Do ENSO-driven hydrological mass redistributions leave a detectable imprint on interannual Length-of-Day variations?

The El Niño–Southern Oscillation (ENSO) is a major driver of interannual climate variability, influencing terrestrial water storage (TWS) via atmospheric teleconnections and thereby affecting the Earth’s rotation through hydrological angular momentum (HAM) variations. To assess the relevance of these hydrological signals, we construct excitation budgets of the axial effective angular momentum and compare them with geodetic excitation from the IERS EOP 20 C04 series. On interannual time scales, the HAM contribution remains small due to the dominance of atmospheric angular momentum and partial compensation by barystatic ocean adjustments. Nevertheless, noticeable differences between the tested hydrological data sets are found. The hydrological model Open Source (OS) LISFLOOD slightly improves the budget closure, while the Land Surface Discharge Model (LSDM) reduces the agreement with geodetic excitation. We, therefore, revisit regional ENSO–TWS relationships and their contribution to axial HAM variability by comparing model-based TWS (OS LISFLOOD and LSDM) with respect to the satellite-based TWS observations from the GRACE/GRACE-FO missions. Lagged cross-correlation analyses identify regions with significant and temporally coherent responses to ENSO, from which basin-integrated HAM functions are derived. An EOF decomposition is used to quantify their contribution to hydrologically excited Length-of-Day (LOD) variability. The results reveal a robust ENSO-related signal in interannual HAM variability, dominated by a small number of tropical basins, with the Amazon basin emerging as the most influential contributor across all data sets. This indicates a low effective spatial dimensionality of the ENSO–HAM coupling. Among the data sets, OS LISFLOOD provides the best agreement with interannual variability, likely reflecting its more detailed representation of hydrological processes, whereas the older LSDM exhibits artificial variability linked to changes in atmospheric forcing of ECMWF operational data and the outdated ERA40 and ERA-Interim reanalyses. GRACE/GRACE-FO confirms the large-scale ENSO signal but is limited by the relatively short observational record.