The originality of this study is to propose a new calibration method based on two calibration phases between Jason-3 and Sentinel-6A (S6A) to better estimate the relative global and regional mean sea level drifts between the two missions. To date, a first calibration phase of approximately 12 months is planned from January 15, 2021, to December 31, 2021, when both satellites will be on the same orbit spaced out by approximately 30 seconds. This calibration will allow for a very accurate assessment of the GMSL bias between Jason-3 and S6A (less than 0.5 mm, see ​ Zawadzki and Ablain, 2016​). A second calibration phase after a few years would reduce the uncertainty levels of the GMSL (global mean seal level) drift estimate. The uncertainty would be low enough to detect any drift detrimental to the stability of the current GMSL record. It would indeed be possible to evaluate the stability between the two satellites with an accuracy at least 3 times better at the global scale than with the most accurate method to date. At regional scales, the second calibration phases would provide regional MSL drift estimates with very good precision. This study also shows that the time spent between the two calibration phases is significantly more sensitive than the length of the second calibration phase for the reduction in uncertainties. Finally, a possible scenario proposed by this study would consist of carrying out the beginning of the second calibration phase approximately 1.5-2 years after the first and for a duration of 3-4 months. This calibration would allow the detection of a relative GMSL drift of approximately 0.15 mm/yr and 0.4-0.5 mm/yr at oceanic basin scales (2000-4000 km).