CoAuthors

Christopher Watson (University of Tasmania, Australia); Matt King (University of Tasmania, Australia); Benoit Legresy (CSIRO, Australia)

Vertical land motion (VLM) is important for a range of geophysical and climate applications and remains a vital connection between relative and absolute estimates of sea level change. VLM also plays a prominent role in evaluating altimeter systematic errors using tide gauge (TG) records. Within this process, it is often assumed that TG VLMs are linear and inferred reliably from nearby GPS sites. This assumption can quickly break down owing to the spatial and temporal limitation of GPS records and many localised drivers of VLM. We develop a space-time approach involving Kalman filtering and smoothing to simultaneously estimate both altimeter systematic errors and VLM in a flexible manner. Our approach combines altimeter minus tide gauge, altimetry crossovers, and GPS observations in a spatio-temporal framework. We assess the performance of the method in the Baltic Sea that is characterised with a dense network of TG and GPS sites. We show the technique can separate TG VLMs from small but significant regional systematic errors in the reference altimetry missions. We quantify local variability of VLM at TGs that otherwise cannot be inferred from interpolation of the GPS velocity field, or from the use of a GIA model.