Annual solar spectral energy distributions in North America

https://doi.org/10.1016/j.polymdegradstab.2020.109380Get rights and content

Highlights

  • Average annual solar spectral energy distributions (SEDs) were calculated based on multi-year spectrophotometric data.

  • Measured data could be fit using spectra calculated using SMARTS 2.9.5 to give the solar SEDs from 280 to 4000 nm.

  • SEDs were calculated for Miami, Florida; Phoenix, Arizona; Edgewater, Maryland; Madison, Wisconsin; and Regina, Saskatchewan.

  • Good agreement was found between independently-obtained datasets and between calculated and measured average broadband radiant energy for the sites.

Abstract

Predicting outdoor service life of polymers and coatings requires knowledge of the spectral power distribution (SPD) of the sunlight and the amount of radiant energy they receive. Standard solar SPDs such as ASTM G173 and ASTM G177 are snapshots of the SPD at a single moment under model conditions. A better measure would be average annual spectral energy distributions (SED) for specific locations, especially benchmark outdoor weathering sites. The annual SED is the energy received at each wavelength over the course of a year. A Smithsonian/NIST collaboration placed SR-18 UV spectroradiometers in Miami, FL, Phoenix, AZ, and Madison, WI during the period 1997-2012, collecting data at ca. 2 nm intervals between 290 and 324 nm. Similar data have been collected at Edgewater, MD for many years. The data are not complete for any year, but gaps could be patched using data from other years. These data sets were compared to data available on-line from ongoing U.S Department of Agriculture (USDA)/Colorado State University outdoor spectroradiometer measurements, after similar patching. The USDA data have fewer points in the UV but extend into the visible portion of the spectrum. The data sets give consistent results and show that the SEDs can be fit by SPDs calculated using the SMARTS 2.9.5 model and satellite ozone data. The annual standard deviations are < 5% for wavelengths > 310 nm. SEDs for average annual radiant energy received by horizontal and 45° south exposures are reported for benchmark sites near Miami and Phoenix. The paper provides a general approach for determining an SED for a location based on SMARTS modeling of the autumnal equinox and subsequent scaling based on reported annual total solar radiant energy.

Introduction

Solar energy exposure is a key factor in polymer degradation during outdoor weathering. Standard solar spectral power distributions (SPD) have been created for several industrial applications such as those in ASTM G173 [1], ASTM G177 [2], ISO/TR 17801 [3], and CIE Document 85 [4]. These standards generally show the maximum irradiance at the defined conditions, essentially summer high noon. However, articles exposed outdoors receive sunlight through all daily and seasonal solar angles, including attenuation by clouds. Solar radiation models such as SMARTS 2.9.5 [5] can calculate the clear-sky solar SPD for any location, time, and orientation. Hourly spectra can then be calculated and averaged to account for solar angles and seasons, but the models do not usually have any provision for effect of clouds on reducing solar irradiance. Service life prediction and predictive accelerated weathering require something different: the spectral energy distribution (SED) incident on a surface over time. An SED is a measure of the amount of solar energy received at each wavelength summed over a period of time to account for all the solar angles and attenuation by clouds. However, no annual solar SED has been published in the literature.

Several projects are ongoing in North America to record the solar SPD every few minutes over the course of many years. The Smithsonian Environmental Research Center (SERC) measured spectra at 18 wavelengths from 290 nm to 324 nm for over a decade at sites in Florida, Arizona, and Wisconsin in collaboration with the National Institute of Standards and Technology (NIST), and has an ongoing measurement program in Maryland. A portion of this data is available for public download [6]. The US Department of Agriculture (USDA) also has a long-running program measuring ultraviolet (UV) and visible radiation at 13 wavelengths at more than 35 sites throughout North America. All of the USDA data are available for download [7]. In principle, obtaining annual solar SEDs is a simple matter of converting the irradiance data to energy and summing over a year. In practice, the spectroradiometers invariably have down time so there are gaps in the data ranging from hours to months in any given year. In addition, the data are not compiled in any way, making it difficult to use.

The objectives of this analysis were: 1) to compile spectroradiometer data near benchmark outdoor exposure sites as well as higher latitude sites in North America; 2) to fit the data to create average annual spectral energy distributions using calculated spectra to fill the gaps; and 3) to evaluate the year-to-year variability in the SEDs.

Section snippets

USDA data

Descriptions of the instrumentation, sites, and data handling are available at the USDA UV-B Monitoring and Research Program website, https://uvb.nrel.colostate.edu/UVB/, as well as in publications [8,9,10]. UV data were acquired with 7 channel Multi-Filter Rotating Shadowband Radiometers (MFRSR) using interference filters with nominal 2-nm full-width at half-maximum (FWHM) bandwidths centered at 300, 305, 311, 317, 325, 332, and 368 nm. The visible-near IR data were acquired on a separate

Results

The Supporting Information includes extensive tables showing all the data compiled by month and year using Method 1, each patched year by Method 2, the input cards used for SMARTS 2.9.5 calculations, fitted spectral energy distributions 280 nm to 4000 nm, and the compiled satellite ozone data used for the calculations. This information should allow the interested reader to reproduce the full spectral energy distributions using SMARTS 2.9.5 for any month at any of the 5 sites.

