Solar irradiance measurements

Abstract

The Sun provides nearly all the energy powering the Earth's climate system, far exceeding all other energy sources combined. The incident radiant energy, the "total solar irradiance," has been measured by an uninterrupted series of temporally overlapping precision space-borne radiometric instruments since 1978, giving a record spanning more than four 11-year solar cycles. Short-term total-irradiance variations exceeding 0.1% can occur over a few days while variations of ~ 0.1% in-phase with the solar cycle are typical. Knowledge of solar variability on timescales longer than the current multi-decadal space-borne record relies on solar-activity proxies and models, which indicate similar-magnitude changes over centuries. Spectrally resolved space-borne irradiance measurements in the ultraviolet have been acquired continuously since 1979, while measurements contiguously spanning the near-ultraviolet to the near-infrared began in 2003. The combination of long-term total- and spectral-irradiance measurements helps determine both the solar causes of irradiance variability, which are primarily due to solar-surface magnetic-activity regions such as sunspots and faculae, and the mechanisms by which solar variability affects the Earth's climate system, with global and regional temperatures responding to variability at solar-cycle and longer timescales. To better understand these solar influences, the most modern total-irradiance instruments are approaching the needed climate-driven measurement accuracy and stability requirements for detection of potential long-term solar-variability trends, while the latest spectral-irradiance instruments are beginning to be able to discern solar-cycle variability. Focusing on the space-borne era where such measurements are the most accurate and stable, this article describes solar-irradiance instrument designs, capabilities, and operational methodologies. It summarizes the many total- and spectral-irradiance measurements available and the measured solar variabilities on timescales from minutes to solar cycles and discusses extrapolations via models to longer timescales. Measurement composites and reference spectra are reviewed. Current capabilities and future directions are described along with the climate-driven solar-irradiance measurement requirements.

Keywords: Earth-energy balance; SSI; Solar climate data record; Solar constant; Solar insolation; Solar irradiance; Solar radiometry; Spectral solar irradiance; TSI; Top-of-atmosphere flux; Total solar irradiance.

Fig. 1

Earth energy imbalance. The Earth-climate system is radiatively in near-equilibrium, with incident short-wave solar radiation approximately balanced by outgoing reflected-solar and long-wave radiation. Units are W m⁻² averaged over the Earth’s surface. Image reproduced with permission from Stephens et al. (2022), copyright by AMS

Fig. 2

Hybrid solar reference spectrum. This high-resolution spectrum representative of solar minimum combines high-spectral-resolution models giving fine spectral detail with lower-resolution SSI measurements normalizing to the correct scale

Fig. 3

Early solar irradiance instrument. Pouillet’s instrument for measuring the TSI by heating a known mass of water in a cylindrical container (a) by sunlight incident on the absorptive surface (b). From the measured rate of temperature change measured by the shielded thermometer (d) and the known heat capacity of water, the incident solar power can be computed, and the surface (b) gives the area over which it is collected. A similarly sized plate (e) enables alignment normal to the incident sunlight. (from Young 1880)

Fig. 4

Earth-climate measurement requirements in the reflected-solar spectral region: Climate-required measurement accuracies and current capabilities of Earth-incoming and -outgoing radiation across the solar spectrum are summarized in this plot from Fox et al. (2022). (Note that the requirements shown state lower uncertainties than imposed on many spacecraft programs, as these are the climate-driven requirements as opposed to the requirements levied on instruments in order to programmatically claim mission success.)

