The Tanlu fault zone, extending over 2400 km from South China to Russia, is one of the most conspicuous tectonic elements in eastern Asia. In this study, we processed the Global Positioning System (GPS) measurements of Anhui Continuously Operating Reference System (AHCORS) between January 2013 and June 2018 to derive a high-precision velocity field in the central and southern segments of the Tanlu fault zone. We integrated the AHCORS data with those publicly available for geodetic imaging of the interseismic coupling and slip rate deficit distribution in the central and southern segments of the Tanlu fault zone. This work aims at a better understanding of strain accumulation and future seismic hazard in the Tanlu fault zone. The result indicates lateral variation of coupling distribution along the strike of the Tanlu fault zone. The northern segment of the Tanlu fault zone has a larger slip rate deficit and a deeper locking depth than the southern segment. Then, we analyzed three velocity profiles across the fault. The result suggests that the central and southern segments of the Tanlu fault zone are characterized by right-lateral strike-slip (0.29–0.44 mm/y) with compression components (0.35–0.76 mm/y). Finally, we estimated strain rates using the least-squares collocation method. The result shows that the dilatation rates concentrate in the region where the principal strain rates are very large. The interface of extension and compression is always accompanied by sudden change of direction of principal strain rates. Especially, in the north of Anhui, the dilatation rate is largest, reaching $3.780\times {10}^{-8}/a$. Our study suggests that the seismic risk in the northern segment of the Tanlu fault zone remains very high for its strong strain accumulation and the lack of historical large earthquakes.

Precise point positioning (PPP) has become a key technique in global navigation satellite systems (GNSS) applications, but its use in real time is limited by long convergence times. Tropospheric delay is one of the major contributors to this limitation. Accurate modeling over China is particularly challenging due to pronounced topographic variations, where empirical or conventional polynomial models fail to capture spatiotemporal characteristics. We propose a height-correction adjusted spherical harmonic model (HASH-Z) that incorporates an exponential height-decay term into a regional spherical cap expansion to account for the vertical dependence of zenith wet delay (ZWD). The model is developed from ZWD estimates obtained with PPP ambiguity resolution (PPP-AR) at a GNSS network in China (∼200 km spacing). Validation against IGS reference products confirms that centimeter-level accuracy of PPP-AR-derived ZWD estimates. Across the modeling stations, root mean square errors (RMSE) remain below 2.0 cm for all sites in winter and for more than 83% of sites in summer. A latitude-dependent variability is evident in winter, but largely obscured in summer due to pronounced atmospheric water vapor fluctuations. Independent tests at stations spanning large height differences further confirm the robustness of the model, with RMSE values of 0.5–1.9 cm in winter and 1.2–2.9 cm in summer. When applied in PPP-AR augmentation, the HASH-Z can significantly accelerate convergence by ∼20% in the vertical component and improve positioning accuracy by 4–12%. These results demonstrate that the HASH-Z model can provide an accurate and efficient representation of tropospheric delay over China.