Total solar irradiance monitoring for climate forcing datasets
Total solar irradiance has been measured from orbit continuously since 1978, but stitching 13+ instruments into one drift-free record is harder than measuring the Sun itself. This page explains how cavity radiometers work, what the 11-year cycle actually delivers to the climate system, and where the calibration bodies are buried.
Sensors
- TIM / SORCE (NASA, 2003–2020): Total Irradiance Monitor on the Solar Radiation and Climate Experiment. Electrically-substituted cavity radiometer; absolute accuracy ~0.035 W/m², precision ~0.001 W/m². Provided the definitive lower TSI value (~1360.8 W/m² at 1 AU) that resolved a long-standing offset between instrument families. Daily cadence.
- TIM / TSIS-1 (NASA, ISS, 2018–present): Total and Spectral Solar Irradiance Sensor, fourth-generation TIM. Continues the SORCE TIM record with improved stray-light control. Mounted on the ISS since December 2017; provides daily TSI observations with an absolute uncertainty target of ±0.03 W/m². Designed to overlap SORCE for inter-calibration.
- VIRGO / SOHO (ESA/NASA, 1996–present): Variability of solar IRradiance and Gravity Oscillations instrument on the Solar and Heliospheric Observatory at L1. Two PMO6-type radiometers plus a sun photometer. Continuous coverage from the L1 Lagrange point; uninterrupted by Earth occultation. Key reference for cycle-24 and cycle-25 TSI behaviour.
- ACRIM III / ACRIMSAT (NASA, 2000–2013): Active Cavity Radiometer Irradiance Monitor III. Flew on the dedicated ACRIMSAT platform. Historically important for the disputed 'ACRIM gap' (1989–1991) between ACRIM I and ACRIM II, during which no overlapping measurements existed and composite records diverge by up to 0.5 W/m².
- PREMOS / PICARD (CNES, 2010–2014): PREcision MOnitor Sensor on the PICARD microsatellite. Provided an independent cross-check during the Solar Cycle 24 minimum. Useful for validating the SORCE TIM absolute scale at a period when multiple instruments were simultaneously operational.
What a cavity radiometer actually does
The principle is disarmingly simple. A blackened conical cavity absorbs incoming solar radiation and heats up. An electrical heater then restores the cavity to its equilibrium temperature, and the power required to do so equals the absorbed solar power. This electrical substitution method ties the measurement to a laboratory-traceable watt rather than to any optical property of a coating or filter, which is why cavity radiometers dominate the TSI record.
The hard part is not the physics. It is controlling stray light, characterising the cavity's absorptance (typically above 0.9999), accounting for thermal gradients, and managing degradation of the aperture edge over years in the ultraviolet-rich solar environment. TIM on SORCE introduced a precision aperture placed at the front of the instrument rather than deep inside, reducing the stray-light problem substantially and producing a TSI value roughly 4.5 W/m² lower than earlier instruments. That discrepancy was not the Sun changing. It was stray light that earlier designs had inadvertently included in their signal.
The 11-year cycle: small signal, large argument
Over a full solar cycle, TSI varies by approximately 1.3 W/m² peak-to-trough, or about 0.1% of the solar constant. At the top of the atmosphere that is not nothing, but the climate forcing it produces is modest. After accounting for Earth's albedo and the geometry of a spherical planet, the effective forcing at the surface is roughly 0.17 W/m² between solar minimum and maximum, compared with the approximately 2.3 W/m² of forcing accumulated from CO₂ since pre-industrial times.
The scientific interest is not in whether solar variability dominates recent warming (the evidence is clear that it does not) but in whether small, persistent offsets between solar minima are detectable and whether they affect decadal climate variability. That question requires a TSI record accurate to better than 0.1 W/m² sustained across decades. No single instrument has survived long enough to answer it alone.
The composite problem: 13 instruments, one record, several disagreements
Continuous satellite TSI measurement began with the Earth Radiation Budget experiment in 1978. Since then, more than a dozen instruments have contributed, each with its own absolute calibration offset, degradation rate, and operational lifetime. Constructing a single composite record requires overlapping successive instruments and applying corrections for drift. Where overlaps are missing, as in the ACRIM gap of 1989 to 1991, different research groups make different assumptions and arrive at composites that disagree on whether TSI at successive solar minima is trending upward, flat, or slightly downward.
Three principal composites exist in the published literature: the ACRIM composite (Willson and colleagues), the PMOD composite (Fröhlich, PMOD/WRC), and the RMIB composite (Dewitte and colleagues at the Royal Meteorological Institute of Belgium). They agree well within a solar cycle but diverge by up to 0.5 W/m² over multi-decadal timescales. The TSIS-1 programme, by maintaining overlap with SORCE and anchoring to the TIM absolute scale, is designed to reduce this ambiguity going forward. It cannot fix the historical record.
