Detectability of Coseismic Offsets and Early Post-Event Coordinate Changes from Public GNSS Products: The 2024 Mw 7.4 Hualien Earthquake

This is the authors' abstract. We don't add key points for this paper.

Public Global Navigation Satellite System (GNSS) products offer rapid access to earthquake deformation, but sparse sampling and shared processing constrain the interpretation of short records. We evaluate the 2024 Mw 7.4 Hualien earthquake using official coseismic offsets, five-minute positions, and daily coordinate series. Fixed-window estimates recover the main coseismic displacement pattern, providing a product-consistency check rather than independent validation. Spatial-block resampling reveals substantial uncertainty in the vertical cross-product comparison. For the first fortnight, we compare constant, linear, logarithmic, and exponential descriptions and distinguish nonconstant variation from resolvable curvature. The linear alternative is essential: most screened coordinate changes do not require a curved transient. Ten components retain screened status under the tested stability, seasonal/step, and common-mode diagnostics, but this conditional retention does not establish a physical deformation mechanism. Under the specified correlated-noise experiment, the intermediate-screen count is not unusual (exceedance probability 0.255); this experiment does not calibrate the complete final-retention procedure. Public products therefore provide useful coseismic consistency information, whereas short-record post-event changes remain sensitive to background treatment and cannot establish a spatially coherent postseismic mechanism.