How Do Atmospheric Tidal Loading Displacements Vary Temporally as well as Across Different Weather Models?

Abstract

AbstractWe assess the impact of varying the mass anomaly sources on the calculation of atmospheric tidal displacement harmonics. Atmospheric mass anomalies are obtained from five state-of-the-art numerical weather models (NWM): DWD’s ICON-Global, ECMWF’s IFS, JMA’s JRA55, ECMWF’s ERA5, and NASA’s MERRA2. To evaluate how the atmospheric tides’ representation in the different models displaces Earth’s crust, we calculate mass harmonics based on a fixed time span (2019.0–2022.0). To evaluate how temporally variable atmospheric tide manifestations are, we also applied a square-root-information filter on displacements spanning seven decades of ERA5. In addition, the variable harmonic atmospheric forcing is used to excite harmonic sea-surface variations employing the barotropic model TiME. The results from the analysis of the five numerical weather models as well as the monthly updated states of ERA5 harmonics are compared. We find that inter-model differences are larger than temporal harmonic modulations for all waves beating at frequencies higher than 1 cpd. We have confirmed that significant modulations are not an artefact in NWM but rather a true effect, and accounting for them might become of relevance for space geodesy at some point as soon as observations increase in spatio-temporal density and accuracy. The global RMS of radial displacements is 0.07 mm (SNR of 16.2 dB) for the “epoch” ensemble and 0.10 mm (SNR of 8.9 dB) for the “NWM” ensemble. We find discrepancies as large as 0.28 mm between harmonics from MERRA2 and early ERA5 batches, which we attribute to data sparsity in the in situ data assimilated into the NWM during the earlier years of the atmospheric reanalysis.