Fig. 5

TSI missions. Nearly 20 TSI missions with publicly available data have contributed to the uninterrupted 47-year data record. Gray shading indicates times before the present

Fig. 6

Space-borne TSI Record. The TSI has been measured from space via an uninterrupted series of overlapping instruments since 1978. Scale differences between instruments are due to calibration differences. Note that TSI fluctuations of ~ 0.1% are in phase with solar-surface magnetic activity over the 11-year solar cycle, as indicated by sunspot numbers (black). (updated regularly at http://spot.colorado.edu/~koppg/TSI)

Fig. 7

ERB cross-section. The single-channel ERB used an inverted-cone geometry inside a cylindrical cavity to absorb incident sunlight, which, in this figure, enters from the left. The precision 0.5-cm² aperture defining the area over which light is collected is immediately in front of the cavity, as on most early instruments, while upstream baffles define the field of view. Image reproduced with permission from Hickey et al. (1988), copyright by Springer

Fig. 8

ERB TSI Data. The daily TSI values are plotted from the ERB, which initiated the space-borne TSI record. Both the uncorrected data, as provided by the original instrument team, and C. Fröhlich’s estimated corrected data are shown for this instrument that, being a single-channel radiometer, was unable to internally monitor its on-orbit degradation

Fig. 9

ACRIM ESR. The back-to-back ACRIM ESR-cavity design provided thermal stability and low sensitivity to common-mode thermal changes. Three such back-to-back conical-cavity pairs per instrument provided redundancy and enabled degradation tracking. (courtesy R. Willson 2012, personal communication)

Fig. 10

ACRIM TSI data. The daily ACRIM-1, -2, and -3 TSI values are plotted. Other than the “ACRIM Gap,” the measurement gap between the ACRIM-1 and ACRIM-2, these instruments provided the primary NASA TSI measurements for 3.5 decades. Scale differences between them are indicative of the uncertainties in these three instruments. Only the ACRIM-3 data have been corrected for internal-instrument scatter, decreasing those values to nearly match the lower TSI value established by the SORCE/TIM; the older ACRIMs’ data were never corrected by the ACRIM team

Fig. 11

ERBE TSI data. This single-channel instrument spanned nearly two solar cycles with very little presumed degradation. It acquired only sparse (bi-weekly) measurements and suffered from high instrument noise

Fig. 12

Cutaway of one VIRGO PMO6 cavity pair. The PMO6 portion of the VIRGO used two pairs of back-to-back cavities with an inverted geometry, such that the cone apex faced toward the entering sunlight, which, in this figure, enters from the left. Note the primary aperture, designated by the circle marked “A,” is deep inside the instrument, immediately in front of the primary ESR cavity. Several upstream baffles and other portions of the instrument are illuminated by sunlight and can scatter additional light into the primary ESR, as described in Sect. 2.4.5. (courtesy of C. Fröhlich, personal communication)

Fig. 13

VIRGO degradation measurements. Measurements with each of the four cavities in the VIRGO enable degradation tracking due to solar exposure. The PMO6V-A shows greater degradation than any prior TSI radiometer. Measurements with PMO6V-B, used 150 × less, help correct this. The primary DIARAD channel, DIARAD-L, shows much less long-term degradation. The DIARAD pairs are used to track each other as well as allow additional corrections to the PMO6 channels. (Plot courtesy of C. Fröhlich, VIRGO PI. The plotted data are on the VIRGO’s “original” scale and have not been corrected to the newer lower value.)

Fig. 14

VIRGO TSI data. The VIRGO provides the longest-duration TSI measurement record of any single instrument

Fig. 15

TIM instrument cutaway. Four black absorptive cavities (two shown) measure solar power passing through precision apertures in a temperature-controlled instrument. In this drawing, sunlight enters from the left, passing through a small front-mounted light-limiting precision aperture before reaching the ESR cavities deep within the instrument interior. This geometry, used on most subsequent TSI instruments, only allows the light desired to be measured into the instrument, reducing internal-instrument scatter that plagued prior instruments. (Compare to VIRGO PMO6 design in Fig. 12.)

