Selected TSI reconstructions, published over roughly the last decade, extending back to 1510. Shown are the ones by Steinhilber et al. (2012), blue), Wu et al. (2018a, yellow), Dasi-Espuig et al. (2016), green), Egorova et al. (2018, purple, shown are four different versions of this model based on the modulation potential used; thick solid line for Usoskin et al., 2016a, thin solid line for McCracken and Beer, 2017, sparse dashed line for Muscheler et al., 2016, and dashed line for the composite series as used by Shapiro et al., 2011), Lean (2018, red), Wang and Lean (2021, black), and Penza et al. (2022, pink). The horizontal black line marks the constraint derived by Yeo et al. (2020b) on the lowest possible level of TSI under the hypothetical assumption of full stoppage of the global dynamo. All series are offset to match the value of Montillet et al. (2022) composite over 1986. There is one exception for the series by Steinhilber et al. (2012), which does not extend to 1986, and for which we used its overlap with the series by Wu et al. (2018a) over 1900–1980, after offsetting the latter to Montillet et al. (2022) composite over 1986. Shown are annual median values, while the numbers in the upper part of the panel denote the conventional solar cycle numbering. The blue shading marks the period of direct TSI measurements, while Ca II K observations exist over the period shaded both in green and blue.