Reconstructing solar irradiance from historical Ca II K observations I. Method and its validation

Context. Knowledge of solar irradiance variability is critical to Earth's climate models and understanding the solar influence on Earth's climate. Direct solar irradiance measurements have only been available since 1978. Reconstructions of past variability typically rely on sunspot data. However, sunspot records provide only indirect information on the facular and network regions, which are decisive contributors to irradiance variability on timescales of the solar cycle and longer. Aims. Our ultimate goal is to reconstruct past solar irradiance variations using historical full-disc Ca II K observations to describe the facular contribution independently of sunspot observations. Here, we develop the method and test it extensively by using modern CCD-based (charge-coupled device) Ca II K observations. We also carry out initial tests on two photographic archives. Methods. We employ carefully reduced and calibrated Ca II K images from 13 datasets, including some of the most prominent series, such as those from the Meudon, Mt Wilson, and Rome observatories. We convert them to unsigned magnetic field maps and then use them as input to the adapted Spectral and Total Irradiance Reconstruction (SATIRE) model to reconstruct total solar irradiance (TSI) variations over the period 1978-2019, for which direct irradiance measurements are available. Results. The reconstructed irradiance from the analysed Ca II K archives agrees well with direct irradiance measurements and existing reconstructions. The model also returns good results on data taken with different bandpasses and images with low spatial resolution. Historical Ca II K archives suffer from numerous inconsistencies, but we show that these archives can still be used to reconstruct TSI with reasonable accuracy provided the observations are accurately processed and the effects of changes in instrumentation and instrumental parameters are identified and accounted for. The reconstructions are relatively insensitive to the TSI reference record used to fix the single free parameter of the model. Furthermore, even employment of a series, itself reconstructed from Ca II K data, as a reference for further reconstructions returns nearly equally accurate results. This will enable the Ca II K archives without an overlap with direct irradiance measurements to be used to reconstruct past irradiance. Conclusions. By using the unsigned magnetic maps of the Sun reconstructed from modern high-quality Ca II K observations as input into the SATIRE model, we can reconstruct solar irradiance variations nearly as accurately as from directly recorded magnetograms. Historical Ca II K observations can also be used for past irradiance reconstructions but need additional care, for example identifying and accounting for discontinuities and changes in the quality of the data with time.