Panel a: reconstructed α versus σ₀ for S₁, full temporal sampling (with a 4 month gap every year) and different methods (see Fig. 10, lower panels, for the color-coding; the curves for methods 3 and 4 are almost indistinguishable here). All noise realizations have been averaged. The true value is indicated by a solid line (only for this panel). Panel b: same, but for the rms RV of . Panel c: same, but for the correlation between and . Panel d: same, but for the rms RV of . Panel e: same, but for the correlation between and . Panel f: same, but for the correlation between and .