Indoor calibration of Sky Quality Meters: Linearity, spectral responsivity and uncertainty analysis

https://doi.org/10.1016/j.jqsrt.2016.03.015Get rights and content

Highlights

  • Verification of the linear response of Sky Quality Meters from 6 to 19 mag/arcsec2 based on a novel indoor procedure.

  • Detailed uncertainty calculation.

  • Measurement intercomparison between 16 instruments.

  • Spectral responsivity analysis.

Abstract

The indoor calibration of brightness sensors requires extremely low values of irradiance in the most accurate and reproducible way. In this work the testing equipment of an ISO 17025 accredited laboratory for electrical testing, qualification and type approval of solar photovoltaic modules was modified in order to test the linearity of the instruments from few mW/cm2 down to fractions of nW/cm2, corresponding to levels of simulated brightness from 6 to 19 mag/arcsec2. Sixteen Sky Quality Meter (SQM) produced by Unihedron, a Canadian manufacturer, were tested, also assessing the impact of the ageing of their protective glasses on the calibration coefficients and the drift of the instruments. The instruments are in operation on measurement points and observatories at different sites and altitudes in Southern Switzerland, within the framework of OASI, the Environmental Observatory of Southern Switzerland. The authors present the results of the calibration campaign: linearity; brightness calibration, with and without protective glasses; transmittance measurement of the glasses; and spectral responsivity of the devices. A detailed uncertainty analysis is also provided, according to the ISO 17025 standard.

Introduction

Public and private street lighting, among many other sources of light, have extended impressively both in small villages and in large cities of the most developed and growing countries in the last decades and have inevitably changed our lifestyle, together with the night skyline. It is well-known [1] that a variety of adverse health effects may be caused by light pollution (or photopollution), which is defined as the alteration of light levels in the outdoor environment (from those present naturally) due to man-made sources of light [2]. The effect is now recognized worldwide, thanks to the activity of associations like International Dark-Sky Association (IDA [3]) and thanks to the last improvements in satellite imaging for defense and meteorological applications [4], [5].

Light pollution is particularly strong in small and densely populated areas such as Southern Switzerland, as highlighted in Fig. 1. The cooperation between Dark-Sky Switzerland, the Sezione della Protezione dell׳Aria dell׳Acqua e del Suolo (SPAAS, the department of air, water and land protection) of canton Ticino and some Swiss astronomical observatories led in 2010 to the creation of a permanent network of monitoring stations all over the region (see Fig. 1c).

One of the main measurement unit in use to date for light pollution monitoring is the Sky Quality MeterTM (SQM) by Unihedron, a Canadian manufacturer specialized in portable tools for metrology in physics and astronomy [6]: 16 of them are now in use among the partners within the framework project OASI, the Environmental Observatory of Southern Switzerland. Measured data can be downloaded on OASI website [7].

Monitoring the measurement data is the mandate of the University of Applied Sciences and Arts of Southern Switzerland (SUPSI, in the Italian acronym), but after 5 years from the first installation, concerns have arisen among partners on the ageing and consequent misalignment of the instruments and on the possible drift of the aged instruments with respect of the new ones. The need of a reliable testing procedure in a controlled, indoor environment led to a measurement campaign held at the ISO 17025 accredited laboratory for photovoltaic terrestrial module testing at SUPSI, whose results are reported in this work.

The experimental equipment of the photovoltaic laboratory have been modified to allow for extremely low reliable level of irradiance for both total and spectral responsivity measurements: the equipment is described in Section 2. The results illustrated in this work confirm those previously published by Cinzano [8], while adding a detailed uncertainty calculation (Section 3) performed according to the requirements of ISO 17025 standard, and extending the measurement of spectral responsivity above the 800 nm limit, where an unexpected non-zero response of the SQM is here reported.

The work will serve as a useful tool and best-practice guideline to research institutes testing similar devices for sky darkness monitoring.

Section snippets

The Sky Quality Meter by Unihedron

Fig. 2a shows a picture of the sixteen SQM testing instruments.

The testing instruments give measurements of brightness as magnitude per steradians (mag/arcsec2), a dimensionless unit proportional to the negative exponential of brightness: thereafter the brighter an object appears in the dark sky, the lower is the measured magnitude. The detector is a crystalline silicon (c-Si) photodiode TSL237 by Texas Advanced Optoelectronic Solutions (TAOS), electrically connected to a current-to-frequency

Uncertainty analysis

The uncertainty calculation for measurements of the relative magnitude m on ORIEL 81172 as described in Section 2.2 was performed according to the ISO 17019 requirements that can be found in the ISO-IEC Guide to the Expression of Uncertainty in Measurement [12].