Spectral irradiance: the part TSI alone cannot tell you
TSI is the integral of all wavelengths. But the Sun does not vary uniformly across the spectrum. Ultraviolet wavelengths below 300 nm vary by several percent over a solar cycle, while visible and near-infrared wavelengths vary by a fraction of a percent. This matters for stratospheric chemistry, ozone photolysis rates, and the indirect pathways by which solar variability may influence surface climate through dynamical coupling. The SIM instrument on SORCE (Solar Irradiance Monitor) provided the first multi-year spectral record, though its long-term trends in the visible generated controversy that has not been fully resolved.
TSIS-1 carries its own SIM successor. The spectral dimension is where the science is moving, and where the calibration challenges are, if anything, more severe than for broadband TSI.
What an analyst can extract from the public record
The SORCE and TSIS-1 data are publicly archived at the Laboratory for Atmospheric and Space Physics (LASP) at the University of Colorado. VIRGO data flow through the SOHO mission archive at ESA and NASA. Daily and 6-hourly TSI values are available in standard ASCII and NetCDF formats with associated uncertainty estimates. For climate modellers, the most commonly used input is one of the three composites described above, typically at monthly resolution, fed into general circulation models as a boundary condition.
Practical uses for a government or research client include: constructing a locally validated TSI forcing time series for a national climate model; auditing which composite a counterpart agency is using and what difference the choice makes to attribution studies; and monitoring TSIS-1 data in near-real-time to flag anomalies before they propagate into operational climate products. Satellize can run this kind of composite-validation and sensitivity analysis as a structured analytical engagement, drawing on the same open archives that underpin the IPCC forcing datasets. The methodology is published; the value is in applying it rigorously and interpreting the result for a non-specialist decision-maker.
Honest limits of the record
The absolute accuracy of any single instrument is around 0.03 to 0.1 W/m², depending on generation. That sounds small, but it is comparable to the multi-decadal trend signal researchers are trying to detect. Instrument degradation in orbit is partially correctable using backup cavities exposed at different duty cycles, but the correction is model-dependent. The historical composites will likely never fully converge on the pre-2003 period.
Cloud cover is not a constraint here: cavity radiometers point at the Sun, not the surface. Revisit is not a constraint either: instruments operate continuously in sunlight. The binding constraints are absolute calibration traceability, long-term stability, and the willingness of successive space agencies to fund overlap periods between missions rather than allowing gaps. The ACRIM gap is a cautionary example of what happens when funding and launch schedules do not cooperate.
Typical figures
| Measurement principle | Electrically-substituted cavity radiometry (active cavity) |
| Spectral range (TSI) | Broadband, ~0.2 to 100 µm (all solar output integrated) |
| Absolute accuracy (TSIS-1 TIM) | ±0.03 W/m² (target); SORCE TIM achieved ~0.035 W/m² |
| Measurement precision (short-term) | ~0.001 W/m² (< 1 ppm) for TIM-class instruments |
| TSI solar-cycle amplitude | ~1.3 W/m² peak-to-trough (~0.1% of ~1361 W/m²) |
| Temporal cadence | Daily (some instruments: 6-hourly or orbit-cadence) |
| Archive depth | Continuous satellite record from November 1978 (ERB/Nimbus-7) |
| Composite record disagreement (multi-decadal) | Up to ~0.5 W/m² between ACRIM and PMOD composites over 1980–2000 |
| Data formats | ASCII, NetCDF-4; distributed via LASP, NASA Earthdata, ESA SOHO archive |
| Coverage | Global (Sun-pointing instruments; no geographic footprint constraint) |
Analytics Satellize can run
| TSI composite validation report | Cross-comparison of ACRIM, PMOD and RMIB composites against TSIS-1 anchor; sensitivity analysis of forcing difference on decadal timescales | Structured PDF report with composite divergence quantified and implications for national climate model boundary conditions stated explicitly |
| Solar forcing time series for GCM input | Selection and quality-screening of preferred composite; application of 1-AU distance correction and geometry factor for effective surface forcing | NetCDF time series file formatted to CMIP6 forcing conventions, with provenance and uncertainty metadata |
| Near-real-time TSIS-1 anomaly monitoring | Automated ingestion of LASP daily TSI product; statistical process control against 27-day solar rotation baseline | Weekly digest flag report; alert notification if daily TSI deviates beyond 3-sigma from rotational baseline |
| Solar-cycle phase characterisation | Sunspot number correlation with TSI record; identification of cycle minimum and maximum epochs using published SIDC sunspot data | Annotated time-series chart with cycle-phase labels for use in attribution study documentation |
| Spectral irradiance sensitivity assessment | Comparison of SIM/SORCE and TSIS-1 SIM spectral records in UV, visible and NIR bands; quantification of stratospheric forcing pathway uncertainty | Technical memo on spectral-vs-broadband forcing uncertainty, suitable for inclusion in national climate assessment annexes |
Who does the work
We can get this done for you. Satellize runs its own analyst desk and a strong science team. You do not buy a data feed and work out what it means; our people source the imagery, run the analysis described on this page, and hand you the answer with its confidence limits stated. Discuss this requirement.