Transitioning from Deterministic Active Fault Maps to Probabilistic Models for Seismic Hazard Assessment: A Global Review and the Bursa-Eskişehir Fault System (Türkiye) Case Study

Transitioning from Deterministic Active Fault Maps to Probabilistic Models for Seismic Hazard Assessment: A Global Review and the Bursa-Eskişehir Fault System (Türkiye) Case Study | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Transitioning from Deterministic Active Fault Maps to Probabilistic Models for Seismic Hazard Assessment: A Global Review and the Bursa-Eskişehir Fault System (Türkiye) Case Study Volkan Karabacak, Çağlar Özkaymak, Ökmen Sümer This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8404630/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Active fault maps provide the geometric backbone for probabilistic seismic hazard assessment (PSHA) and surface-fault-rupture considerations, yet what qualifies as an “active fault” varies markedly among countries and use-cases. Here we compile and compare 35 widely used mapping protocols and databases, evaluate current practice in Türkiye, and propose a harmonization framework that (i) makes the selected time window and classification logic explicit and (ii) encodes evidence quality and geometric/segmentation uncertainty in a traceable way. As a case study we apply the workflow to the multi-segment Bursa–Eskişehir Fault System (BEFS) and calculate a segment-based Cycle Ratio, CR = elapsed time since the most recent event / median recurrence interval, as a relative indicator of seismic-cycle position. In BEFS, CR spans from very low values (e.g., the Bursa segment, ~0.08) to values near or above unity (e.g., the İnegöl West segment, ~1.07), highlighting strong along-strike variability in inferred stress accumulation. We emphasize that CR is not an earthquake-timing forecast; rather, it provides an auditable prioritization layer that can be combined with slip rate, segmentation uncertainty, and event-history constraints when building time-dependent source models. The proposed workflow offers a reproducible path for incorporating Türkiye’s rapidly growing paleoseismological record into next-generation seismogenic source databases. Active fault mapping Probabilistic Seismic Hazard Assessment (PSHA) Seismic cycle Bursa–Eskişehir Fault System Time-dependent hazard Türkiye 1. Introduction The success of earthquake hazard and risk mitigation strategies depends largely on the reliability of the answer to the question: "Where, when, and how will the earth's crust rupture?" The geological equivalent of this answer lies in the accurate definition of the spatial distribution, geometry, segmentation architecture, and earthquake-generating characteristics of active faults. Today, Probabilistic Seismic Hazard Analyses (PSHA) and engineering design spectra directly utilize source zoning and parameters such as fault length, slip type, slip rate, recurrence interval (RI), and maximum credible earthquake magnitude (M max ) as inputs (Cornell, 1968; McGuire, 2004). While these parameters are essential for modern hazard models, traditional deterministic maps often provide only spatial location ('where'), lacking the critical temporal and parametric data ('when' and 'how often') required to reduce epistemic uncertainties in PSHA. Therefore, bridging the gap between geological fieldwork and engineering demand requires a shift from static inventories to dynamic source models. Similarly, site selection decisions for land-use planning and critical infrastructure projects must consider not only ground acceleration hazards but also "near-fault" hazards such as surface faulting and permanent displacement. In this context, active fault maps serve as the fundamental "interface" where pure geological information is translated into engineering parameters and planning decisions. However, this critical interface is not bound by a single international terminology or a standardized time window. The approaches of countries are shaped more by regulatory needs and risk perception than by scientific consensus. While some countries accept Holocene activity (approximately the last 11,000 years) as a "sufficiently young and well-defined" threshold for regulatory zoning purposes, others adopt a wider geological time window on the scale of the Late Pleistocene. On the other hand, international databases focus on the earthquake-generating potential of faults through a PSHA-compatible "seismogenic source" approach rather than age thresholds. For instance, on a global scale, the GEM Global Active Faults Database (GAF-DB) aims to gather active fault geometries and attributes under a single roof by combining regional compilations (Styron and Pagani, 2020), whereas the European-scale EFSM20 and ESHM20 frameworks structure active faults to serve directly as geological inputs for probabilistic hazard models (Basili et al., 2022). Türkiye, located in a high deformation rate sector of the Eastern Mediterranean–Anatolian transition zone within the Alpine–Himalayan belt, faces high seismic risk potential alongside the geological diversity brought by this tectonic context (Fig. 1a). Destructive earthquakes causing tens of thousands of casualties in the last century, and most recently the February 6, 2023 Kahramanmaraş earthquakes, have once again revealed the catastrophic scale this natural phenomenon can reach. The first and most fundamental stage of reducing earthquake damage is the mapping and characterization of the active faults generating the earthquakes in the field. However, today this need extends far beyond merely drawing fault traces on a map. The standard now required is the reporting of parametric information, such as segment structure, rupture zone geometry, short and long-term slip rates, recurrence intervals, and displacement amounts per event together with data quality and uncertainties. The first holistic national-scale mapping initiative in Türkiye began with the 1:1,000,000 scale Active Fault Map of Türkiye published by the General Directorate of Mineral Research and Exploration (MTA) in 1992, which was updated at different scales between 2004 and 2011 in light of advances in earthquake geology and increasing data accumulation (Emre et al., 2013). Following this process, a "Türkiye Paleoseismology Research" vision was developed to systematically investigate paleoseismological behavior; particularly in the post-2023 period, parametric data production on priority faults and segments has gained momentum within the framework of a broad-participation platform. This paper addresses this outlined framework through two complementary datasets: (1) A comprehensive inventory analysis compiling and classifying global active fault mapping protocols (Supplementary File 1), and (2) An application example based on segment-based paleoseismological event sequences, recurrence interval (RI), and "cycle ratio" indicators on the Bursa–Eskişehir Fault System (BEFS) (Fig. 1b) (Supplementary File 2). The primary objective of this study is not to criticize Türkiye's existing mapping standard, but to propose a gradual and practicable scientific framework that can elevate the rapidly growing paleoseismological and parametric database to a "probabilistic and parametric" mapping level compatible with modern international approaches. Importantly, the mapping framework proposed in this paper is not intended to replace the existing deterministic Active Fault Map of Türkiye produced by MTA, which remains the authoritative national baseline for fault‐trace location and age‐based activity classes (Emre et al., 2013, 2018). Instead, consistent with international practice where multiple active‐fault products coexist for different end‐users and decision contexts (Supplementary File 1), we propose an additional, risk‐oriented cartographic layer that goes beyond trace depiction by translating fault behavior and its uncertainties (e.g., segmentation confidence, recurrence statistics, and cycle‐stage indicators) into a PSHA‐compatible prioritization interface. Accordingly, the proposed classification should be read as a complementary product designed to sit alongside trace maps and regulatory zoning tools, not as an alternative to them. FIGURE 1 2. Conceptual Framework: Time Windows and Classification Types In the discipline of active fault mapping, two fundamental methodological parameters shape the philosophy of map creation and its ultimate intended use: (i) The time window, which determines the geological time interval of activity upon which a fault is deemed "active," and (ii) The classification type, which defines the systematic logic by which this activity information is coded and reported. 2.1. Time Window Approaches Time windows most frequently referenced in literature and practice can be categorized according to the regulatory or scientific purposes of the mapping: Holocene-Focused (10–12 ka): This is the narrowest time interval, preferred especially in applications yielding direct legal/regulatory consequences such as construction restrictions, the establishment of surface faulting setback zones, and zoning planning. For example, the Alquist–Priolo Earthquake Fault Zoning Act in California (conservation.ca.gov) and similar national regulations (Sözbilir et al., 2018) base their approach on this window to minimize construction risk on active fault traces. Late Pleistocene / Late Quaternary (100–130 ka): Covering not only the last 10,000 years but also the last glacial and interglacial periods, this approach offers a broader "recent geological time" perspective to avoid overlooking faults with long seismic recurrence periods. Quaternary (2.58 Ma): This is the most widely accepted upper limit in national-scale fault inventories and global databases. However, within this broad time interval, the level of evidence and the precision of separation into sub-epochs (Pleistocene/Holocene) can vary between regions. Pliocene–Quaternary (~5 Ma) and Broad-Scale Neotectonic Windows: These broad time intervals, generally used in active tectonic atlases and regional geodynamic compilations, aim to understand the evolution of the current deformation regime. This diversity stems from the "purpose-oriented" nature of mapping rather than scientific inconsistency. Indeed, there is no obligation for zoning maps constituting the basis of spatial planning and PSHA-based source models to use the same time thresholds. 