Final Report Summary - GPS4SEALEVELS (Advancing understanding of sea level change through improved vertical land movement estimates)
The project “GPS4SeaLevelS” (330103) addresses the ability to correct the vertical land motion (VLM) at tide gauges (TG) for sea level studies and hence a step-change in the accuracy of sea level reconstructions over the last two centuries.
The majority of sea level studies, including those that form the foundation for syntheses presented by the Intergovernmental Panel on Climate Change, use solely models of Glacial Isostatic Adjustment (GIA) to account for VLM at TGs. These models contain known biases over regional scales (e.g. Europe), and capture only one component of VLM, neglecting various processes that exist over different spatial and temporal scales. Direct observation of VLM using GPS is therefore vital. Despite many advances in this technique, systematic errors remain unaccounted for, limiting the precision that can be achieved. In addressing these packages, the objective is to reduce biases and uncertainty in sea level change estimates toward 0.1 mm/yr.
The objectives of this fellowship are:
- To improve the precision and accuracy of GPS position time series. This will have a major impact not only to the sea level community, but across a broad range of geosciences (e.g. surface mass loading, reference frames).
- To integrate these GPS estimates of VLM into the international satellite altimeter validation process, thus enabling accurate determination of altimeter drift biases.
- To derive a VLM velocity field for application to the global network of long-running high quality TGs to improve assessment of long term sea level change at global and regional scales.
This fellowship will provide high level training and international exposure from the leading geodesy and altimetry validation group at the University of Tasmania, Australia. It also represents an invaluable opportunity for knowledge transfer back to the University of La Rochelle, France, bringing a significant advance in the understanding of sea level change allowing the Fellow to continue his scientific development at the highest level in Europe.
Since the beginning of the project, the fellow has extended his data processing abilities being trained in new processing software (GIPSY/OASYS software package). Using the processing facilities and training at the University of Tasmania, the fellow has develop his own processing scheme in GIPSY/OASIS and has successfully reprocessed the International GNSS Service core network using the most updated models and corrections. The obtained training will allow the fellow to estimate the VLM of any GPS station accurately and faster than his previous abilities. For instance, the fellow has used this new expertise to investigate the coastal deformation of the Antarctic continent after a large earthquake and its implications for modelling the GIA. Under this context, the fellow has also received training in modelling the solid Earth response and the computation of surface deformation from co- and post-seismic events. The fellow has also investigated the impact of interannual deformation from modelled surface fluid mass changes on the estimated GPS velocities at global scale. In particular, the fellow bounded current GPS vertical velocity errors used to correct decadal to century-long tide gauges and identified regions where longer GPS time series are needed to mitigate this effect.
The fellow has also investigated the effects of far-field multipath in the GPS observations, which is one of the main systematic errors still present in state-of-the-art GPS time series. While investigating this topic, the fellow found a new way to use observations affected by multipath transforming this problem into an innovative method to remotely and continuously level co-located GNSS and tide gauge stations. Using this new GNSS reflectometry technique, the fellow has obtained significant results including a new technique that provides the sea level community with a means to tying the vertical position of the reference points of co-located tide gauge and GNSS stations remotely, precisely and at no added cost. This method has the potential to solve the problem of the missing levelling connections at TGs, enabling further research on the connection of national height datums and mean dynamic topography of the sea surface at the coast.
The fellow was also trained on the methodology to estimate satellite altimeter bias drift using tide gauge observations and GPS estimates of VLM. In particular, the fellow has acquired new expertise in extracting and processing satellite altimetry observations using the Radar Altimeter Database System (RADS). This has enabled him to extract and compare sea level observations from tide gauge observations in order to estimate VLM.
The fellow finalized a new reprocessed GPS solution which includes more than 700 GPS stations globally distributed with data from 1995 onwards. The VLM estimates obtained from this new solution were compared for the first time against several solutions available from different groups, including the contribution to the last International Terrestrial Reference Frame realization (ITRF2014) coordinated under the second reprocessing campaign of the International GNSS Service. In this context, the fellow has created and currently chairs the Working Group 3.2 within the International Association of Geodesy addressing the comparison of existing GPS vertical velocity solutions.
Finally, the fellow has investigated the propagation of different systematic and stochastic errors in VLM corrections at TGs and their propagation into sea-level change reconstructions. A special focus was paid to separate the impact on global and regional sea-level change estimates. The errors considered covered a wide range of processes from unmodelled real and spurious land motion effects, reference frame uncertainty and deficiencies of GIA modelling. In particular, his research has addressed for the first time the global impact of the relative VLM errors between TGs and distant GPS stations.
The main conclusion of this research is that, under certain conditions including the GPS station location (vs loading and large earthquakes) and the time series length, the impact of GPS velocity errors on sea-level change estimates is well at the targeted level of 0.1 mm/yr. However, the GPS velocities rely on a reference frame which uncertainties, especially the stability of the origin drift along the rotation axis, propagate into sea-level change estimates by an error up to 0.2 mm/yr. Future research needs to concentrate on reducing the uncertainties related to the realization of the terrestrial reference frame.
