Solar Radiation Measurement and Solar Radiometers

Abstract

An instrument able to measure electromagnetic radiation, in its different forms and spectral ranges, is called a radiometer. This chapter focuses on the radiometers used for sensing solar radiation and on the measurements of different components and types of solar irradiance. As a simple classification, we will distinguish between broadband and spectral (narrowband) sensors. First, the fundamentals of physical sensors used to measure solar radiation are briefly described. Then, importance about calibration methods and uncertainty, as well as the structure of the traceability chain in the magnitude of solar irradiance, are presented. Next, solar radiometers and measurement techniques are described, starting from direct radiation in Earth’s surface, global irradiance in horizontal and tilted surfaces, diffuse irradiance, and finally another kind of radiation sensor. This structure is not casual but follows a path similar to that of the traceability chain, starting from the more accurate to the less accurate instruments. There are two additional sections devoted to the measurement of the spectral distribution of irradiance and to the measurement of aerosol contents in the atmosphere by using filter radiometers.

…too little interest is devoted to the calibration of instruments and quality of data. Measurements which are not reliable are useless.

WMO Technical note 172 (1981)

Appendix: Current Status of the WRR

1.1 1.1    The SI Laboratory Absolute Radiometers, the Space Radiometers, and the SI-WRR Conflict

As described in Sect. 2, the development of absolute cavity radiometers at the end of the decade of 1960, initially by the JPL and soon by other laboratories and commercial companies, provided at the end the foundation by the WMO of the WRR as the top reference standard in the solar irradiance scale, and the designation of PMOD (Davos) as the WRC where the group of WSG standards were conservated, maintained, and disseminated.

On the other hand, electrical substitution-based cavity radiometers continued evolving in the environment of the NMI laboratories, until the development of the cryogenic absolute radiometers (Quinn and Martin 1985). Unlike solar absolute radiometers, which work at ambient temperature, cryogenic radiometers work with reference temperatures in the cold reservoir between 2 and 20 K, by using liquid He and N. However, their application was first the determination of the Stephan–Boltzmann constant and of the thermodynamic temperature in radiation thermometry (Fox 2001).

Shortly after that, a primary standard radiometer was developed in the NPL (UK) for applications in optical radiometry (Martin et al. 1985). An improved design, by using a mechanical cooling engine to reach temperatures of 15 K, resulted in a compact instrument that became the standard to implement at NMI level the fundamental unit of candela (cd) and its derived units (lumen, lux) in the SI (Fox et al. 1995). Figure 31 shows a sketch of this kind of cryogenic radiometer. Additional reviews about the characteristics and operation of these absolute radiometers, their design, and their historical evolution can be found elsewhere (Hengstberger 1989; Fox and Rice 2005).

Fig. 31

Basic structure of a cryogenic radiometer for measurements of radiant power of lasers in NMI laboratories. Taken from Fox et al. (1995)

Due to the somewhat independent evolution of solar irradiance scale, based on the WRR, with respect to the SI optical radiometry scales at NMIs, based on cryogenic radiometers, intercomparisons between both scales were necessary and were done in a repeated form to determine their mutual transference and equivalence, and to check the stability of the results (Romero et al. 1991, 1995; Finsterle et al. 2008; Fehlmann et al. 2012). The intercomparison process is not immediate because of the differences between relative intensities of every scale and due to the different operation modes of the instruments, what forced the use of transfer standards (trap detectors) in some cases.

First, two comparisons (1991 and 1995) gave as result differences below 0.3% and within the respective uncertainty of each scale, what was considered as reasonable. The third comparison in 2005 produced an excessively low result and doubts about the linearity of the transfer detectors used were posed. In 2010, a new intercomparison was carried out, with results at the same level of the two first comparisons for measurements in power mode, but with differences of (0.34 ± 0.18)% between scales (WRR above SI) for measurements in irradiance mode. Discrepancies mainly arose due to the different modes of operation of the standards (irradiance versus power, light beam entirely covering or not the input port, see Fig. 32). Successive comparison along 2012 and 2014 confirmed differences of (0.31 ± 0.6)% in the ratio WRR/SI (Suter et al. 2012; Finsterle 2015) but without overlapping the difference between respective uncertainties, what leaves in question the transfer between scales.

