pubmed.ncbi.nlm.nih.gov

Placing limits on long-term variations in quiet-Sun irradiance and their contribution to total solar irradiance and solar radiative forcing of climate - PubMed

Placing limits on long-term variations in quiet-Sun irradiance and their contribution to total solar irradiance and solar radiative forcing of climate

Mike Lockwood et al. Proc Math Phys Eng Sci. 2020 Jun.

Abstract

Recent reconstructions of total solar irradiance (TSI) postulate that quiet-Sun variations could give significant changes to the solar power input to Earth's climate (radiative climate forcings of 0.7-1.1 W m-2 over 1700-2019) arising from changes in quiet-Sun magnetic fields that have not, as yet, been observed. Reconstructions without such changes yield solar forcings that are smaller by a factor of more than 10. We study the quiet-Sun TSI since 1995 for three reasons: (i) this interval shows rapid decay in average solar activity following the grand solar maximum in 1985 (such that activity in 2019 was broadly equivalent to that in 1900); (ii) there is improved consensus between TSI observations; and (iii) it contains the first modelling of TSI that is independent of the observations. Our analysis shows that the most likely upward drift in quiet-Sun radiative forcing since 1700 is between +0.07 and -0.13 W m-2. Hence, we cannot yet discriminate between the quiet-Sun TSI being enhanced or reduced during the Maunder and Dalton sunspot minima, although there is a growing consensus from the combinations of models and observations that it was slightly enhanced. We present reconstructions that add quiet-Sun TSI and its uncertainty to models that reconstruct the effects of sunspots and faculae.

Keywords: quiet-Sun magnetic fields; radiative forcing of climate; total solar irradiance.

© 2020 The Author(s).

PubMed Disclaimer

Conflict of interest statement

We declare we have no competing interests.

Figures

Figure 1.
Figure 1.

Total solar irradiance (TSI) variations after 1995. (a) The SATIRE (Spectral And Total Irradiance REconstructions) ‘S’ reconstructions of two contributions to TSI: the integrated facular brightening (fb: daily values in pink and means over Carrington rotation (CR) intervals in mauve) and sunspot darkening (sd: daily means in grey, CR means in black) [15]. (b) The PMOD (Physikalisch-Meteorologisches Observatorium Davos) composite of TSI observations (TSIPMOD: daily means in cyan, CR means in blue) [10]. (c) The RMIB (Royal Meteorological Institute of Belgium) composite of TSI observations (TSIPMOD: daily means in grey, CR means in black) [11] splined with data from the TIM (Total Irradiance Monitor) on the SORCE (SOlar Radiation and Climate Experiment) satellite (daily values in orange, CR means in red) [16]. The join of the spline is the start of the TIM data on 25 February 2003 and made by re-calibrating the RMIB data using a constant offset quantified using the mean values of the two datasets over the first year of TIM observations. The CR means in green are PMOD data with the best-fit offset of 0.23 W m−2 used to fill in a gap in the SORCE data. (d) Comparison of CR means of two observation series: PMOD composite (in blue) and the RMIB/TIM spline (in red). (Online version in colour.)

Figure 2.
Figure 2.

Model reconstructions of TSI since 1600. The SEA11 reconstruction shown is that based on 10Be abundances from South Pole ice cores [30] and the EEA18 reconstruction is that based on a combination of various 10Be and 14C abundance records (referred to as PHI-US16 [31]). See text for details of other reconstructions. (Online version in colour.)

Figure 3.
Figure 3.

Analysis of the relative drifts in TSI observation series. All data are means over CR intervals and the difference between a given series (generic total solar irradiance TSIX) and the PMOD composite, TSIPMOD, is plotted as a function of date: OX,95 is the zero-level offset for the TSIX series that makes the linear variation fitted to the TSIX data equal to that for TSIPMOD on 1 January 1995. The orange squares are for the TIM data; the mauve circles are RMIB composite; the blue triangles are for data from the ACRIM-3 (Active Cavity Radiometer Irradiance Monitor) instrument on the ACRIMSAT spacecraft [12] and the green diamonds are for data from TCTE (Total Solar Irradiance Calibration Transfer Experiment), which operated from November 2013 until June 2019. The yellow circles are for the Community Composite, v. 1 (CCv01 [14]). The coloured lines are the best-fit linear variations to the data points and are surrounded by paler-shaded areas giving the 2 − σ uncertainty in that linear regression (see key). The rates of drift for these instruments compared with the PMOD composite are given in the bottom left of the plot. The plot also shows data points and best-fit linear regression for SOVAP (SOlar VAriability Picard) on the Picard satellite [17] as purple inverted triangles and a purple dot-dash line. The 2 years of data from the TSIS-1 (Total and Solar Irradiance Sensor) satellite are shown as stars. The black dashed line is for the PREMOS (PREcision MOnitor Sensor) instrument on Picard (operated for 2010–2013) from the analysis of Kopp [42,43]. (Online version in colour.)

