Mon Oct 13, F. Flechtner
• Sorted by Date • Sorted by Last Name of First Author •
Abrykosov, Petro and Pail, Roland, 2025. Demonstrating the potential for the reduction of temporal aliasing through tailored stochastic modelling of non-tidal atmosphere and ocean model uncertainties in closed-loop simulations. Journal of Geodesy, 99(7):54, doi:10.1007/s00190-025-01980-4.
• from the NASA Astrophysics Data System • by the DOI System •
@ARTICLE{2025JGeod..99...54A,
author = {{Abrykosov}, Petro and {Pail}, Roland},
title = "{Demonstrating the potential for the reduction of temporal aliasing through tailored stochastic modelling of non-tidal atmosphere and ocean model uncertainties in closed-loop simulations}",
journal = {Journal of Geodesy},
keywords = {GRACE, GRACE Follow-on, Stochastic Modelling, Time-variable gravity, Mathematical Sciences, Statistics},
year = 2025,
month = jul,
volume = {99},
number = {7},
eid = {54},
pages = {54},
abstract = "{The imperfections of geophysical background models (BM) widely applied
in GRACE/GRACE-FO data processing pose one of the primary
limitations towards the gravity field retrieval performance.
With regard to ocean tide (OT) models, it could be shown that
incorporating prior knowledge on the spatial distribution of
uncertainties in terms of error variance-co-variance matrices
(VCMs) has the potential to reduce temporal aliasing in
designated spectral bands. It is therefore reasonable to assume
that the same approach can be beneficial for the mitigation of
aliasing related to errors within models representing the non-
tidal atmospheric and oceanic (AO) contributions. Unlike in the
case of OT, however, the uncertainties of the AO components
feature variations not only in space, but also in time. In this
contribution, we propose a method for the derivation of
stationary and non-stationary error VCMs on the basis of the
AOe07 time series, as well as the methodology for their
respective application in the data processing chain on the basis
of error propagation. The added value of these error VCMs is
assessed in a series of numerical closed-loop simulations for a
GRACE-type mission scenario. The impact of these error VCMs is
studied with respect to their spatial resolution as well as the
extent of correlation between model samples, and also in
combination with the stochastic information of other error
sources (OT, sensor noise). It is shown that in a best-case
scenario, the combined stochastic modelling of BM errors can
reduce the retrieval error by 35\% on average when applying
stationary error information for AO, and by 60\% when applying
non-stationary error VCMs. In a more realistic scenario where a
mismatch between the observed and stochastically modelled error
is introduced, the improvements are in comparison smaller, but
nevertheless constitute 10 and 18\%, respectively. It is also
shown that the joint stochastic modelling of all error sources
is crucial to improve the gravity solution, while applying
stochastic modelling only for individual contributors may even
degrade the performance. Additionally, it is demonstrated that
the inclusion of BM error models is applicable for a double-
pair-based gravity retrieval in the same manner.}",
doi = {10.1007/s00190-025-01980-4},
adsurl = {https://ui.adsabs.harvard.edu/abs/2025JGeod..99...54A},
adsnote = {Provided by the SAO/NASA Astrophysics Data System}
}
Generated by
bib2html_grace.pl
(written by Patrick Riley
modified for this page by Volker Klemann) on
Mon Oct 13, 2025 16:16:53
GRACE-FO
Mon Oct 13, F. Flechtner