Uncertainty contributions are classified as type A (method of evaluation of uncertainty by the statistical analysis of series of observations) or type B (other methods, e.g. manufacturer datasheet, scientific literature, reasonable

Linearity

The linear response of the 16 SQMs was tested, measuring the increasing magnitude at decreasing low irradiance levels of Table 2, from 250 μW/cm2 to 0.79 nW/cm2, corresponding to magnitude values from 6 to 19 mag/arcsec2.

The deviation from the linear regression of data was then calculated and plotted as a function of magnitude: results showed standard deviations from the linear regression ranging from 0.11 mag/arcsec2 to 0.14 mag/arcsec2. Fig. 11 reports the linearity test of three different

Conclusions

The results of an indoor calibration campaign of 16 SQMs by Unihedron were presented, confirming the linear response of the instruments at extremely low irradiances at several orders of magnitude. An original method with a Class A solar simulator equipped with a set of ND filters (whose linearity was tested) was used, with a detailed uncertainty analysis highlighting the impact of various sources of measurement uncertainty.

The tested SQMs showed large deviations when measured at the same level

Acknowledgements

The authors would like to gratefully acknowledge the financial support of Repubblica e Cantone Ticino, within the collaboration with “OASI” (Grant number 1359 of the 18th March 2014). They would also like to thank the international partners of the Dark-Sky association for their precious comments on the results of this work.

References (12)

  • F.J. Verheijen

    Photopollution: artificial light optic spatial control systems fail to cope with. Incidents, causation, remedies

    Exp Biol

    (1985)
  • Hollan J. What is light pollution, and how do we quantify it? In: Dark-sky 2008 conference paper, Vienna; August 2008....
  • ...
  • C.D. Elvidge et al.

    Radiance calibration of DMSP-OLS low-light imaging data of human settlements

    Remote Sens Environ

    (1999)
  • P. Cinzano et al.

    The artificial night sky brightness mapped from DMSP satellite Operational Linescan System measurements

    Mon Not R Astron Soc

    (2000)
  • ...
There are more references available in the full text version of this article.

Cited by (22)

  • Measurements of sky brightness at Bosscha Observatory, Indonesia

    2020, Heliyon
    Citation Excerpt :

    The measured lamp brightness in mpass units decreased with increasing power to the lamp, in the same way that astronomical magnitudes become smaller as sources become brighter. We changed the brightness to a linear scale in nano-Lambert units and found a good linear response, with the goodness of fit corresponding to an R-squared value of 0.97 (Pravettoni et al., 2016). We connected the device to a laptop via a USB extension cable of about 5 m long.

  • Monitoring transition: Expected night sky brightness trends in different photometric bands

    2019, Journal of Quantitative Spectroscopy and Radiative Transfer
    Citation Excerpt :

    Our goal is to model the expected evolution of the night sky brightness, measured in different photometric channels and at different distances from the sources, along the transition process of substitution of traditional gas-discharge lamps by LEDs. To that end we consider several photometric detection bands of practical interest, including the photopic [49] and scotopic [50] spectral eye sensitivities, the specific photometric band of the Sky Quality Meter (Unihedron, Calgary, Canada) [51–52] used in several night sky brightness measurement networks across the world, and from which extensive dataset records exist [32–39], and the Day-Night Band of the Visible and Infrared Imaging Radiometer Suite (VIIRS-DNB) onboard the Suomi-NPP satellite [53–54]. Our approach is intentionally kept simple, considering only direct radiance propagation, averaged ground reflections and single-scattering processes, but it is expected to be reasonably comprehensive as to be useful for describing the principal traits of the expected light pollution evolution trends.

  • Light pollution offshore: Zenithal sky glow measurements in the mediterranean coastal waters

    2018, Journal of Quantitative Spectroscopy and Radiative Transfer
  • PePSS - A portable sky scanner for measuring extremely low night-sky brightness

    2018, Journal of Quantitative Spectroscopy and Radiative Transfer
    Citation Excerpt :

    Therefore, there is much interest and need for devices that can operate under distinct illumination levels. Sky Quality Meters (SQMs) are known to have roughly a linear response between 6 and 19 mag/arcsec2 [19] with a deviation from linearity smaller than 2.6% [4]. However, the spectral bandwidth of photodiode responsivity is as wide as some hundreds of nanometers, and the same also applies to quantum efficiency.

  • Experimentations on luminescent glazing for solar electricity generation in buildings

    2017, Optik
    Citation Excerpt :

    The use of these solar panels in the case of the distributed approach requires their installation on roofs or facades of buildings according to known technology of BIPV (Building Integrated Photovoltaic) or BAPV (Building Added Photovoltaic) [4–6]. The BIPV panels are quite expensive panels and do not allow a clear view of the surrounding landscape through the glass as either in the window or in opening roof (Sky Light) [7–9]. Furthermore, the orientation and tilt are critical to the performance index of the system made with BIPV and BAPV or Sky Light modules [10–12].

View all citing articles on Scopus
View full text