2.2. Classification Types: Age, Evidence, and Source Modelling The global inventory analysis within Supplementary File 1 indicates that four main methodological approaches are adopted in the classification of active faults: Age-Based (AGE): The most basic approach where classification seeks to answer only whether the "last movement occurred within a specific time window." Age + Evidence/Confidence (AGE+EVID / AGE+CONF): A more qualified classification type where, in addition to chronological information, the type of data documenting the activity (paleoseismology, geomorphology, historical/instrumental seismicity, etc.) and the level of confidence in the data are coded. Capable Fault Approach: Used especially in critical infrastructure projects such as nuclear power plants and dams, this approach focuses not only on the fault's past activity but also on its "potential to generate surface displacement/rupture in the near future" (Machette, 2000). Seismogenic Source (SOURCE-MODEL): The most advanced level where the fault is modeled as a "seismogenic source" that can be used directly as input in PSHA. Here, in addition to fault trace geometry, numerical parameters such as slip rate, moment balance, maximum magnitude (M max ), and segment connectivity are integrated into the system. The European-scale EFSM20 and ESHM20 projects are current and successful examples of this approach (Basili et al., 2022). 3. Dataset and Methodological Framework This research relies on two fundamental datasets that complement each other: a comparative analysis of global active fault protocols and a regional application example (i.e. BEFS). The global dataset (Supplementary File 1), forming the first pillar of the study, was created by compiling 35 national/institutional active fault maps and databases from at least 30 different countries and regions. Each product in this inventory is presented in tables coded based on time window ("time_bin") and classification type ("class_bin") to make them suitable for analysis. The "time_bin" coding used in the analysis of the global inventory allows for the comparison of different national approaches. Examination of the inventory distribution reveals a predominance of Quaternary-focused products (QUAT, approx. 15 maps), followed by Late Quaternary/Late Pleistocene scale (LQ, LP100/LP125) studies, PSHA-focused source models (SRC), and Holocene-focused regulatory approaches (HOLO). This distribution suggests that there is no "single correct time window" in active fault mapping; instead, the combination of "purpose, scale, and risk tolerance" is the determining factor. Similarly, "class_bin" codes, which examine whether maps carry only fault traces or attribute and modeling capacity, reveal that basic inventory types like AGE+MAP (approx. 10 maps) are common, but SOURCE-MODEL (3) structures that can be directly integrated into seismic hazard analyses (PSHA) remain more limited. Furthermore, the scarcity of harmonized multi-source inventories (HARMONIZED) indicates that the need for standardization in international data integration persists. The BEFS example (Supplementary File 2), constituting the regional application part of the study, includes segment-based paleoseismological event sequences for the Ulubat, Bursa, İnegöl, and Eskişehir faults, recurrence interval (RI) estimates obtained via Bayesian age modeling, and the time elapsed since the last event/RI ratio (cycle ratio). Event chronologies and age constraints used in this scope were compiled from Karabacak et al. (2021) for the Ulubat Fault, Karabacak & Sancar (2025) for the Bursa and İnegöl faults, and Elma et al. (2024, 2025) and Karabacak et al. (2025) for the Eskişehir Fault. In the BEFS analyses, two main indicators were utilized considering the available datasets: 1) Recurrence Interval (RI), representing the median and uncertainty range of the segment's paleoseismic recurrence period, and 2) Cycle Ratio (CR), representing the ratio of the time elapsed since the last surface rupture to the RI. In this methodology, a cycle ratio of ~1 and above is interpreted as a relative prioritization indicator signifying that the segment has entered the late stage of its cycle; the 0.5–1 range indicates increasing stress accumulation, while a value <0.5 signifies the segment is in a relatively early stage (Supplementary File 2). This approach ensures that seismic hazard is addressed with intervals containing uncertainty rather than deterministic singular values. However, the method has a critical limitation; the cycle ratio serves to create a comparative risk window for decision-makers rather than definitively "dating" the expected earthquake, offering a statistical risk projection rather than a claim that an earthquake will occur soon. To translate the segment-based 'Cycle Ratio' (CR) indicators that defined as the ratio of elapsed time to the median recurrence interval into a spatially coherent and risk-oriented cartographic model, a statistical classification method was adopted. Instead of utilizing arbitrary linear intervals, the distribution of CR values derived from the Türkiye-wide paleoseismological inventory (Supplementary File 3, Table 1) was subjected to the Fisher–Jenks natural breaks optimization algorithm (Fisher, 1958; Jenks, 1967). The Fisher–Jenks optimization was specifically selected over arbitrary linear intervals to minimize intra-class variance and maximize the deviation between class means. This ensures that the proposed risk categories (C1–C5) reflect the inherent statistical clustering of the stress accumulation stages derived from the national paleoseismological inventory, providing an objective basis for prioritization. The optimization analysis indicated that a four-class solution for the pre-saturation interval (0 < CR ≤ 1.0) yielded the highest 'Goodness of Variance Fit' (GVF ≈ 0.97), providing the most representative delineation of the seismic cycle from early post-seismic relaxation to the late accumulation phase. Consequently, the mapping legend was structured into six distinct categories: an 'undefined' class (C0) for data-poor segments, four sequential classes for the standard cycle (C1–C4), and a specific 'limit exceeded' or 'overdue' category (C5) for segments where the elapsed time exceeds the characteristic recurrence interval (CR > 1.0) (Supplementary File 3, Table 2). This classification architecture allows the active fault map to evolve from a static trace inventory into a graded priority interface suitable for decision-making processes. 4. Global Active Fault Mapping Protocols: Trends and Distinctive Examples When the general landscape of active fault mapping studies created in different countries on a national and institutional basis (Supplementary File 1) is examined, it is observed that studies diversify along three main axes (purpose, scale, and modelling level) rather than following a standard template. While the ultimate goal of mapping (regulatory zoning, national hazard modelling, or academic/geodynamic inventory production) determines the quality of the product, the data resolution difference between 1:1,000,000 scale continental/country compilations and 1:5,000 local scale zoning maps reveals the heterogeneous structure of the inventory. Similarly, basic maps offering only fault trace and kinematic class show distinct differences in terms of modelling maturity compared to systems hosting parameters such as slip rate, recurrence interval (RI), and maximum magnitude (M max ) in their databases. In the last decade, global and continental databases integrating national compilations have accelerated the search for standardization and a common language for this diversity. In this context, the GEM Global Active Faults Database (GAF-DB) gathers worldwide active faults under a single roof with attributes such as geometry, kinematics, slip rate, and references (Styron and Pagani, 2020), while the Active Faults of Eurasia Database (AFEAD) offers a highly detailed geodatabase by reporting tens of thousands of fault objects across Eurasia in a consistent structure (Basili et al., 2022). The fundamental contribution of these databases to the literature is moving the "which fault is active?" debate from solely a time window constraint to an axis of evidence type, parameter set, and uncertainty trio. However, it is a clear reality that these global initiatives require a comprehensive "metadata and harmonization" process for national data to become interoperable, rather than directly replacing national regulatory standards. From a regulatory framework perspective, the Alquist–Priolo Earthquake Fault Zoning Act in California constitutes a distinctive example with its approach to reducing surface faulting hazard. This protocol mandates the creation of "Earthquake Fault Zones" around active fault traces and requires construction decisions in these zones to be supported by field investigations (conservation.ca.gov). Associating "activity" mostly with Holocene activity and well-defined fault traces, this approach aims to minimize geological uncertainty via "on-site evidence" before reflecting it directly onto construction risk. On the other hand, approaches like EFSM20 and ESHM20 in Europe treat active faults directly as an input for probabilistic seismic hazard analyses (PSHA) (hazard.efehr.org). Here, faults are modelled not just as geographic traces but as seismogenic sources for which earthquake rate estimations can be made. Therefore, the primary need in this perspective is not only the location of the fault but also geometric and kinematic parameter sets allowing for moment balance and earthquake rate calculations. This transforms the fault mapping process from a deterministic "map" production into a probabilistic "model" construction. One of the best institutional examples of this transformation is the Quaternary Fault and Fold Database (QFault) operated by the USGS in the USA, which links the "Quaternary fault" definition to a systematic meta-data scheme, presenting fault geometry and segmentation alongside age/time of last movement, slip rate intervals, and literature references in a singular record logic (Machette et al., 2004). Thus, the inventory ceases to be merely a cartographic product and transforms into an updatable and traceable infrastructure providing data for hazard modeling. Similarly, Alquist–Priolo zones in California are an advanced regulatory tool connecting surface rupture hazard directly to fieldwork and engineering applications via a "legal obligation + on-site verification" combination, rather than being just a national database (California Geological Survey). The Active Fault Database of Japan, maintained by AIST/GSJ, incorporates "hazard degree" indicators such as behavioral segmentation of faults and slip rate classes into the classification within a framework based on Late Quaternary activity (120–130 ka) (AIST/GSJ; Research Group for Active Faults of Japan, 1991). New Zealand's NZAFD similarly defines activity within an approximately 125 ka window but exhibits close integration with guidelines for transferring surface rupture hazard to land-use decisions via recurrence interval and displacement parameters (GNS Science; Kerr et al., 2003). On the Eastern Mediterranean scale, platforms like NOAFaults and GreDaSS in Greece produce evidence/confidence degree classes by associating active fault geometry with the "seismogenic source" definition (Ganas et al., 2013; GreDaSS Consortium, 2014); in Italy, ISPRA's ITHACA approach demonstrates that a targeted data-criterion architecture can be applied for engineering site selection, specifically focusing on "capable fault" and surface faulting hazard (ISPRA). These examples emphasize that the two-layered classification centered on time_bin and class_bin proposed in Supplementary File 1 finds practical correspondence in mature systems through meta-data fields and uncertainty indicators. 