Abstract
Compared to the ITSG-Grace2016 release, multiple improvements within the processing chain have been implemented: updated background models, co-estimation of tides and stochastic modeling of satellite orientation measurements.
The ITSG-Grace2018 release is based on Level-1B Release 03 data and the AOD1B Release 06 dealiasing product. It includes unconstrained monthly and Kalman smoothed daily solutions, as well as a static gravity field, covering the complete GRACE observation period. Inaccuracies in background models contribute a significant amount of noise towards the overall error budget of this solution. We mitigate this effect by co-estimating unmodeled signals at various tidal frequencies, and high-frequency temporal variations in the background models through constrained daily gravity field solutions. The constraints used for the combined estimation of daily gravity field variations are now based on improved error estimates for the dealiasing models.
In an effort to better model all known error sources, we propagate orientation uncertainties derived from star camera/angular acceleration sensor fusion to the antenna offset correction for the K/Ka-Band observation. This enables the disentanglement of the stationary noise of the K-Band system and the non-stationary noise of the antenna offset correction. A model for bias jumps in the K/Ka-Band observations at the day-night boundary is introduced.
Investigations indicate a noise reduction within the monthly solutions of about 20-30 percent. The reprocessed release is presented and selected parts of the processing chain, as well as their effect on the estimated gravity field solutions, are discussed.
The ITSG-Grace2018 release is based on Level-1B Release 03 data and the AOD1B Release 06 dealiasing product. It includes unconstrained monthly and Kalman smoothed daily solutions, as well as a static gravity field, covering the complete GRACE observation period. Inaccuracies in background models contribute a significant amount of noise towards the overall error budget of this solution. We mitigate this effect by co-estimating unmodeled signals at various tidal frequencies, and high-frequency temporal variations in the background models through constrained daily gravity field solutions. The constraints used for the combined estimation of daily gravity field variations are now based on improved error estimates for the dealiasing models.
In an effort to better model all known error sources, we propagate orientation uncertainties derived from star camera/angular acceleration sensor fusion to the antenna offset correction for the K/Ka-Band observation. This enables the disentanglement of the stationary noise of the K-Band system and the non-stationary noise of the antenna offset correction. A model for bias jumps in the K/Ka-Band observations at the day-night boundary is introduced.
Investigations indicate a noise reduction within the monthly solutions of about 20-30 percent. The reprocessed release is presented and selected parts of the processing chain, as well as their effect on the estimated gravity field solutions, are discussed.
| Original language | English |
|---|---|
| Publication status | Published - 10 Dec 2018 |
| Event | AGU Fall Meeting 2018 - Walter E. Washington Convention Center, Washington, D.C., United States Duration: 10 Dec 2018 → 14 Dec 2018 https://fallmeeting.agu.org/2018/ |
Conference
| Conference | AGU Fall Meeting 2018 |
|---|---|
| Country/Territory | United States |
| City | Washington, D.C. |
| Period | 10/12/18 → 14/12/18 |
| Internet address |
Fields of Expertise
- Sustainable Systems