Fig. 16

SORCE/TIM TSI data. SORCE/TIM, being a new instrument design that achieved roughly 10 × lower uncertainties than prior instruments, established the now-accepted lower mean TSI value of 1361 W m⁻²

Fig. 17

PICARD PREMOS and SOVAP TSI data. The PREMOS was the first TSI instrument to transfer TSI calibrations from the ground-based SI-traceable TRF to space. Its results support the now-accepted lower TSI value established by the SORCE/TIM. The SOVAP reported measurements much lower than those on the VIRGO/DIARAD but still higher than other concurrent TSI instruments. Other than the scale difference, the two agree well on TSI variability

Fig. 18

TCTE/TIM TSI data. The TCTE/TIM was a mission to help guarantee overlap between the predecessor SORCE and the successor TSIS-1 TIMs

Fig. 19

FY-3E SIAR ESR cutaway. The SIAR uses three back-to-back ESR pairs, such as the one shown here. The precision aperture is mounted immediately sunward of the ESR and deep within the instrument, as in older instrument designs. Image reproduced with permission from Song et al. (2021), copyright by the author(s)

Fig. 20

FY-3E TSI data FY-3E/SIM is the best of a series of Chinese Space Agency instruments acquiring TSI measurements that is providing publicly available data. The FY-3E/DARA data have also recently been released. These data are inexplicably lower than the mission’s SIM instrument (or any other currently flying instrument) by more than their combined estimated uncertainties. [The FY-3E mission is owned and operated by the China Meteorological Administration (https://doi.org/10.1007/s11207-021-01794-5). The FY 3E DARA data are retrieved and processed via the cooperation between the Changchun Institute of Optics, Fine Mechanics and Physics Chinese Academy of Sciences (CIOMP/CAS, China), the China Meteorological Administration (CMA, China), and PMOD/WRC (Davos, Switzerland) and are available at the Interdisciplinary Earth Data Alliance (IEDA) https://doi.org/10.60520/IEDA/113059]

Fig. 21

TSIS-1/TIM TSI data. The TSIS-1/TIM, an updated SORCE/TIM instrument, continues the NASA and NOAA climate-data record of TSI

Fig. 22

CTIM TSI data. The CTIM is a technology-demonstration mission using miniature ESRs. As a CubeSat-based demonstration, its measurement record has more gaps than most prior instruments, but it agrees well with concurrent instruments in both temporal variability and absolute scale

Fig. 23

TSI data ca 2012. This is a figure using TSI data from early 2012 to show the state of the measurements at that time. The TSI instruments prior to the SORCE/TIM were in consensus agreement on values near 1366 W m⁻², making the SORCE/TIM values of 1361 W m⁻² initially suspect. After nearly 10 years of community discussion and lab tests, the higher values were found to be erroneous. The ACRIM-3 values were retroactively lowered in ground processing in 2011 and then the VIRGO values in 2014. [Includes data from: www.ngdc.noaa.gov/stp/SOLAR/solar.html (NIMBUS7/ERB, ERBS/ERBE, NOAA9, and NOAA10); http://www.acrim.com (ACRIM1, ACRIM2, and ACRIM3); the VIRGO team via ftp://ftp.pmodwrc.ch courtesy of the VIRGO Experiment on the cooperative ESA/NASA Mission SOHO from VIRGO Team through PMOD/WRC, Davos, Switzerland; http://lasp.colorado.edu/home/sorce/data/tsi-data/ (SORCE/TIM).]

Fig. 24

SSI measurements and related data. This plot by Schöll et al. (2016) summarizing the SOLID Composite input data shows select space-borne instruments acquiring SSI measurements and other related data as a function of time and wavelength. Instrument data are shown by the colored boxes, time-specific reference spectra by the vertical lines spanning their wavelength range, and proxy data by the horizontal lines at the top

Fig. 25

ATLAS reference spectrum. This classic reference spectrum is a composite built around SOLSPEC measurements from ATLAS shuttle flights

Fig. 26

SOLAR-ISS-V2.0 SOLSPEC reference spectrum

Fig. 27

WHI reference spectrum. The lower resolution at visible and NIR wavelengths in this spectrum compared to other reference spectra is because the WHI reference spectrum is purely instrument based and does not attempt to improve spectral resolution with other (ground or aircraft) measurements