5. Active Fault Mapping in Türkiye: The MTA Approach and Transforming Needs The first standardized source on a national scale for understanding the active fault architecture in Türkiye is the 1992 Active Fault Map of Türkiye by MTA (General Directorate of Mineral Research and Exploration). This map served as the primary reference for earthquake hazard and planning studies for many years. With methodological progress in earthquake geology and data increase, active fault maps were renewed at different scales between 2004–2011; the explanatory text and classification logic of the map were published by Emre et al. (2013, 2018). The current requirement is for these maps to become "parametric." In other words, the database must carry not only where the fault is, but also how it behaves and with what uncertainties its earthquake generation potential is defined. In this direction, the primary goal of paleoseismology-based studies conducted under MTA coordination is: (i) detailing structural/geometric features, (ii) determining historical/prehistoric earthquakes resulting in surface rupture, (iii) constraining recurrence intervals and slip rates, and (iv) transferring results to a national database. The platform approach, which accelerated after February 6, 2023, aims to close the paleoseismological data gap in numerous fault segments in Türkiye. For instance, an action was taken by bringing together Türkiye's state institutions to reduce earthquake risk. Implemented under the leadership of expert scientists, the "Platform" (P/SISMO-TURK Project, project no: 107G001)) targets the investigation of Türkiye's Active Faults in all aspects by nearly 200 scientists. This rapid data production brings two critical questions to the agenda simultaneously: How will this new data be integrated into deterministic "trace maps"? How will a "probabilistic" active fault classification/mapping be designed for PSHA and risk mitigation decisions? At this point, the main lesson emerging from the global inventory (Supplementary File 1) is this: Türkiye's need is not to select a single time window, but to design a multi-layered classification and harmonization parameters capable of serving both regulatory/planning and PSHA-focused usage through a single database. This multi-layered view also implies that the proposed probabilistic/parametric classes are designed to complement, not displace, deterministic trace mapping. In practical terms, the deterministic map continues to serve as the primary reference for fault-trace geometry and regulatory/planning workflows, whereas the additional layer developed here targets PSHA-driven hazard and risk applications (e.g., insurance, infrastructure prioritization, and scenario-based mitigation) by explicitly incorporating parameter ranges and uncertainty. This two-track architecture reflects the “portfolio” logic observed in mature mapping systems, in which trace inventories, capable-fault/zoning products, and source-model datasets are maintained together for different purposes (Supplementary File 1). 6. Case Study: Bursa–Eskişehir Fault System 6.1. Nature of the Study Area and Selection Criteria BEFS, extending from the south of the Marmara Region to the interior of Central Anatolia (Figs. 1b and 2a), qualifies as an ideal key application area for this study with its multi-segmented structure and tectonic diversity. While the Ulubat, Bursa, and İnegöl faults forming the western wing of the system are in direct interaction with dense industrial and residential areas (İ.e. Bursa, İnegöl and Eskişehir Cities), the Eskişehir Fault on the easternmost represents a critical tectonic corridor in the interior of Anatolia (Fig. 2a). This diversity allows for the combined evaluation of both high-risk urban centers and segments showing differences in data density. Furthermore, the selection of the BEFS is driven by data reliability and internal consistency. The co-authors of this study were directly involved in the geological mapping and paleoseismological trenching campaigns across these segments (e.g., Karabacak et al., 2021; Karabacak & Sançar, 2025; Elma et al., 2025). This first-hand engagement allows for expert control over the quality of the input data and the management of epistemic uncertainties regarding event horizons and chronological constraints. Additionally, as a priority focus area within the ongoing national 'P/SISMO-TURK' project, the BEFS ensures a continuous flow of high-resolution parametric data, making it an ideal dynamic laboratory for testing the proposed probabilistic framework. FIGURE 2 6.2. Segment-Based Recurrence Interval and Cycle Ratio Analyses Segment-based recurrence interval (RI), dates of last events, elapsed time, and cycle ratios (Elapsed Time/RI) compiled from the dataset detailed in Supplementary File 2 are presented in Table 1. These data constitute the basis for understanding the current positions of fault segments within the seismic cycle. The findings in Table 1 reveal two critical patterns regarding system characteristics. First, a distinct heterogeneity is observed within the system; cycle ratios within the same fault system range widely from very low values like 0.08 to critical levels like 1.07, indicating the completion of the seismic cycle. Second, as seen in the Ulubat East segment example, the generation of different scenarios (A and B) demonstrates the necessity of transparently managing uncertainty instead of a deterministic single-scenario approach. Table 1. RI and Cycle Ratio in the BEFS segments. Fault Segment RI (median; 5%–95%) (yr) Last Event Elapsed Time (median; 5%–95%) (yr) Cycle Ratio Ulubat Central segment 5502 (1804–7531) AD 1139 886 (875–903) 0.16 East segment – Scenario A 4181 (3975–4387) AD 175 1850 (1649–2052) 0.44 East segment – Scenario B 4019 (3649–4392) AD 13 2012 (1640–2381) 0.50 Bursa Bursa segment 2117 (579–5647) AD 1855 170 (168–172) 0.08 İnegöl West segment 2395 (948–3452) BC 541 2565 (2450–2724) 1.07 Eskişehir Segment 1 3278 (3121–3432) AD 542 ± 144 ~1480 0.45 Segment 2 3221 (2892–3552) BC 800 ± 100 ~2825 0.88 Segment 3 3046 (2148–3940) BC 886 ± 350 ~2900 0.95 Segment 4 3174 (2826–3523) AD 750 ± 50 ~1275 0.40 6.3. Fault System Scale Relative Stress Accumulation Classification The cumulative view of segment-based data at the fault system scale and the relative stress accumulation classification are summarized in Table 2. Analysis results point to an accumulation pattern that can be characterized as a late-stage in terms of the seismic cycle, particularly for the İnegöl West segment and certain sections of the Eskişehir Fault (especially Seg. 2 and Seg. 3). However, it must be strongly emphasized that this classification is not an earthquake timing prediction, but a data-based risk prioritization metric. Table 2. Summary of relative stress accumulation at the fault scale along the BEFS. Fault No. of Segments Median Cycle Ratio Max. Cycle Ratio Relative Classification (Stress Accumulation) Ulubat 3 0.44 0.50 Moderate Bursa 2 0.08 0.08 Low İnegöl 3 1.07 1.07 High (Limit Exceeded) Eskişehir 4 0.66 0.95 High 6.4. Stress Transfer and Multi-Segment Rupture Dynamics In multi-fault systems, seismic hazard cannot be reduced to the independent behavior of a single segment. As conceptually outlined by Field et al. (2014), the rupture probability of neighboring segments can change dynamically with stress transfer and inter-segment connectivity. Although paleoseismological data reduce uncertainty regarding the "renewal timing" of segments, geodesy (GNSS), seismotectonic modeling, and historical earthquake catalogs must be evaluated in an integrated manner to fully quantify stress transfer at the system scale. In this context, the BEFS example serves as a conducive pilot region for making the transition to "probabilistic active fault mapping" and the need for standardization visible in Türkiye, thanks to its data heterogeneity and high interaction with residential areas. 7. Discussion: Transition from Deterministic Mapping to Parametric Active Fault Modeling 7.1. Why "Probabilistic Active Fault Map"? The existing Active Fault Map of Türkiye has filled a significant gap in literature and practice by providing a standard answer to the question of "spatial distribution of source faults" on a country scale. However, today, the requirements of Probabilistic Seismic Hazard Analysis (PSHA) and risk mitigation strategies have moved beyond merely knowing the geographic location of the fault. Modern seismotectonic approaches mandate the reporting of parameters such as fault geometry, segmentation structure, slip rate, recurrence interval (RI), and maximum magnitude (M max ) together with their epistemic and aleatory uncertainties. This necessity means the active fault map must evolve from a static "inventory product" into a continuously updatable, dynamic "model input" for hazard analyses. Therefore, the transition from deterministic mapping to parametric and probabilistic mapping is not a preference but a scientific and practical necessity. 