Same as Fig. 7 but for GJ 486. In this case, we find a discrepancy between the lines observed with COS/G130M and those observed with COS/G160M and COS/G230L. It is possible that this discrepancy is due to the large flares observed during the COS/G130M observations. We only use lines observed with COS/G160M and COS/G230L (darker purple points with 1σ errors) to estimate the Lyα flux with the UV– UV correlation method. The Lyα values from the reconstruction and the UV–UV estimation agree to within 2σ.