Fig. 28

SAO2010 reference spectrum. Being a combination of high spectral resolution ground- and balloon-based measurements with radiometric space-based measurements, even on the scale shown here, the SAO2010 spectrum has higher spectral resolution than other reference spectra

Fig. 29

HSRS extended. The high-resolution version of the HSRS Extended spectrum (also known as the FS-HSRS, for “full spectrum”) is plotted along with the uncertainties

Fig. 30

Comparisons of reference spectra. The four plotted reference spectra are ratioed to their mean. Differences are largest above 1500 nm, where the older (ATLAS-3 and WHI) spectra are up to ~ 8% higher than the TSIS-1/SIM-based HSRS

Fig. 31

SSI measurements from 200 to 205 nm from the earlier SBUV-instrument series along with those from other instruments during the era. Image reproduced with permission from DeLand et al. (2004), copyright by AGU

Fig. 32

SCIAMACHY SSI reference spectrum from 27 February 2003. The spectrum is more accurately termed a beginning-of-life “baseline” spectrum from which relative degradation corrections are tracked, as it has several spectral gaps. (data from https://www.iup.uni-bremen.de/UVSAT/data/solarreference/) (see Hilbig et al. 2018)

Fig. 33

SCIAMACHY SSI time series. (from https://www.iup.uni-bremen.de/UVSAT/data/solartimeseries/, based on Hilbig et al. 2020)

Fig. 34

SORCE TSI and binned SCIAMACHY SSI. While SCIAMACHY can provide useful short-term solar-variability data, the longer-term (solar-cycle) results are influenced by residual instrument-stability effects. (See Hilbig et al. .)

Fig. 35

SORCE TSI and binned SSI Over a 5-month period. The SSI integrated over 100-nm bins about the shown center wavelengths generally follows the TSI over this 5-month period over which the SIM is relatively stable. The longest wavelengths were not scanned regularly, so have gaps

Fig. 36

SORCE TSI and binned SSI. The SORCE TSI (top plot) shows solar-cycle variability over the 17-year mission duration. The binned SORCE SSI measurements (bottom 6 plots) lack the stability to definitively show realistic solar-cycle responses. The apparent trends are likely from uncorrected instrument artifacts

Fig. 37

SORCE TSI and binned MuSIL SSI. The MuSIL approach corrects for some of the long-term SORCE/SIM trends by forcing agreement with solar proxies. These data suggest an out-of-phase solar-cycle dependence in the NIR

Fig. 38

SORCE TSI and SIMc V.2 SSI. The SIMc corrects the long-term SORCE/SIM data by applying constraints to follow the TSI. These spectral bins are larger than in prior plots since the SIMc wavelengths are reported on sparser scales than the instrument data

Fig. 39

Spectral ranges of TROPOMI, OMI, GOME, and SCIAMACHY. (Cropped version of Figure 2 in Veefkind et al. 2012)

Fig. 40

TSIS-1/SIM Spectral Resolution. The TSIS-1/SIM spectral resolution varies between ~ 1 and 34 nm over the 200 to 2400-nm range. The SORCE/SIM is similar

Fig. 41

TSIS-1 TSI and binned SSI. The TSIS-1 TSI (top plot) shows increasing activity heading into Solar Cycle 25 that is reasonably similar to the binned TSIS-1 SSI measurements (bottom 6 plots), indicating the greatly improved stability of this instrument’s SIM (compare to Fig. 36)

Fig. 42

VIRGO SPM data. The degradation-corrected SPM daily data from the blue, green, and red channels are plotted for the early part of the SoHO mission spanning the maximum of Solar Cycle 23. The annual-period oscillations are likely an uncorrected thermal effect, being in-phase with the Sun-SoHO distance. (Data courtesy of C. Fröhlich, ‘SPM_lev20a_d_170496_290508.idl’)