7.2. From Multi-National Comparison to National Standard: Harmonization Parameters The global inventory analysis presented in this study (Supplementary File 1) proposes a parameter set to ensure the interoperability of protocols in different countries rather than directly equating them. For Türkiye, a practicable harmonization core capable of meeting both regulatory zoning (surface faulting hazard) and PSHA-focused source modeling needs within a single national database can be gathered under the following headings: Time Window Label (time_bin): The chronological interval defining the fault's activity (e.g., QUAT, LQ, HOLO). Classification Label (class_bin): Coding determining the quality of the product (e.g., AGE+MAP, AGE+EVID, SOURCE-MODEL). Evidence Type and Quality (EVID/CONF): The source of data documenting activity (paleoseismology, geomorphology, historical/instrumental record, geodesy) and the confidence level attributed to this data. Geometry and Segmentation Confidence: Precision of fault trace location, uncertainties regarding segment boundaries, and presence of distributed deformation zones instead of singular fault traces. Parametric Fields: Short/long-term slip rates, recurrence interval (RI), single event displacement, fault width/depth, $M_{max}$ values, and statistical uncertainty ranges for all these parameters. Modeling Compatibility: Direct usability of data in PSHA processes (moment balance, earthquake occurrence rate models, logic-tree branches). Adopting this parameter set as a common language will enable the multi-purpose use of the national database. 7.3. Critical Requirement Pointed Out by the BEFS Example: Uncertainty Management Specific to the BEFS, the scenarios produced particularly for the Ulubat East segment represent a practical application of the probabilistic and parametric mindset in Türkiye (Supplementary File 2). The fundamental philosophy here is not to ignore uncertainty, but to model it and make it visible. This approach creates value in two main dimensions: Scientifically, it prevents data deficiency or interpretation differences from being masked by a "singular and definite value." Practically, it offers decision-makers (municipalities, infrastructure operators, insurance sector, and risk governance units) a more realistic risk projection containing "ranges and probabilities" instead of a deterministic judgment. 7.4. Limitations and Manageable Uncertainty This proposed transformation process requires the clear acceptance and management of certain limitations: Data Constraints: The number of paleoseismological trenches and dating data is still limited in many fault segments; this may cause uncertainty ranges to widen. Geomorphological Ambiguity: In areas where segment boundaries cannot be clearly traced geomorphologically or in distributed deformation zones, the concept of a mappable "single trace" may remain insufficient. Interpretation Risk: Derived indicators like Cycle Ratio are not definite earthquake timing predictions, but merely tools for relative prioritization among segments. There is a risk that this data may be misinterpreted as "an earthquake will definitely happen soon." Use-case boundary: The cycle ratio based classes and the resulting Türkiye-scale map are intended as a national-scale, comparative prioritization tool for hazard and risk governance. They are not a substitute for site-specific surface-fault-rupture zoning, engineering microzonation, or legal setbacks, which require higher-resolution mapping, targeted paleoseismological/geomorphic investigations, and on-site verification in the project area. Misuse outside this intended scale and purpose may lead to erroneous “safe/unsafe” interpretations. Consequently, clearly coding and reporting these limitations is not a weakness of probabilistic mapping, but strictly its strongest aspect. This is because uncertainty is thereby removed from being an unknown and transformed into a manageable risk parameter. 7.5. Transfer of Cycle Ratio Based Classification to Mapping at Türkiye Scale The rapid increase in paleoseismological trench data and recurrence interval (RI) solutions based on these data in Türkiye makes it possible to go beyond producing active fault maps solely with "trace/presence-absence" logic. However, on a national scale, data scope and quality are heterogeneous among segments; while RI and/or time of last event are well-constrained in some segments, uncertainties are high or basic parameters are undefined in others. Therefore, a new mapping example requires a simple yet traceable classification scheme that (i) offers a comparable prioritization metric to decision-makers, and (ii) clearly shows data deficiency and uncertainty. In the approach proposed in this study, the seismic cycle position for each fault/segment is expressed by the normalized "cycle ratio" (CR = elapsed time / RImedian). In the map legend, CR values in the 0-1 range are shown with sequential color tones, establishing a readable visual hierarchy from the early stage of the cycle (lighter tone) to the late/critical stage (darker tone). In determining class intervals, the Fisher-Jenks "natural breaks" approach, which best represents the natural clustering of the Türkiye-wide CR distribution ((Supplementary File 1), Table 1), was used, and solutions minimizing intra-class variance were preferred in selecting the number of classes (Fisher, 1958; Jenks, 1967; Jenks and Caspall, 1971). Within this framework, the 0-1 range was divided into four main classes; the CR>1 condition was highlighted separately with hatching/thick contours instead of color to distinguish the segments with CR > 1 (cycle-limit exceedance) (Table 3). The critical component of the proposed classification is keeping undefined/uncertain records as a separate category on the map (C0). Thus, the map makes visible not only high or low-priority segments but also gaps in terms of national scale parametric data production; this also functions as a prioritization tool for future paleoseismological and geodetic studies. In Figure 2b, the Cycle Ratio (CR) based national fault segment classification (Table 3), determined from paleoseismological data obtained from active faults across the entire Turkish mainland, has been adapted for the BEFS and its immediate surroundings, which were selected as the key study area, in order to make first Parametric Active Fault Map representation concrete for practical use. Finally, it should be emphasized that CR classes are designed not to produce an earthquake-timing prediction, but to offer a relative cycle-prioritization scale among segments. Therefore, in transferring classes to planning/PSHA applications, it is essential to evaluate them together with additional parameters such as evidence type, data confidence level, slip rate, segment connectivity, and multi-segment rupture probability. Table 3. Cycle Ratio (CR) based national fault segment classification and map representation (Adapted from Supplementary File 3, Table 2). Class Cycle Ratio (CR) Seismic Cycle Position Map Representation C0 Uncertain / Undefined Insufficient data (No RI or last event age, or unreliable) Gray tone; dotted/thin line C1 0.00-0.21 Very early stage of cycle (post-event) Light green tone C2 >0.21-0.40 Early stage (stress accumulation begins) Yellow tone C3 >0.40-0.75 Middle stage (accumulation evident) Orange tone C4 >0.75-1.00 Late stage / critical period (approach to RI) Red tone C5 >1.00 Cycle limit exceedance (CR > 1; RI threshold exceeded) Red tone with surrounded hatching and/or thick contour 8. Conclusions This paper reviews the practical diversity of “active fault” definitions and mapping standards worldwide, and uses that comparison to frame a set of minimum, transparent metadata fields for national fault inventories in Türkiye. The inventory of 35 protocols highlights that time window, activity criteria, and mapping purpose are commonly coupled, and that a single, universal “active” definition is rarely achievable across regulatory zoning, engineering applications, and PSHA-oriented source modeling. We then demonstrate the proposed workflow on the BEFS by compiling published paleoseismological constraints at the segment scale and deriving a Cycle Ratio (CR = elapsed time / RI_median) as a relative indicator of seismic-cycle position. The BEFS example shows strong along-strike variability, with low CR values for some segments (e.g., Bursa segment, ~0.08) and values approaching or exceeding unity in others (e.g., İnegöl West segment, ~1.07). CR should not be interpreted as a deterministic earthquake-timing statement. Its meaning depends on the quality of event chronologies, the representativeness of estimated recurrence intervals, and the adopted renewal-model assumptions. Accordingly, CR is best used as an auditable prioritization layer to identify where additional data, alternative segment scenarios, or sensitivity testing would most improve downstream hazard models. For practical transfer to seismogenic source databases, we recommend reporting (at minimum) the adopted time window and classification logic, evidence type and confidence, geometry/segmentation uncertainty, and parametric fields relevant to time-dependent modeling (e.g., RI distributions and, where available, slip-rate constraints). In this sense, the existing deterministic Active Fault Map of Türkiye remains a critical geometric baseline, while an uncertainty-coded parametric layer such as the one illustrated here can be maintained as a complementary product within the same evolving database. Declarations Ethics approval and consent to participate Not applicable. This study does not involve human participants, human data, or animals. Consent for publication Not applicable. Availability of data and materials All data and materials supporting the findings of this study are included within the manuscript and its Supplementary Files. Additional information can be provided by the corresponding author upon reasonable request. Competing interests The authors declare that they have no competing interests. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Authors’ contributions VK conceived the study, developed the methodological framework, compiled and analyzed the datasets, and wrote the first draft of the manuscript. Co-authors ÇÖ and ÖS compiled and analyzed the datasets, contributed to data interpretation, figure preparation, critical revision of the manuscript, and approved the final version. Acknowledgements The authors thank colleagues and institutions that provided access to published datasets and cartographic resources used in this study. We also thank the anonymous reviewers in advance for their constructive comments that will help improve the manuscript. Supplementary information Supplementary information accompanies this paper and is provided as Supplementary Files 1–3. ● Supplementary File 1: Global "Active Fault Mapping" protocols ● Supplementary File 2: Bursa–Eskişehir Fault System (BEFS) paleoseismological analysis ● Supplementary File 3: Paleoseismological based earthquake cycle analysis of Türkiye's Active Faults Use of AI and AI-assisted technologies The authors declare that no generative AI or AI-assisted technologies were used to generate scientific content, interpret results, or draw conclusions. Digital tools were used for language editing and formatting only, and the authors take full responsibility for the integrity and originality of the work. References Basili, R., Danciu, L., Beauval, C., Sesetyan, K., Vilanova, S. P., Adamia, S.,... & Zupančič, P. (2022). European Fault-Source Model 2020 (EFSM20): online data on fault geometry and activity parameters. Istituto Nazionale di Geofisica e Vulcanologia (INGV). https://doi.org/10.13127/efsm20 Begg, J.G., Mouslopoulou , V., Heron, D., Nicol, A., 2025. AFG - Active Faults Greece: a comprehensive geomorphologybased 1:25,000 fault database. Scientific Data, 12:1853 | https://doi.org/10.1038/s41597-025-06283-z California Geological Survey (CGS). (2018). Earthquake Fault Zones: A Guide for Government Agencies, Property Owners / Developers, and Geoscience Practitioners for Assessing Fault Rupture Hazards in California (Special Publication 42, Revised 2018). Sacramento, CA: California Department of Conservation. Caputo R., Chatzipetros A., Pavlides S. and Sboras S. (2012): The Greek Database of Seismogenic Sources (GreDaSS): state-of-the-art for northern Greece. Ann. Geophys., 55(5), 859-894. Cornell, C. A. (1968). Engineering seismic risk analysis. Bulletin of the Seismological Society of America, 58(5), 1583–1606. https://doi.org/10.1785/BSSA0580051583 D’Agostino, N., Métois, M., Koci, R., Duni, L., Kuka, N., Ganas, A., et al. (2020). Active crustal deformation and rotations in the southwestern Balkans from continuous GPS measurements. Earth and Planetary Science Letters, 539, 116246. https://doi.org/10.1016/j.epsl.2020.116246 Duman TY, Çan T, Emre Ö, Kadirioğlu FT, Başarır Bastürk N, Kılıc¸ T, Arslan S, Özalp S, Kartal RF, Kalafat D, et al. 2018. Seismotectonic database of Turkey. Bull Earthquake Eng. 16(8):3277–3316. doi: 10.1007/s10518-016-9965-9. Elma, İ., Özçelik, B., Karabacak, V., Özkaymak, Ç. ve Sümer, Ö. (2024). First paleoseismological evidence from the İnönü–Oklubalı segment of the Eskişehir Fault. Turkish Journal of Earthquake Research , 6 , 349–368. Elma, İ., Safarov, M. A., Karabacak, V., … Özkaymak, Ç. (2025). Paleosismolojik Veriler Işığında Eskişehir Fayı’nın Batı Segmentinin Deprem Davranışı / Earthquake Behaviour of the Western Segment of the Eskişehir Fault in Light of Paleoseismological Data. Türkiye Jeoloji Bülteni, 68(3), 333-356. https://doi.org/10.25288/tjb.1647559 Emre Ö, Duman TY, Özalp S, Şaroğlu F, Olgun S¸, Elmacı H, Çan T. 2018. Active fault database of Turkey. Bull Earthquake Eng. 16(8):3229–3275. doi: 10.1007/s10518-016-0041-2. Emre, Ö., Duman, T. Y., Özalp, S., Elmacı, H., Olgun, Ş., & Şaroğlu, F. (2013). Active Fault Map of Türkiye with Explanatory Text (Scale 1:1,250,000). Special Publication Series-30. Ankara: General Directorate of Mineral Research and Exploration (MTA). Emre, Ö., Duman, T. Y., Özalp, S., Şaroğlu, F., Olgun, Ş., Elmacı, H., & Çan, T. (2018). Active fault database of Türkiye. Bulletin of Earthquake Engineering, 16(8), 3229–3275. https://doi.org/10.1007/s10518-016-0041-2 Field, E. H., Arrowsmith, R. J., Biasi, G. P., Bird, P., Dawson, T. E., Felzer, K. R.,... & Zeng, Y. (2014). Uniform California Earthquake Rupture Forecast, Version 3 (UCERF3)—The Time-Independent Model. Bulletin of the Seismological Society of America, 104(3), 1122–1180. https://doi.org/10.1785/0120130164 Fisher, W.D. (1958). On grouping for maximum homogeneity. Journal of the American Statistical Association, 53(284), 789-798. Ganas, A., Oikonomou, I. A., & Tsimi, C. (2013). NOAfaults: a digital database for active faults in Greece. Bulletin of the Geological Society of Greece, 47(2), 518–530. https://doi.org/10.12681/bgsg.11079 ITHACA Working Group. (2019). ITHACA (ITaly HAzard from CApable faulting), A database of active capable faults of the Italian territory (Version Dec 2019). ISPRA Geological Survey of Italy. http://sgi.isprambiente.it/ithacaweb/ Jenks, G.F. (1967). The data model concept in statistical mapping. International Yearbook of Cartography, 7, 186-190. Jenks, G.F., & Caspall, F.C. (1971). Error on choroplethic maps: definition, measurement, reduction. Annals of the Association of American Geographers, 61(2), 217-244. Karabacak, V., & Sançar, T. (2025). Bursa İli Yerleşim Alanlarının Deprem Tehlikesi: Ulubat, Bursa ve İnegöl Diri Faylarının Paleosismolojik Özellikleri / Earthquake Hazard in Urban Areas of Bursa Province: Paleosismological Properties of the Ulubat, Bursa and İnegöl Active Faults. Türkiye Jeoloji Bülteni, 68(2), 151-176. https://doi.org/10.25288/tjb.1642150 Karabacak, V., Özkaymak, Ç., Sümer, Ö. (2024). Determination of Paleoseismological Characteristics of the Eskişehir Fault and the Dodurga Fault, TÜBİTAK Project No. 123G010, Progress Report. Karabacak, V., Sançar, T., Sağlam Selçuk, A., & Büyükdeniz, Y. (2021). Paleoseismicity of the Ulubat Fault: Inferences on Seismic Behaviour of the Southern Branch of the North Anatolian Fault Zone, South Marmara. Turkish Journal of Earthquake Research, 3(1), 1–19. https://doi.org/10.46464/tdad.909358 Kerr, J., Nathan, S., Van Dissen, R., Webb, P., Brunsdon, D., & King, A. (2003). Planning for Development of Land on or Close to Active Faults: A guideline to assist resource management planners in New Zealand. Ministry for the Environment (MfE). Machette, M. N. (2000). Active, capable, and potentially active faults—a paleoseismic perspective. Journal of Geodynamics, 29(3–5), 387–392. https://doi.org/10.1016/S0264-3707(99)00060-5 Machette, M. N., Haller, K. M., Dart, R. L., & Rhea, S. B. (2004). Quaternary fault and fold database of the United States. U.S. Geological Survey Open-File Report 03-417. https://doi.org/10.3133/ofr03417 McGuire, R. K. (2004). Seismic Hazard and Risk Analysis (Monograph MNO-10). Oakland, CA: Earthquake Engineering Research Institute (EERI). Özkaymak Ç (2015). Tectonic analysis of the Honaz Fault (western Anatolia) using geomorphic indices and the regional implications. Geodinamica Acta 27 (2-3): 110-129. P/SISMO-TÜRK. (2023–2026). Determination of the paleoseismological characteristics of Türkiye’s active faults (TÜBİTAK Project No. 107G001), Research Project, TÜBİTAK Pavlides S, Caputo R, Sboras S, Chatzipetros A, Papathanasiou G, Valkaniotis S. The Greek catalogue of active faults and database of seismogenic sources. Bull Geol Soc. Greece 2010;43(1):486–94. Pavlides, S., Drakatos, G., Zouros, N. (2024). Active Tectonics and Seismicity in Greece. In: Darques, R., Sidiropoulos, G., Kalabokidis, K. (eds) The Geography of Greece. World Regional Geography Book Series. Springer, Cham. https://doi.org/10.1007/978-3-031-29819-6_25. Research Group for Active Faults of Japan. (1991). Active Faults in Japan: Sheet Maps and Inventories (Revised ed.). Tokyo: University of Tokyo Press. Sozbilir, H., Ozkaymak, C., Uzel, B., & Sumer, O. (2018). Criteria for surface rupture microzonation of active faults for earthquake hazards in urban areas. In Handbook of Research on Trends and Digital Advances in Engineering Geology (pp. 187–230). Hershey, PA: IGI Global. https://doi.org/10.4018/978-1-5225-2709-1.ch006 Styron, R., & Pagani, M. (2020). The GEM Global Active Faults Database. Earthquake Spectra, 36(1_suppl), 160–180. https://doi.org/10.1177/8755293020944182 Şafak Yaşar L., Tiryakioğlu İ., Aktuğ B., Erdoğan H. and Özkaymak Ç., (2025) Determination of Anatolian Plate’s tectonic block boundaries with clustering analysis using GNSS sites velocities, Geomatics, Natural Hazards and Risk, 16:1, 2446588, DOI: 10.1080/19475705.2024.2446588 Taymaz, T., Yılmaz, Y., Dilek, Y., 2007. The geodynamics of the Aegean and Anatolia: introduction. Geological Society, London, Special Publications, 291, 1–16. Viltres, R., Jónsson, S., Alothman, A. O., Liu, S., Leroy, S., Masson, F., Doubre, C., Reilinger R., (2022). Present-day motion of the Arabian plate. Tectonics, 41, e2021TC007013. https://doi.org/10.1029/2021TC007013 Zelenin, E., Bachmanov, D., Garipova, S., Trifonov, V., & Kozhurin, A. (2022). The Active Faults of Eurasia Database (AFEAD): the ontology and design behind the continental-scale dataset. Earth System Science Data, 14(10), 4489–4503. https://doi.org/10.5194/essd-14-4489-2022 Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8404630","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":574733749,"identity":"2879b4a4-02be-40af-b0c9-4585bff5ef71","order_by":0,"name":"Volkan Karabacak","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABAElEQVRIiWNgGAWjYBACCSBmBjOYmY///lMB4jI3EKmFnS1BgucMiMtIrBZ+HgMJ3jYQk4AWyfazDz8X1NjJSzbzGBhIzquN5m8HavlRsQ2nFmmedGPpGceSDWczsxUkGG47njvjMGMDY8+Z2zi1yDGkAbWxHWCcx8y84UDitmO5DUAtzIxteLTwP2P+zfPvgP08ZgbDhoNzjuXOJ6RFWiKNTZq37UDibGYWY8bGhprcDYS0SM54xmY9sy85eWYzWxozw7EDuRuBWg7i84vE+TTm2wXf7GxnnD98jJmhpi533vnDBx/8qMCtBR0cBpMHiFYPBHWkKB4Fo2AUjIIRAgDsyVXyRBQQcwAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0003-2581-7984","institution":"Eskişehir Osmangazi University","correspondingAuthor":true,"prefix":"","firstName":"Volkan","middleName":"","lastName":"Karabacak","suffix":""},{"id":574733750,"identity":"362d552e-735b-4d06-aa43-b812044ee760","order_by":1,"name":"Çağlar Özkaymak","email":"","orcid":"","institution":"Afyon Kocatepe University: Afyon Kocatepe Universitesi","correspondingAuthor":false,"prefix":"","firstName":"Çağlar","middleName":"","lastName":"Özkaymak","suffix":""},{"id":574733751,"identity":"8f8a1b8b-de4e-4cb4-a682-a087d94ba989","order_by":2,"name":"Ökmen Sümer","email":"","orcid":"","institution":"Dokuz Eylul University: Dokuz Eylul Universitesi","correspondingAuthor":false,"prefix":"","firstName":"Ökmen","middleName":"","lastName":"Sümer","suffix":""}],"badges":[],"createdAt":"2025-12-19 12:20:16","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8404630/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8404630/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":100600555,"identity":"75500a2d-5cb1-4e3c-a1d6-1b841332d270","added_by":"auto","created_at":"2026-01-19 14:48:42","extension":"jpeg","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":982949,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8404630/v1/959e4921438ecb7de5bed4dc.jpeg"},{"id":100599953,"identity":"27651550-1f5d-4543-83e8-5d54216e231d","added_by":"auto","created_at":"2026-01-19 14:45:48","extension":"jpeg","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":613973,"visible":true,"origin":"","legend":"","description":"","filename":"Figure2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8404630/v1/1f8a44aa394ee7e2e09da082.jpeg"},{"id":100599949,"identity":"65d312e3-e8fa-4e12-abb6-0e3bdf506e89","added_by":"auto","created_at":"2026-01-19 14:45:47","extension":"xml","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8376,"visible":true,"origin":"","legend":"","description":"","filename":"beeeBEEED2501221.xml","url":"https://assets-eu.researchsquare.com/files/rs-8404630/v1/56494566c3a840039a2a561a.xml"},{"id":100600231,"identity":"9500df07-56f8-47ca-8566-c372b42cf26c","added_by":"auto","created_at":"2026-01-19 14:47:01","extension":"xml","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1100,"visible":true,"origin":"","legend":"","description":"","filename":"BEEED250122122467.go.xml","url":"https://assets-eu.researchsquare.com/files/rs-8404630/v1/2d1ce1fee10f249dabf7e0ce.xml"},{"id":100599952,"identity":"1c38dee5-b445-432f-afa1-af8eca43b669","added_by":"auto","created_at":"2026-01-19 14:45:48","extension":"xml","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":969,"visible":true,"origin":"","legend":"","description":"","filename":"BEEED2501221Import.xml","url":"https://assets-eu.researchsquare.com/files/rs-8404630/v1/106488de6f4b0734fee63b01.xml"},{"id":107707724,"identity":"e62b0ed4-6b33-4377-a9ef-e38c7fe7efaa","added_by":"auto","created_at":"2026-04-24 09:21:00","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":311332,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8404630/v1/7a2bc30a-d2aa-47de-b458-307b0155c33d.pdf"}],"financialInterests":"","formattedTitle":"Transitioning from Deterministic Active Fault Maps to Probabilistic Models for Seismic Hazard Assessment: A Global Review and the Bursa-Eskişehir Fault System (Türkiye) Case Study","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eThe success of earthquake hazard and risk mitigation strategies depends largely on the reliability of the answer to the question: \"Where, when, and how will the earth's crust rupture?