How Different ECMWF Atmospheric Models Impact the Estimation of Sea-Level Trends

In the context of satellite altimetry, numerical weather models (NWMs) are relevant sources to derive the wet path delay (WPD) caused by the water vapor and cloud liquid water content in the atmosphere. Over regions where WPD based on observations are not available, these may be the only source of information. This article presents an assessment of the state-of-the-art NWMs from the European Centre for Medium-Range Weather Forecasts (ECMWF), the operational model (ECMWF OP), and the latest reanalysis (ERA5), for estimating the WPD of satellite altimeter observations, with focus on the direct impact of using different models on the estimation of altimeter-derived sea-level trends. Results show that, due to regular model updates, ECMWF OP possesses significant discontinuities that induce errors in sea-level trend which, in some regions, can exceed 1 mm yr
⁻¹
in absolute value, over periods of 7–8 years. The major impacts occur over the periods when Jason-2 and Jason-3 are in the reference orbit (phase A). Since this is the only model adopted in most altimeter products, it is strongly recommended that the most recent reanalysis model (currently ERA5) is included. Overall, ERA5 is the best compromise between model accuracy and stability, for sea/water-level studies and generally all climate applications, over regions where WPD based on observations is not available, such as some coastal and inland water regions.