Tuna, Şahin Çağlar2026-04-072026-04-0720261570-761X1573-145610.1007/s10518-025-02349-w2-s2.0-105026498315https://hdl.handle.net/123456789/14144https://doi.org/10.1007/s10518-025-02349-wThis study presents a high-resolution, site-specific probabilistic framework for assessing soil liquefaction hazard by explicitly modeling subsurface spatial variability and seismic hazard. Gaussian Random Field (GRF) modeling and Monte Carlo simulation were used to generate 1,000 realizations of cone tip resistance (qc) and sleeve friction (fs), calibrated from 39 Cone Penetration Tests (CPTs). Liquefaction triggering was evaluated using the CSR-CRR procedure, and the depth-integrated Liquefaction Potential Index (LPI) was computed for each realization. Monte Carlo simulations were performed under a fixed seismic demand (PGA = 0.174 g) to isolate the effects of soil spatial variability, while seismic hazard uncertainty was subsequently incorporated by convolving site-specific fragility functions with PGA-based hazard curves TBDY ( Turkish Building Earthquake Code, 2018). Logistic regression produced site-calibrated fragility relationships between LPI and the probability of liquefaction (P[Liq]), enabling multi-threshold evaluations for LPI > 15, 25, and 45. The risk density peaked near PGA approximate to 0.18 g, indicating that moderate ground motions dominate liquefaction risk. The annual probability of exceeding LPI > 15 was approximately 1.92 x 10(-3) (approximate to 0.192%/yr; similar to 1 in 521 years), indicating a non-negligible annual liquefaction hazard. This framework enhances realism in liquefaction risk estimation and provides actionable guidance for performance-based geotechnical design, seismic mitigation, and hazard-informed infrastructure planning in regions susceptible to soil liquefaction.eninfo:eu-repo/semantics/closedAccessFragility CurvesSoil LiquefactionGaussian Random FieldPerformance-Based DesignMonte Carlo SimulationSpatial VariabilityArticle