\" The geological equivalent of this answer lies in the accurate definition of the spatial distribution, geometry, segmentation architecture, and earthquake-generating characteristics of active faults.\u003c/p\u003e\n\u003cp\u003eToday, Probabilistic Seismic Hazard Analyses (PSHA) and engineering design spectra directly utilize source zoning and parameters such as fault length, slip type, slip rate, recurrence interval (RI), and maximum credible earthquake magnitude (M\u003csub\u003emax\u003c/sub\u003e) as inputs (Cornell, 1968; McGuire, 2004). While these parameters are essential for modern hazard models, traditional deterministic maps often provide only spatial location ('where'), lacking the critical temporal and parametric data ('when' and 'how often') required to reduce epistemic uncertainties in PSHA. Therefore, bridging the gap between geological fieldwork and engineering demand requires a shift from static inventories to dynamic source models. Similarly, site selection decisions for land-use planning and critical infrastructure projects must consider not only ground acceleration hazards but also \"near-fault\" hazards such as surface faulting and permanent displacement. In this context, active fault maps serve as the fundamental \"interface\" where pure geological information is translated into engineering parameters and planning decisions.\u003c/p\u003e\n\u003cp\u003eHowever, this critical interface is not bound by a single international terminology or a standardized time window. The approaches of countries are shaped more by regulatory needs and risk perception than by scientific consensus. While some countries accept Holocene activity (approximately the last 11,000 years) as a \"sufficiently young and well-defined\" threshold for regulatory zoning purposes, others adopt a wider geological time window on the scale of the Late Pleistocene. On the other hand, international databases focus on the earthquake-generating potential of faults through a PSHA-compatible \"seismogenic source\" approach rather than age thresholds. For instance, on a global scale, the GEM Global Active Faults Database (GAF-DB) aims to gather active fault geometries and attributes under a single roof by combining regional compilations (Styron and Pagani, 2020), whereas the European-scale EFSM20 and ESHM20 frameworks structure active faults to serve directly as geological inputs for probabilistic hazard models (Basili et al., 2022).\u003c/p\u003e\n\u003cp\u003eTürkiye, located in a high deformation rate sector of the Eastern Mediterranean–Anatolian transition zone within the Alpine–Himalayan belt, faces high seismic risk potential alongside the geological diversity brought by this tectonic context (Fig. 1a). Destructive earthquakes causing tens of thousands of casualties in the last century, and most recently the February 6, 2023 Kahramanmaraş earthquakes, have once again revealed the catastrophic scale this natural phenomenon can reach.\u003c/p\u003e\n\u003cp\u003eThe first and most fundamental stage of reducing earthquake damage is the mapping and characterization of the active faults generating the earthquakes in the field. However, today this need extends far beyond merely drawing fault traces on a map. The standard now required is the reporting of parametric information, such as segment structure, rupture zone geometry, short and long-term slip rates, recurrence intervals, and displacement amounts per event together with data quality and uncertainties.\u003c/p\u003e\n\u003cp\u003eThe first holistic national-scale mapping initiative in Türkiye began with the 1:1,000,000 scale Active Fault Map of Türkiye published by the General Directorate of Mineral Research and Exploration (MTA) in 1992, which was updated at different scales between 2004 and 2011 in light of advances in earthquake geology and increasing data accumulation (Emre et al., 2013). Following this process, a \"Türkiye Paleoseismology Research\" vision was developed to systematically investigate paleoseismological behavior; particularly in the post-2023 period, parametric data production on priority faults and segments has gained momentum within the framework of a broad-participation platform.\u003c/p\u003e\n\u003cp\u003eThis paper addresses this outlined framework through two complementary datasets: (1) A comprehensive inventory analysis compiling and classifying global active fault mapping protocols (Supplementary File 1), and (2) An application example based on segment-based paleoseismological event sequences, recurrence interval (RI), and \"cycle ratio\" indicators on the\u0026nbsp;Bursa–Eskişehir Fault System (BEFS)\u0026nbsp;(Fig. 1b) (Supplementary File 2).\u003c/p\u003e\n\u003cp\u003eThe primary objective of this study is not to criticize Türkiye's existing mapping standard, but to propose a gradual and practicable scientific framework that can elevate the rapidly growing paleoseismological and parametric database to a \"probabilistic and parametric\" mapping level compatible with modern international approaches. Importantly, the mapping framework proposed in this paper is not intended to replace the existing deterministic Active Fault Map of Türkiye produced by MTA, which remains the authoritative national baseline for fault‐trace location and age‐based activity classes (Emre et al., 2013, 2018). Instead, consistent with international practice where multiple active‐fault products coexist for different end‐users and decision contexts (Supplementary File 1), we propose an additional, risk‐oriented cartographic layer that goes beyond trace depiction by translating fault behavior and its uncertainties (e.g., segmentation confidence, recurrence statistics, and cycle‐stage indicators) into a PSHA‐compatible prioritization interface. Accordingly, the proposed classification should be read as a complementary product designed to sit alongside trace maps and regulatory zoning tools, not as an alternative to them.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eFIGURE 1\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e"},{"header":"2. Conceptual Framework: Time Windows and Classification Types","content":"\u003cp\u003eIn the discipline of active fault mapping, two fundamental methodological parameters shape the philosophy of map creation and its ultimate intended use: (i) The time window, which determines the geological time interval of activity upon which a fault is deemed \"active,\" and (ii) The classification type, which defines the systematic logic by which this activity information is coded and reported.\u003c/p\u003e\u003ch4\u003e2.1. Time Window Approaches\u003c/h4\u003e\u003cp\u003eTime windows most frequently referenced in literature and practice can be categorized according to the regulatory or scientific purposes of the mapping:\u003c/p\u003e\u003cul class=\"decimal_type\"\u003e\n \u003cli\u003eHolocene-Focused (10–12 ka): This is the narrowest time interval, preferred especially in applications yielding direct legal/regulatory consequences such as construction restrictions, the establishment of surface faulting setback zones, and zoning planning. For example, the Alquist–Priolo Earthquake Fault Zoning Act in California (conservation.ca.gov) and similar national regulations (Sözbilir et al., 2018) base their approach on this window to minimize construction risk on active fault traces.\u003c/li\u003e\n \u003cli\u003eLate Pleistocene / Late Quaternary (100–130 ka): Covering not only the last 10,000 years but also the last glacial and interglacial periods, this approach offers a broader \"recent geological time\" perspective to avoid overlooking faults with long seismic recurrence periods.\u003c/li\u003e\n \u003cli\u003eQuaternary (2.58 Ma): This is the most widely accepted upper limit in national-scale fault inventories and global databases. However, within this broad time interval, the level of evidence and the precision of separation into sub-epochs (Pleistocene/Holocene) can vary between regions.\u003c/li\u003e\n \u003cli\u003ePliocene–Quaternary (~5 Ma) and Broad-Scale Neotectonic Windows: These broad time intervals, generally used in active tectonic atlases and regional geodynamic compilations, aim to understand the evolution of the current deformation regime.\u003c/li\u003e\n\u003c/ul\u003e\u003cp\u003eThis diversity stems from the \"purpose-oriented\" nature of mapping rather than scientific inconsistency. Indeed, there is no obligation for zoning maps constituting the basis of spatial planning and PSHA-based source models to use the same time thresholds.\u003c/p\u003e\u003ch4\u003e2.2. Classification Types: Age, Evidence, and Source Modelling\u003c/h4\u003e\u003cp\u003eThe global inventory analysis within Supplementary File 1 indicates that four main methodological approaches are adopted in the classification of active faults:\u003c/p\u003e\u003cul\u003e\n \u003cli\u003eAge-Based (AGE): The most basic approach where classification seeks to answer only whether the \"last movement occurred within a specific time window.