Figure 4.
Figure 4.

(a) The variation in the heliospheric modulation parameter, Φ . The mauve dots are annual means from the analysis of data from neutron monitors, calibrated using direct measurements of the cosmic ray spectrum performed by the Payload for Antimatter Matter Exploration and Light-nuclei Astrophysics (PAMELA) space-borne spectrometer [46]. The black line is a second-order polynomial fit of the reconstruction of Usoskin et al. [44] based on cosmogenic isotope data and modelling using the open solar flux since 1868 derived from geomagnetic data by Lockwood et al. [35]. This has been extended to 2019 using a linear regression spline of observed counts from neutron monitors. (b) Running means of the Φ values shown in (a), 〈Φ〉τ, over the interval of duration τ years prior to the time in question. These are shown as an anomaly relative to the value of 〈Φ〉τ for 2019. Blue, orange and red lines are for τ of 11, 22 and 33 years, respectively. (Online version in colour.)

Figure 5.
Figure 5.

(a) The variation of modulation parameter Φ over 1995–2019: the green line shows daily means, 〈Φ〉1day, and the black line shows means over CR periods, 〈Φ〉CR. (b) Means of Φ over prior intervals of duration τ of 5.5 years (mauve line), 11 years (blue line), 16.5 years (green line) and 22 years (orange line). (Online version in colour.)

Figure 6.
Figure 6.

Scatter plots of the SATIRE-3D model results with the TIM observations. (a) The logarithms of the number of daily samples in bins are 0.05 W m−2 by 0.05 W m−2 in extent, colour-coded and plotted as a function of observed (TSIo) and modelled (TSIm) total solar irradiance. (b) Scatter plot of annual means of TSIm as a function of TSIo (mauve squares) and means over CR intervals (black dots). The dashed line shows TSIm = TSIo. (Online version in colour.)

Figure 7.
Figure 7.

Scatter plots of the SATIRE-3D model results with the PMOD data composite in the same format as figure 6. (Online version in colour.)

Figure 8.
Figure 8.

Variations of quiet-Sun irradiance, Q = TSIo−TSIm, for the 14 combinations of TSIo and TSIm used in this paper and listed in table 1. The green lines show daily values and the black lines are averages over solar CR periods (of duration approx. 27 days). (a) RMIB/TIM data, EMPIRE model; (b) PMOD data, EMPIRE model; (c) TIM data, EMPIRE model; (d) RMIB data, EMPIRE model; (e) RMIB/TIM data, SATIRE-S model; (f) PMOD data, SATIRE-S model; (g) TIM data, SATIRE-S model; (h) RMIB data, SATIRE-S model; (i) RMIB/TIM data, NRLTSIv2 model; (j), PMOD data, NRLTSIv2 model; (k) TIM data, NRLTSIv2 model; (l) RMIB data, NRLTSIv2 model; (m) TIM data, SATIRE-3D model; (n) PMOD data, SATIRE-3D model. (Online version in colour.)

Figure 9.
Figure 9.

Analysis of the relationship of CR means of quiet-Sun irradiance Q, derived from the difference between the RMIB/TIM data composite and the EMPIRE model, and the heliospheric modulation potential averaged over the prior interval of duration τ, 〈Φ〉τ. (a). The correlation coefficient r between Q and 〈Φ〉τ as a function of τ. (b) The difference between the Q values at the end of the Maunder minimum and for 2019 from the best-fit linear regression for the τ in question. (c) The significance S(τ) in the difference between the peak |r| (rp, at τ = τp) and that for general τ, r(τ), computed using the Meng-Z test. The horizontal dashed and dot-dash lines are at S = 1 − 0.68 and S = 1 − 0.95 and so mark the 1 − σ and 2 − σ levels. In all three panels, the mauve, blue, green and orange vertical lines are at τ of 5.5, 11, 16.5 and 22 years, respectively (the time variations shown in figure 4b). Between the dashed lines is where the difference between |rp| and |r(τ)| is not large enough to be significant at the 1 − σ level. (Online version in colour.)

