25 Sapritsky V I: S39, S128 Saunders R D: S124 Scholze F: S224 Schrama C A: S177 Seidel J: S39 Seo J C: S196 Shaw P-S: S124 Sherman J: S93 Shin D J: S196 Shirley E L: S5 Shitomi H: S185 Smart M P: S55 Smith A: S1 Smith A W: S13 Sperfeld P: S39, S66, S111 Sperling A: S66 Stock K D: S66, S111, S163 Stock M: S208 Stolyarevskaya R I: S39 Suarez-Romero J G: S189 Sun J: S85

Tarrio C: S229 Taubert D R: S35 Tepichín-Rodríguez E: S189 Theocharous E: S136 Thornagel R: S107 Toivanen P: S220 Tümmler J: S224

Ulm G: S101, S107, S224 Usadi E: S192

van der Ham E W M: S177 Vernyi A: S25 Vest R E: S141, S145

Winter S: S66 Wittchen T: S66 Woods T: S78

Xiong X: S85, S89 Xu G: S21

Yarbrough M A: S81 Ylianttila L: S17, S120 Yoon H W: S30, S124, S154, S172

Zama T: S115 Zolotarevsky Y: S25

In partnership with Institute of Physics Publishing (IOPP), Metrologia embarks on a new phase of its development with the publication of volume 40. This first issue publishes the proceedings of the NEWRAD 2002 conference. In the past, NEWRAD special issues have typically taken of the order of fifteen months or so from the time of the meeting to appear in print. The fact that in this case the time to publication is less than nine months is testimony to one of the many advantages that IOPP can bring to the publication of the journal, and in the future even shorter times are envisaged.

Readers will notice that the composition of the Editorial Board has changed. I should like to take this opportunity to thank the outgoing members of the Board for the valuable input they have provided over the years, generously contributing their time and expertise for the advancement of the journal. At the same time I extend a welcome to the new members, confident in the expectation that they will use their wide knowledge and experience to ensure that the high standards of Metrologia are maintained.

As the new Chairman of the NEWRAD Committee, I am delighted to introduce this special issue of Metrologia that records the proceedings of NEWRAD 2002, which was held at the National Institute of Standards and Technology, Gaithersburg, Maryland on 20-24 May 2002. As in previous conferences, there was a range of contributions with a strong emphasis on applications, and the Conference provided the usual valuable opportunity to listen to the needs of user communities. As a result of resource constraints, National Metrology Institutes and other standards bodies are, these days, targeting their activities more and more onto topics of the highest priority. Events such as NEWRAD therefore provide a unique opportunity to gather the latest ideas, to assess the state of the art and to help assign priorities. As in the past, it also was an occasion at which collaborations between various communities were strengthened and during which much exchange of information and ideas could take place. The convivial atmosphere created by Al Parr and his NIST colleagues did much to stimulate this interchange and we are grateful to them for their efficient local organization.

The change of Committee Chairman also provides me with an opportunity to offer a few remarks on the departure of Terry Quinn from the BIPM. One of the founders of NEWRAD, Terry has, of course, made major personal contributions to the field and has championed this conference since its first meeting in 1985. We owe him a great debt of gratitude and wish him a long and happy retirement. For my part, following Terry as Director of the BIPM and as Chairman of your committee is a huge challenge. I thoroughly enjoyed my period as President of the Consultative Committee for Photometry and Radiometry and look forward to continuing my association with the community at future NEWRAD meetings.

We plan to return to Davos for the next NEWRAD in late September/early October 2005 and are grateful to Werner Schmutz of the World Radiation Centre for his help in the arrangements. He may be contacted by telephone on +41 (081) 417 5100 and by e-mail at w.schmutz@pmodwrc.ch. I hope to see you there!

Previous NEWRAD meetings and, before that, the meetings on 'New developments and Applications in Optical Radiometry' were at:

• 1985: Cambridge, Massachusetts;

• 1988: National Physical Laboratory, Teddington;

• 1990: World Radiation Centre, Davos;

• 1994: Physicalisch Technische Bundesanstalt, Berlin; and

• 1997: Optical Sciences Centre, University of Arizona.