Vertical deformation and residual altimeter systematic errors around continental Australia inferred from a Kalman-based approach

We further developed a space-time Kalman approach to investigate time-fixed and time-variable signals in vertical land motion (VLM) and residual altimeter systematic errors around the Australian coast, through combining multi-mission absolute sea-level (ASL), relative sea-level from tide gauges (TGs) and Global Positioning System (GPS) height time series. Our results confirmed coastal subsidence in broad agreement with GPS velocities and unexplained by glacial isostatic adjustment alone. VLM determined at individual TGs differs from spatially interpolated GPS velocities by up to similar to 1.5 mm/year, yielding a similar to 40% reduction in RMSE of geographic ASL variability at TGs around Australia. Our mission-specific altimeter error estimates are small but significant (typically within similar to +/- 0.5-1.0 mm/year), with negligible effect on the average ASL rate. Our circum-Australia ASL rate is higher than previous results, suggesting an acceleration in the similar to 27-year time series. Analysis of the time-variability of altimeter errors confirmed stability for most missions except for Jason-2 with an anomaly reaching similar to 2.8 mm/year in the first similar to 3.5 years of operation, supported by analysis from the Bass Strait altimeter validation facility. Data predominantly from the reference missions and located well off narrow shelf regions was shown to bias results by as much as similar to 0.5 mm/year and highlights that residual oceanographic signals remain a fundamental limitation. Incorporating non-reference-mission measurements well on the shelf helped to mitigate this effect. Comparing stacked nonlinear VLM estimates and altimeter systematic errors with the El Nino-Southern Oscillation shows weak correlation and suggests our approach improves the ability to explore nonlinear localized signals and is suitable for other regional- and global-scale studies.