SMARTS 2.9.5 fits to the data

The purpose of making the calculated fits to the data were 1) to find a rational way to “connect the dots” and 2) to allow calculations of integrated energy for comparison to measured data. There was no rationale or theoretical justification for finding a method other than obtaining a good fit. The average hourly solar SPD for the autumnal equinox provided a remarkably good fit to the annual average data for all sites except for Regina. This is reasonable given that the equinox provides an

Conclusions

Incomplete solar spectrophotometric data can be patched and averaged to give useful solar spectral energy distributions. The SERC and USDA data sets show good agreement with the SERC data on average 2-5% lower than the USDA data for nearby sites. Standard deviations of annual averages generally are < 5% for wavelengths > 310 nm. Annual variations in stratospheric ozone probably account for the high variability in wavelengths in the UVB band. However, the UVB constitutes only a small fraction of

CRediT authorship contribution statement

James E. Pickett: Methodology, Data curation, Writing - review & editing. Patrick J. Neale: Data curation, Writing - review & editing. Jacob P. Pickett: Data curation, Formal analysis, Writing - review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

We thank the USDA/Colorado State University team for access to their extensive data sets of solar spectral data. We thank Becky Olson at Colorado State University for downloading the Regina, SK data for us. The “NASA” data in Table 5 were obtained from the NASA Langley Research Center (LaRC) POWER Project funded through the NASA Earth Science/Applied Science Program. We especially thank Dr. Chris White at NREL for making the SERC data available on-line and facilitating the contacts that made

References (30)

  • J.E. Pickett et al.

    Global weathering of engineering thermoplastics

    Polym. Degrad. Stabil.

    (2005)
  • ASTM G173-03, Standard Tables for Reference Solar Spectral Irradiances: Direct Normal and Hemispherical on 37° Tilted...
  • ASTM G177-03, Standard Tables for Reference Solar Ultraviolet Spectral Distributions: Hemispherical on 37° Tilted...
  • Plastics — Standard table for reference global solar spectral irradiance at sea level — Horizontal, relative air mass 1

    (2014)
  • CIE 085-1989, Solar Spectral Irradiance, Commission Internationale de...
  • C.A. Gueymard, SMARTS 2.9.5, Simple Model of the Atmospheric Radiative Transfer of Sunshine,...
  • https://www.nist.gov/el/materials-and-structural-systems-division-73100/solar-data; accessed Feb. 10,...
  • https://uvb.nrel.colostate.edu/UVB/uvb-dataAccess.jsf; accessed Feb. 10,...
  • D.S. Bigelow et al.

    The USDA ultraviolet radiation monitoring program

    Bull. Am. Meteorol. Soc.

    (1998)
  • W. Gao et al.

    An ultraviolet radiation monitoring and research program for agriculture

  • K. Lantz

    The 1997 North American interagency intercomparison of ultraviolet spectroradiometers including narrowband filter radiometers

    J. Res. Natl. Inst. Stand. Technol.

    (2002)
  • E. Early

    The 1995 North American interagency intercomparison of ultraviolet monitoring spectroradiometers

    J. Res. Natl. Inst. Stand. Technol.

    (1998)
  • P. Neale et al.

    The Smithsonian-NIST-USDA-FPL network for monitoring solar ultraviolet irradiance: Comparison of radiometer measurements and radiative transfer model calculations

  • https://serc.si.edu/research/laboratories/photobiology-and-solar-radiation/uv-radiation-and-climate-monitoring,...
  • https://www.esrl.noaa.gov/gmd/grad/neubrew/SatO3DataTimeSeries.jsp, accessed Feb. 11,...
  • Cited by (1)

    • Biochemical and photochemical mechanisms that produce different UV-induced mutation spectra

      2021, Mutation Research - Fundamental and Molecular Mechanisms of Mutagenesis
      Citation Excerpt :

      Is this a relevant dose to simulate natural sunlight? According to the recorded sunlight spectra that were measured in Florida and Arizona in the US [45], and another reference spectrum (https://www.astm.org/Standards/G177.htm), the energy of 300+/- 5 nm wavelength constitutes approximately 0.15−0.22 % of total UV(A + B) radiance on the ground. Therefore, 5 kJ/m2 of 300+/- 5 nm light is equivalent to 2,300−3,300 kJ/m2 of total UV(A + B; 280−400 nm) radiance, which is roughly equivalent to 2.8–4.2 days of sunlight in those geographical regions.

    View full text