Accurate modeling of traveling ionospheric disturbances (TIDs) is essential for characterizingtheir spatiotemporal variations and mitigating their effects on Global Navigation Satellite System (GNSS)precise positioning. This study develops a regional medium‐scale TID (MSTID) propagation model usingBeiDou geostationary orbit (GEO) total electron content (TEC) data. First, we propose a singular spectrumanalysis (SSA)‐based interpolation method to reconstruct the possible data gaps of BDS GEO TEC time series,enhancing its TID detection stability. Second, we select well‐distributed GNSS reference stations in HongKong, grouping them into triplets to estimate average TID parameters (main frequency, amplitude, velocity, anddirection) using cross‐spectral analysis. Finally, a regional MSTID propagation model is constructed using acosine function, incorporating the average phases, frequency, and amplitudes. The model was firstly validatedthrough the measured TID signal from the HKSC, HKST, and HKKT stations on 29 October 2018. Resultsindicate strong agreement between the model‐estimated and measured TID signals. Validation results show themodel reduces TID‐induced ionospheric delay errors by 48.2%, 50.3%, and 42.7% without accounting for thetime lag and further by 55.0%, 53.1%, and 45.1% with optimal temporal alignment for HKSC, HKST, andHKKT, respectively. Another MSTID event on 30 April 2016, was also selected to further validate the generalapplicability of the proposed TID model under different propagation characteristics. It confirms the model'saccuracy and reliability. This study provides an effective approach for precise TID characterization and supportsmitigation strategies for GNSS positioning errors caused by traveling ionospheric disturbances.