Fig. 43

Pictorial TSI composite. Measurement continuity helps correct for instrument-scale differences shown in Fig. 6, allowing each instrument’s data to be scaled to the more-accurate current TSI values for the creation of composite solar-irradiance records

Fig. 44

Traditional TSI composites. Three formerly prominent, traditional, PI-created TSI composites show different trends relative to the 1986 solar minimum (indicated by the dashed blue line in each plot) that are due to differing selections of how the contributing instruments are weighted or corrected for use in each composite. These trend differences are indicative of long-term uncertainties in the measurement record, which limit the ability to definitively discern a secular trend over the measurement period. Note also the absolute scales differ for each composite. The monthly sunspot number is shown in the bottom plot (black)

Fig. 45

Community consensus TSI composite. The Community Consensus TSI Composite uses an unbiased statistics-based methodology to estimate instrument uncertainties and weightings to produce a TSI composite over the space-borne TSI-measurement era. The daily TSI values are shown in red with a near-annual smoothing in black using the left-hand vertical axis. The blue points indicate the statistical uncertainties in the composite using the right-hand vertical axis (note the large differences in axis scales). The light-blue values show the TSI measurements with those uncertainties added, although they are nearly indistinguishable from the data themselves due to the scale of the uncertainties. Note that the uncertainties are lower for the more recent (and more accurate) instruments and at times having a greater number of available instruments. (from http://spot.colorado.edu/~koppg/TSI)

Fig. 46

SOLID composite. This surface plot shows the SOLID SSI data over the spacecraft era. The full data extend from 1951 through 2014. Variations in time are nearly imperceptible on this scale

Fig. 47

GSFC SSI3 composite. This SSI composite spans a broad wavelength range by combining instrument measurements and solar proxies. The average over a 100-nm mid-visible bandpass is shown

Fig. 48

Short-term TSI variability. The TSI varies continually at the ~ 0.01% level on timescales of minutes due to the superposition of convection and oscillations on the visible solar disk (updated from Kopp et al. 2005a, b)

Fig. 49

Flare measurement in TSI. The X17 flare on 28 October 2003 caused an abrupt, short-duration 0.028% increase in the TSI. Data shown are from the SORCE/TIM with 50-s cadence (red) and the GOES XRS (green). The TSI measurements enable estimates of the total radiant flare energy

Fig. 50

TSI response due to sunspot and faculae center-to-limb variations. This figure shows the TSI variability due to solar-rotation timescale passages of sunspots (top) and faculae (bottom) with their different center-to-limb variations. Image reproduced with permission from Solanki and Fligge (2002), copyright by ESA

Fig. 51

SSI sensitivity to TSI variability. On solar-rotation timescales, the SSI varies by nearly twice the TSI at wavelengths near 400 nm, by the same relative amount at around 650 nm, and by half of the relative TSI variability in the NIR (top plot). This correlation holds for both empirical data (orange; from SORCE/ TIM) and models (green; SATIRE-like model). The individual contributions from sunspots and faculae are shown in the bottom plot. Image reproduced with permission from Kopp et al. (2024), copyright by the author(s)

Fig. 52

“Original” PMOD TSI composite. The PMOD TSI composite by Fröhlich shows peak-to-peak TSI variability of slightly less than 0.1% in each of the three solar cycles observed during the space-borne measurement record, with that variability being in phase with solar activity. Different colors indicate the binary selections of each instrument used in the creation of the composite. The right-hand vertical scale indicates the more-accurate, currently accepted absolute value. [Courtesy of C. Fröhlich and the VIRGO team via ftp://ftp.pmodwrc.ch/pub/Claus/ISSI_WS2005/ISSI2005a_CF.pdf]

Fig. 53

Historical TSI reconstructions. TSI reconstructions provide estimates of solar variability from the Maunder Minimum to the present and are largely based on the sunspot record (adapted from Kopp 2014)