Quantifying the precision of retracked Jason-2 sea level data in the 0-5 km Australian coastal zone

This paper aims to quantify the precision of retracked Jason-2 altimeter sea level data (2008-2016) in the 0-5 km Australian coastal zone. A new retracking strategy has been developed to achieve the highest possible precision of coastal sea-level data. Coastal waveforms were classified into four groups according to the adaptability of dedicated coastal retrackers, from which each group was reprocessed using an optimal retracker. A modified Brown model was introduced to reprocess the quasi-specular waveforms over calm waters. A seamless transition of sea surface heights from different retrackers was achieved by estimating and removing the significant wave height dependent height differences. In order to generate the 20-Hz sea level anomaly (SLA) and total water level envelope (TWLE) at the coast, the regional along-track mean sea surface (MSS) with a high spatial resolution (-300 m) was estimated. Compared to the global MSS model, the use of the along-track MSS has dramatically reduced the SLA variance by 140 cm2 in the 0-15 km coastal strip. The results from the evaluation of data precision for Jason-2 ground tracks and the validation against tide gauges show that this new retracking strategy can retrieve more reliable sea level data nearshore. The precision of 20-Hz sea level data in the 0-5 km Australian coastal strip ranges typically from 5 cm to 6 cm in most coastal areas, which is only -1 cm lower than that beyond 5 km off the coastline. However, in Northwest Australian and some coastal regions, the data precision drops to >= 7 cm due partly to the specific coastal sea states and topography. In addition, it is found that the altimeter and tide gauge TWLE time series have a high correlation (>0.8) in the 0-5 km distance band, indicating that coastal sea-level data are useful for applications such as estimation of river plumes, long-term variation of water level in lakes and rivers, and the analysis of storm surges.