Coastal Sea State Uncertainty From a Triple Collocation Analysis of Observations During the Sentinel-6 Michael Freilich – Jason-3 Tandem Phase Experiment

The growing satellite record of sea state observations is becoming increasingly important for climate change research, to improve ocean and weather forecasts and to inform climate change mitigation and investment strategies. In this context, coastal processes and impacts are of particular concern, driving a yet stronger research imperative. The Copernicus Sentinel-6 Michael Freilich (S6-MF) mission was launched in November 2020 by the European Space Agency to succeed Jason-3 (J3) as the long term satellite altimetry reference mission. S6-MF commissioning involved a unique 12-month Tandem Experiment during which S6-MF flew approximately 30 seconds behind J3 on the same ground tracks, resulting in an unprecedented global dataset of quasi-simultaneous collocated altimeter sea state measurements in Low-Resolution Mode (LRM) and Synthetic Aperture Radar (SAR) mode.

In this work, this unique dataset is examined to evaluate uncertainties in altimeter significant wave height (Hs) observations from the two missions in different operating modes and different sea state conditions. A particular focus is placed on the evaluation of uncertainties in the coastal zone by exploiting the large number of moored buoys located near the coast of the U.S. S6-MF and J3 data are compared with buoy measurements and reanalysis data using, amongst other methods, triple collocation (TC) analysis. Attention is paid to both the collocation methodology and possible correlation of random errors. Results indicate that, over both global and coastal oceans, J3 and S6-MF Low-Resolution Hs are almost identical, with near-zero bias, low RMS difference and very high correlation. This very high correlation precludes the use of triple collocation to the J3/S6-MF-SAR/buoy triplets. Comparing S6-MF SAR with J3 LRM and buoys confirms the positive sea-state dependent bias in SAR Hs. Triple collocation of J3, S6-MF and buoys reveals the sensitivity of measurement uncertainty to collocation criteria, particular in coastal areas. Further, we show how sea state dependence of measurement uncertainty varies between oceanic and coastal settings. In general, we find that steeper spatial gradients of sea state typically associated with coastal regions can hamper interpretation of TC analyses without undue consideration. These findings demonstrate the value of the Tandem Experiment to evaluate uncertainty and provide evidence of the stability and/or enhancements of new mission data contributing to the growing satellite climate record.