The majority of sea level studies, including those that form the foundation for syntheses presented by the Intergovernmental Panel on Climate Change, use solely models of Glacial Isostatic Adjustment (GIA) to account for VLM at TGs. These models contain known biases over regional scales (e.g. Europe), and capture only one component of VLM, neglecting various processes that exist over different spatial and temporal scales. Direct observation of VLM using GPS is therefore vital. Despite many advances in this technique, systematic errors remain unaccounted for, limiting the precision that can be achieved. In addressing these packages, the objective is to reduce biases and uncertainty in sea level change estimates toward 0.1 mm/yr.
The objectives of this fellowship are:
- To improve the precision and accuracy of GPS position time series. This will have a major impact not only to the sea level community, but across a broad range of geosciences (e.g. surface mass loading, reference frames).
- To integrate these GPS estimates of VLM into the international satellite altimeter validation process, thus enabling accurate determination of altimeter drift biases.
- To derive a VLM velocity field for application to the global network of long-running high quality TGs to improve assessment of long term sea level change at global and regional scales.
This fellowship will provide high level training and international exposure from the leading geodesy and altimetry validation group at the University of Tasmania, Australia. It also represents an invaluable opportunity for knowledge transfer back to the University of La Rochelle, France, bringing a significant advance in the understanding of sea level change allowing the Fellow to continue his scientific development at the highest level in Europe.
Since the beginning of the project, the fellow has extended his data processing abilities being trained in new processing software (GIPSY/OASYS software package). Using the processing facilities and training at the University of Tasmania, the fellow has develop his own processing scheme in GIPSY/OASIS and has successfully reprocessed the International GNSS Service core network using the most updated models and corrections. The obtained training will allow the fellow to estimate the VLM of any GPS station accurately and faster than his previous abilities. For instance, the fellow has used this new expertise to investigate the coastal deformation of the Antarctic continent after a large earthquake and its implications for modelling the GIA. Under this context, the fellow has also received training in modelling the solid Earth response and the computation of surface deformation from co- and post-seismic events. The fellow has also investigated the impact of interannual deformation from modelled surface fluid mass changes on the estimated GPS velocities at global scale. In particular, the fellow bounded current GPS vertical velocity errors used to correct decadal to century-long tide gauges and identified regions where longer GPS time series are needed to mitigate this effect.
The fellow has also investigated the effects of far-field multipath in the GPS observations, which is one of the main systematic errors still present in state-of-the-art GPS time series. While investigating this topic, the fellow found a new way to use observations affected by multipath transforming this problem into an innovative method to remotely and continuously level co-located GNSS and tide gauge stations. Using this new GNSS reflectometry technique, the fellow has obtained significant results including a new technique that provides the sea level community with a means to tying the vertical position of the reference points of co-located tide gauge and GNSS stations remotely, precisely and at no added cost. This method has the potential to solve the problem of the missing levelling connections at TGs, enabling further research on the connection of national height datums and mean dynamic topography of the sea surface at the coast.
The fellow was also trained on the methodology to estimate satellite altimeter bias drift using tide gauge observations and GPS estimates of VLM. In particular, the fellow has acquired new expertise in extracting and processing satellite altimetry observations using the Radar Altimeter Database System (RADS). This has enabled him to extract and compare sea level observations from tide gauge observations in order to estimate VLM.
The fellow finalized a new reprocessed GPS solution which includes more than 700 GPS stations globally distributed with data from 1995 onwards. The VLM estimates obtained from this new solution were compared for the first time against several solutions available from different groups, including the contribution to the last International Terrestrial Reference Frame realization (ITRF2014) coordinated under the second reprocessing campaign of the International GNSS Service. In this context, the fellow has created and currently chairs the Working Group 3.2 within the International Association of Geodesy addressing the comparison of existing GPS vertical velocity solutions.
Finally, the fellow has investigated the propagation of different systematic and stochastic errors in VLM corrections at TGs and their propagation into sea-level change reconstructions. A special focus was paid to separate the impact on global and regional sea-level change estimates. The errors considered covered a wide range of processes from unmodelled real and spurious land motion effects, reference frame uncertainty and deficiencies of GIA modelling. In particular, his research has addressed for the first time the global impact of the relative VLM errors between TGs and distant GPS stations.
The main conclusion of this research is that, under certain conditions including the GPS station location (vs loading and large earthquakes) and the time series length, the impact of GPS velocity errors on sea-level change estimates is well at the targeted level of 0.1 mm/yr. However, the GPS velocities rely on a reference frame which uncertainties, especially the stability of the origin drift along the rotation axis, propagate into sea-level change estimates by an error up to 0.2 mm/yr. Future research needs to concentrate on reducing the uncertainties related to the realization of the terrestrial reference frame.