Figure 10.
Figure 10.

Analysis of the relationship of CR means of quiet-Sun irradiance Q, derived from the difference between the RMIB data composite and the SATIRE-S model, and the heliospheric modulation potential averaged over the prior interval of duration τ, 〈Φ〉τ. The format is as for figure 9. (Online version in colour.)

Figure 11.
Figure 11.

Example of the regression analysis between CR means of the heliospheric modulation potential, averaged over the prior interval of duration τ = 11 years, 〈Φ〉τ, and the quiet Sun irradiance, Q, derived from the difference between the SATIRE-S model and (a,b) the PMOD data composite and (c,d) the RMIB/TIM data composite. The time series (a,c) are derived from the regression fits given in (b,d). The mauve circles in (b,d) are the data points and are fitted with the linear, ordinary least-squares (OLS) regression line shown in black. The cyan squares are the same data with the addition of a drift factor given by the difference between the blue line in figure 3 and that for the data composite in question (the baseline drift in the ACRIM-3 data), and the blue line is the OLS linear regression fit to these data. For the orange triangles, the added drift is that of the TIM data (w.r.t. the data composite used) and the red line is the OLS regression fit. The black, red and blue lines in (a,c) are the time variations derived from the 〈Φ〉τ variation from the three best-fit regression lines shown in (b,d), respectively. (Online version in colour.)

Figure 12.
Figure 12.

The same as figure 11, but using the SATIRE-3D model. (Online version in colour.)

Figure 13.
Figure 13.

Time variations derived in the way exemplified by figures 11 and 12 for all the model–data combinations listed in table 1. The black lines used the PMOD data composite and the mauve lines the RMIB/TIM data composite (or the TIM or RMIB data individually) and separate lines for τ = 11 years and τ = 22 years. The cyan and orange lines are, respectively, the mean and the median variations for the ensemble of 28 variations. The grey area shows the maximum and minimum values of all the maximum deviations of each ensemble member using the error estimates derived from figure 3 and implemented as shown in figures 1 and 12. (Online version in colour.)

Figure 14.
Figure 14.

Model reconstructions of TSI since the Maunder minimum. (a) Two model reconstructions based on sunspot number, TSIm, the red line giving the NRLTSIv2 reconstruction [56] and the blue line the SATIRE-T2 reconstruction [23,24]. In (b) the quiet-Sun irradiance variation ΔQ shown in figure 13 is added to the NRLTSIv2 reconstruction, and in (c) it is added to the SATIRE-T2 reconstruction. In each case, the line is the median value for each year and the grey area is bounded by the maximum and minimum of the set of ΔQ variations shown in figure 13. The data in the plot are available in the electronic supplementary material. (Online version in colour.)

Similar articles

References

    1. Gray LJ, et al. 2010. Solar influences on climate. Rev. Geophys. 48, RG4001 (10.1029/2009RG000282) - DOI
    1. Solanki SK, Krivova NA, Haigh JD. 2013. Solar irradiance variability and climate. Annu. Rev. Astron. Astrophys. 51, 311–351. (10.1146/annurev-astro-082812-141007) - DOI
    1. Owens MJ, Lockwood M, Hawkins E, Usoskin IG, Jones GS, Barnard LA, Schurer A, Fasullo J. 2017. The Maunder minimum and the Little Ice Age: an update from recent reconstructions and climate simulations. J. Space Weather Space Clim. 7, A33 (10.1051/swsc/2017034) - DOI
    1. Shapiro AI, Schmutz W, Rozanov E, Schoell M, Haberreiter M, Shapiro AV, Nyeki S. 2011. A new approach to the long-term reconstruction of the solar irradiance leads to large historical solar forcing. Astron. Astrophys. 529, 1–8. (10.1051/0004-6361/201016173) - DOI
    1. Egorova T, Schmutz W, Rozanov E, Shapiro AI, Usoskin I, Beer J, Tagirov RV, Peter T. 2018. Revised historical solar irradiance forcing. Astron. Astrophys. 615, 1–10. (10.1051/0004-6361/201731199) - DOI

LinkOut - more resources