\"\u003c/li\u003e\n \u003cli\u003eAge + Evidence/Confidence (AGE+EVID / AGE+CONF): A more qualified classification type where, in addition to chronological information, the type of data documenting the activity (paleoseismology, geomorphology, historical/instrumental seismicity, etc.) and the level of confidence in the data are coded.\u003c/li\u003e\n \u003cli\u003eCapable Fault Approach: Used especially in critical infrastructure projects such as nuclear power plants and dams, this approach focuses not only on the fault's past activity but also on its \"potential to generate surface displacement/rupture in the near future\" (Machette, 2000).\u003c/li\u003e\n \u003cli\u003eSeismogenic Source (SOURCE-MODEL): The most advanced level where the fault is modeled as a \"seismogenic source\" that can be used directly as input in PSHA. Here, in addition to fault trace geometry, numerical parameters such as slip rate, moment balance, maximum magnitude (M\u003csub\u003emax\u003c/sub\u003e), and segment connectivity are integrated into the system. The European-scale EFSM20 and ESHM20 projects are current and successful examples of this approach (Basili et al., 2022).\u003c/li\u003e\n\u003c/ul\u003e"},{"header":"3. Dataset and Methodological Framework","content":"\u003cp\u003eThis research relies on two fundamental datasets that complement each other: a comparative analysis of global active fault protocols and a regional application example (i.e. BEFS).\u003c/p\u003e\u003cp\u003eThe global dataset (Supplementary File 1), forming the first pillar of the study, was created by compiling 35 national/institutional active fault maps and databases from at least 30 different countries and regions. Each product in this inventory is presented in tables coded based on time window (\"time_bin\") and classification type (\"class_bin\") to make them suitable for analysis.\u003c/p\u003e\u003cp\u003eThe \"time_bin\" coding used in the analysis of the global inventory allows for the comparison of different national approaches. Examination of the inventory distribution reveals a predominance of Quaternary-focused products (QUAT, approx. 15 maps), followed by Late Quaternary/Late Pleistocene scale (LQ, LP100/LP125) studies, PSHA-focused source models (SRC), and Holocene-focused regulatory approaches (HOLO). This distribution suggests that there is no \"single correct time window\" in active fault mapping; instead, the combination of \"purpose, scale, and risk tolerance\" is the determining factor. Similarly, \"class_bin\" codes, which examine whether maps carry only fault traces or attribute and modeling capacity, reveal that basic inventory types like AGE+MAP (approx. 10 maps) are common, but SOURCE-MODEL (3) structures that can be directly integrated into seismic hazard analyses (PSHA) remain more limited. Furthermore, the scarcity of harmonized multi-source inventories (HARMONIZED) indicates that the need for standardization in international data integration persists.\u003c/p\u003e\u003cp\u003eThe BEFS example (Supplementary File 2), constituting the regional application part of the study, includes segment-based paleoseismological event sequences for the Ulubat, Bursa, İnegöl, and Eskişehir faults, recurrence interval (RI) estimates obtained via Bayesian age modeling, and the time elapsed since the last event/RI ratio (cycle ratio). Event chronologies and age constraints used in this scope were compiled from Karabacak et al. (2021) for the Ulubat Fault, Karabacak \u0026amp; Sancar (2025) for the Bursa and İnegöl faults, and Elma et al. (2024, 2025) and Karabacak et al. (2025) for the Eskişehir Fault.\u003c/p\u003e\u003cp\u003eIn the BEFS analyses, two main indicators were utilized considering the available datasets: 1) Recurrence Interval (RI), representing the median and uncertainty range of the segment's paleoseismic recurrence period, and 2) Cycle Ratio\u0026nbsp;(CR), representing the ratio of the time elapsed since the last surface rupture to the RI. In this methodology, a cycle ratio of ~1 and above is interpreted as a relative prioritization indicator signifying that the segment has entered the late stage of its cycle; the 0.5–1 range indicates increasing stress accumulation, while a value \u0026lt;0.5 signifies the segment is in a relatively early stage (Supplementary File 2). This approach ensures that seismic hazard is addressed with intervals containing uncertainty rather than deterministic singular values. However, the method has a critical limitation; the cycle ratio serves to create a comparative risk window for decision-makers rather than definitively \"dating\" the expected earthquake, offering a statistical risk projection rather than a claim that an earthquake will occur soon.\u003c/p\u003e\u003cp\u003eTo translate the segment-based 'Cycle Ratio' (CR) indicators that defined as the ratio of elapsed time to the median recurrence interval into a spatially coherent and risk-oriented cartographic model, a statistical classification method was adopted. Instead of utilizing arbitrary linear intervals, the distribution of CR values derived from the Türkiye-wide paleoseismological inventory (Supplementary File 3, Table 1) was subjected to the Fisher–Jenks natural breaks optimization algorithm (Fisher, 1958; Jenks, 1967). The Fisher–Jenks optimization was specifically selected over arbitrary linear intervals to minimize intra-class variance and maximize the deviation between class means. This ensures that the proposed risk categories (C1–C5) reflect the inherent statistical clustering of the stress accumulation stages derived from the national paleoseismological inventory, providing an objective basis for prioritization.\u003c/p\u003e\u003cp\u003eThe optimization analysis indicated that a four-class solution for the pre-saturation interval (0 \u0026lt; CR ≤ 1.0) yielded the highest 'Goodness of Variance Fit' (GVF ≈ 0.97), providing the most representative delineation of the seismic cycle from early post-seismic relaxation to the late accumulation phase. Consequently, the mapping legend was structured into six distinct categories: an 'undefined' class (C0) for data-poor segments, four sequential classes for the standard cycle (C1–C4), and a specific 'limit exceeded' or 'overdue' category (C5) for segments where the elapsed time exceeds the characteristic recurrence interval (CR \u0026gt; 1.0) (Supplementary File 3, Table 2). This classification architecture allows the active fault map to evolve from a static trace inventory into a graded priority interface suitable for decision-making processes.\u003c/p\u003e"},{"header":"4. Global Active Fault Mapping Protocols: Trends and Distinctive Examples","content":"\u003cp\u003eWhen the general landscape of active fault mapping studies created in different countries on a national and institutional basis (Supplementary File 1) is examined, it is observed that studies diversify along three main axes (purpose, scale, and modelling level) rather than following a standard template. While the ultimate goal of mapping (regulatory zoning, national hazard modelling, or academic/geodynamic inventory production) determines the quality of the product, the data resolution difference between 1:1,000,000 scale continental/country compilations and 1:5,000 local scale zoning maps reveals the heterogeneous structure of the inventory. Similarly, basic maps offering only fault trace and kinematic class show distinct differences in terms of modelling maturity compared to systems hosting parameters such as slip rate, recurrence interval (RI), and maximum magnitude (M\u003csub\u003emax\u003c/sub\u003e) in their databases.\u003c/p\u003e\u003cp\u003eIn the last decade, global and continental databases integrating national compilations have accelerated the search for standardization and a common language for this diversity. In this context, the GEM Global Active Faults Database (GAF-DB) gathers worldwide active faults under a single roof with attributes such as geometry, kinematics, slip rate, and references (Styron and Pagani, 2020), while the Active Faults of Eurasia Database (AFEAD) offers a highly detailed geodatabase by reporting tens of thousands of fault objects across Eurasia in a consistent structure (Basili et al., 2022). The fundamental contribution of these databases to the literature is moving the \"which fault is active?\" debate from solely a time window constraint to an axis of evidence type, parameter set, and uncertainty trio. However, it is a clear reality that these global initiatives require a comprehensive \"metadata and harmonization\" process for national data to become interoperable, rather than directly replacing national regulatory standards.\u003c/p\u003e\u003cp\u003eFrom a regulatory framework perspective, the Alquist–Priolo Earthquake Fault Zoning Act in California constitutes a distinctive example with its approach to reducing surface faulting hazard. This protocol mandates the creation of \"Earthquake Fault Zones\" around active fault traces and requires construction decisions in these zones to be supported by field investigations (conservation.ca.gov). Associating \"activity\" mostly with Holocene activity and well-defined fault traces, this approach aims to minimize geological uncertainty via \"on-site evidence\" before reflecting it directly onto construction risk.\u003c/p\u003e\u003cp\u003eOn the other hand, approaches like EFSM20 and ESHM20 in Europe treat active faults directly as an input for probabilistic seismic hazard analyses (PSHA) (hazard.efehr.org). Here, faults are modelled not just as geographic traces but as seismogenic sources for which earthquake rate estimations can be made. Therefore, the primary need in this perspective is not only the location of the fault but also geometric and kinematic parameter sets allowing for moment balance and earthquake rate calculations. This transforms the fault mapping process from a deterministic \"map\" production into a probabilistic \"model\" construction.\u003c/p\u003e\u003cp\u003eOne of the best institutional examples of this transformation is the Quaternary Fault and Fold Database (QFault) operated by the USGS in the USA, which links the \"Quaternary fault\" definition to a systematic meta-data scheme, presenting fault geometry and segmentation alongside age/time of last movement, slip rate intervals, and literature references in a singular record logic (Machette et al., 2004). Thus, the inventory ceases to be merely a cartographic product and transforms into an updatable and traceable infrastructure providing data for hazard modeling. Similarly, Alquist–Priolo zones in California are an advanced regulatory tool connecting surface rupture hazard directly to fieldwork and engineering applications via a \"legal obligation + on-site verification\" combination, rather than being just a national database (California Geological Survey).