Fig. 32

(Left) Results of WRR to SI intercomparisons. Taken from Fehlmann et al. (2012). (Right) different operation modes for radiometers of the SI and WRR scales (power versus irradiance). Taken from Suter et al. (2012)

The third group of absolute cavity radiometers, of great importance for this exposition, is that formed by solar radiometers used for the determination of total solar irradiance (TSI) and related quantities in successive space satellite and shuttle missions since the 1970s. Though their fundamental structure is quite similar to that of terrestrial absolute radiometers (TSI level of the order of ~1365 W m⁻² ± 3.5% of yearly oscillation), there are differences in two important working conditions in space: operation under vacuum (absence of air convection and atmospheric pressure effects) and operation at very low reference temperatures. Successive generations of instruments and space missions (e.g., NIMBUS7/ERB, SMM/ACRIM1, UARS/ACRIM2, SOHO/VIRGO, SORCE/TIM, ACRIMSAT/ACRIM3) have introduced progressive improvements in their design (Fröhlich 2013) and have contributed to the current recording of more than 35 years of TSI data. These space measurements allowed not only to determine the solar constant but also its natural variability in periodic 11-year cycles (corresponding to sunspots cycles), which is in the order of 0.1% (Yeo et al. 2014). However, data obtained from different experiments and instruments in space were not consistent with reference to their absolute irradiance values (see Fig. 33). The value of 1361.1 W m⁻² has been confirmed by a new revision with an estimated standard uncertainty of 0.5 W m⁻² (Gueymard 2018).

Fig. 33

Total solar irradiance recording by absolute radiometers in different space missions since 1978. Taken from www.acrim.com (as available in July/2018)

Differences were particularly enhanced when total irradiance monitor (TIM) radiometer (Kopp et al. 2005) went into operation in 2003 and measured values 0.35% lower than those of the variability of the solar irradiance and gravity oscillations (VIRGO) mission (Fröhlich et al. 1997). Besides new research into the origin of these differences, a new laboratory able to compare the twin reserve instruments (kept on Earth), the total solar irradiance radiometer facility (TFR) in the Laboratory for Atmospheric and Space Physics (LASP, Univ. Colorado, USA) was created (Kopp et al. 2007). This advanced installation allows the absolute radiometers to work both in power and irradiance modes, under good vacuum and well normal atmospheric pressure conditions, and it is then suitable to compare different types of instruments and irradiance scales. Thanks to TFR, it was checked how part of the differences found in space was due to the respective traceabilities of radiometers to WRR and SI scales. A new intercomparison WRR/SI with the PMO/PREMOS radiometer was carried out in TRF (Schmutz et al. 2013), and equivalent differences were found in the ground as in space (Fehlmann et al. 2012).

These results with space absolute radiometers contributed to demonstrate how WRR irradiance scale was out of concordance or equivalence to SI irradiance scale due to operating and functional differences between instruments and reference standards.

1.2 1.2    Current Status of WRR

Current lack of transference, equivalence, and/or compatibility between WRR and SI is being objected of an in-depth revision by WMO and CIPM, who agreed to cooperate to ensure that meteorological data could be adequately traced to SI. WRR is nowadays forming an “island of traceability” (Finsterle 2015), temporary out from SI, due to the WRR/SI ratio differences higher than 0.3% and the uncertainty associated to the comparison results.

Status of the WRR at a technical level is also delicate because many of the instruments originally integrating the WSG had to be ruled out of the group because of malfunctioning or drift. Currently, the WRR is implemented with at least four of the six surviving instruments, but some of them have cumulated more than 35 years of operation and can fail at any moment. Therefore, it is urgent the incorporation of new components to the WSG, or to search for new standard references, alternative to WSG, with enough precision, stability and low uncertainty, even by holding an irradiance scale based on artifacts.

A possible solution to the problem could be the use as a reference of a new absolute cavity radiometer called cryogenic solar absolute radiometer (CSAR) (Martin and Fox 1993), developed in collaboration between PMOD/WRC, METAS (Switzerland) and NPL (UK). CSAR bases their outdoor measurements in a supplementary unit, monitor for integrated transmittance (MITRA) (Walter et al. 2014), which is responsible for detect changes on window transmittance. As a whole, CSAR and MITRA present an impressive accuracy (150 ppm) in the determination of solar irradiance.

The operation of a radiometer with cryogenic temperatures allows the use of larger cavities, with enhanced absorptivity, and thanks to a reduction in thermal gradients in the cavity, it ensures the equivalence between radiative/thermal heating and electrical heating. However, these low temperatures require the use of vacuum for operating the cavities and to add an optical window whose spectral transmittance can change due to ambient temperature and intensity of received radiation. MITRA allows introducing corrections due to these factors in a synchronous form with CSAR measurements. As radiometer, CSAR can also work at ambient temperature without requiring cooling (Fig. 34).

Fig. 34

Picture of the WSG realizing the WRR, together with other cavity radiometers, and of the new CSAR and MITRA devices (left-lower side on the tracker). Taken at PMOD/WRR (Davos) in 2015

First functional probes of CSAR on the ground and first intercomparisons against cryogenic radiometers of SI scale laboratories seem to have given very promising results (Walter 2016) in terms of stability and traceability to SI. However, it is necessary to wait for the CIMO/WMO working group to decide what is the solution for the near future for the solar irradiance scale. The huge technical capabilities, complexity, and economical budget of an instrument like CSAR or of an installation as TFR-LASP do not seem to be easily expandable concepts to other NMIs in an extensive form.