\u003c/p\u003e\u003cp\u003eThe Active Fault Database of Japan, maintained by AIST/GSJ, incorporates \"hazard degree\" indicators such as behavioral segmentation of faults and slip rate classes into the classification within a framework based on Late Quaternary activity (120–130 ka) (AIST/GSJ; Research Group for Active Faults of Japan, 1991). New Zealand's NZAFD similarly defines activity within an approximately 125 ka window but exhibits close integration with guidelines for transferring surface rupture hazard to land-use decisions via recurrence interval and displacement parameters (GNS Science; Kerr et al., 2003).\u003c/p\u003e\u003cp\u003eOn the Eastern Mediterranean scale, platforms like NOAFaults and GreDaSS in Greece produce evidence/confidence degree classes by associating active fault geometry with the \"seismogenic source\" definition (Ganas et al., 2013; GreDaSS Consortium, 2014); in Italy, ISPRA's ITHACA approach demonstrates that a targeted data-criterion architecture can be applied for engineering site selection, specifically focusing on \"capable fault\" and surface faulting hazard (ISPRA). These examples emphasize that the two-layered classification centered on \u003cem\u003etime_bin\u003c/em\u003e and \u003cem\u003eclass_bin\u003c/em\u003e proposed in Supplementary File 1 finds practical correspondence in mature systems through meta-data fields and uncertainty indicators.\u003c/p\u003e"},{"header":"5. Active Fault Mapping in Türkiye: The MTA Approach and Transforming Needs","content":"\u003cp\u003eThe first standardized source on a national scale for understanding the active fault architecture in Türkiye is the 1992 Active Fault Map of Türkiye by MTA (General Directorate of Mineral Research and Exploration). This map served as the primary reference for earthquake hazard and planning studies for many years. With methodological progress in earthquake geology and data increase, active fault maps were renewed at different scales between 2004–2011; the explanatory text and classification logic of the map were published by Emre et al. (2013, 2018).\u003c/p\u003e\u003cp\u003eThe current requirement is for these maps to become \"parametric.\" In other words, the database must carry not only where the fault is, but also how it behaves and with what uncertainties its earthquake generation potential is defined. In this direction, the primary goal of paleoseismology-based studies conducted under MTA coordination is: (i) detailing structural/geometric features, (ii) determining historical/prehistoric earthquakes resulting in surface rupture, (iii) constraining recurrence intervals and slip rates, and (iv) transferring results to a national database.\u003c/p\u003e\u003cp\u003eThe platform approach, which accelerated after February 6, 2023, aims to close the paleoseismological data gap in numerous fault segments in Türkiye. For instance, an action was taken by bringing together Türkiye's state institutions to reduce earthquake risk. Implemented under the leadership of expert scientists, the \"Platform\" (P/SISMO-TURK Project, project no: 107G001)) targets the investigation of Türkiye's Active Faults in all aspects by nearly 200 scientists.\u003c/p\u003e\u003cp\u003eThis rapid data production brings two critical questions to the agenda simultaneously:\u003c/p\u003e\u003col\u003e\n \u003cli\u003eHow will this new data be integrated into deterministic \"trace maps\"?\u003c/li\u003e\n \u003cli\u003eHow will a \"probabilistic\" active fault classification/mapping be designed for PSHA and risk mitigation decisions?\u003c/li\u003e\n\u003c/ol\u003e\u003cp\u003eAt this point, the main lesson emerging from the global inventory (Supplementary File 1) is this: Türkiye's need is not to select a single time window, but to design a multi-layered classification and harmonization parameters capable of serving both regulatory/planning and PSHA-focused usage through a single database. This multi-layered view also implies that the proposed probabilistic/parametric classes are designed to complement, not displace, deterministic trace mapping. In practical terms, the deterministic map continues to serve as the primary reference for fault-trace geometry and regulatory/planning workflows, whereas the additional layer developed here targets PSHA-driven hazard and risk applications (e.g., insurance, infrastructure prioritization, and scenario-based mitigation) by explicitly incorporating parameter ranges and uncertainty. This two-track architecture reflects the “portfolio” logic observed in mature mapping systems, in which trace inventories, capable-fault/zoning products, and source-model datasets are maintained together for different purposes (Supplementary File 1).\u003c/p\u003e"},{"header":"6. Case Study: Bursa–Eskişehir Fault System","content":"\u003ch4\u003e6.1. Nature of the Study Area and Selection Criteria\u003c/h4\u003e\u003cp\u003eBEFS, extending from the south of the Marmara Region to the interior of Central Anatolia (Figs. 1b and 2a), qualifies as an ideal key application area for this study with its multi-segmented structure and tectonic diversity. While the Ulubat, Bursa, and İnegöl faults forming the western wing of the system are in direct interaction with dense industrial and residential areas (İ.e. Bursa, İnegöl and Eskişehir Cities), the Eskişehir Fault on the easternmost represents a critical tectonic corridor in the interior of Anatolia (Fig. 2a). This diversity allows for the combined evaluation of both high-risk urban centers and segments showing differences in data density. Furthermore, the selection of the BEFS is driven by data reliability and internal consistency. The co-authors of this study were directly involved in the geological mapping and paleoseismological trenching campaigns across these segments (e.g., Karabacak et al., 2021; Karabacak \u0026amp; Sançar, 2025; Elma et al., 2025). This first-hand engagement allows for expert control over the quality of the input data and the management of epistemic uncertainties regarding event horizons and chronological constraints. Additionally, as a priority focus area within the ongoing national 'P/SISMO-TURK' project, the BEFS ensures a continuous flow of high-resolution parametric data, making it an ideal dynamic laboratory for testing the proposed probabilistic framework.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e\u003cem\u003eFIGURE 2\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\u003ch4\u003e6.2. Segment-Based Recurrence Interval and Cycle Ratio Analyses\u003c/h4\u003e\u003cp\u003eSegment-based recurrence interval (RI), dates of last events, elapsed time, and cycle ratios (Elapsed Time/RI) compiled from the dataset detailed in Supplementary File 2 are presented in Table 1. These data constitute the basis for understanding the current positions of fault segments within the seismic cycle. The findings in Table 1 reveal two critical patterns regarding system characteristics. First, a distinct heterogeneity is observed within the system; cycle ratios within the same fault system range widely from very low values like 0.08 to critical levels like 1.07, indicating the completion of the seismic cycle. Second, as seen in the Ulubat East segment example, the generation of different scenarios (A and B) demonstrates the necessity of transparently managing uncertainty instead of a deterministic single-scenario approach.\u003c/p\u003e\u003cp\u003eTable 1. RI and Cycle Ratio in the BEFS segments.\u003c/p\u003e\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"595\"\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eFault\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSegment\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 113px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRI (median; 5%–95%) (yr)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eLast Event\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eElapsed Time (median; 5%–95%) \u0026nbsp;(yr)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCycle Ratio\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd rowspan=\"3\" valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eUlubat\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eCentral segment\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 113px;\"\u003e\n \u003cp\u003e5502 (1804–7531)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003eAD 1139\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e886 (875–903)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.16\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eEast segment – Scenario A\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 113px;\"\u003e\n \u003cp\u003e4181 (3975–4387)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003eAD 175\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e1850 (1649–2052)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.44\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eEast segment – Scenario B\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 113px;\"\u003e\n \u003cp\u003e4019 (3649–4392)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003eAD 13\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e2012 (1640–2381)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.50\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eBursa\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eBursa segment\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 113px;\"\u003e\n \u003cp\u003e2117 (579–5647)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003eAD 1855\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e170 (168–172)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.08\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eİnegöl\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eWest segment\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 113px;\"\u003e\n \u003cp\u003e2395 (948–3452)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003eBC 541\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e2565 (2450–2724)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e1.07\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd rowspan=\"4\" valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eEskişehir\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eSegment 1\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 113px;\"\u003e\n \u003cp\u003e3278 (3121–3432)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003eAD 542 ± 144\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e~1480\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.45\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eSegment 2\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 113px;\"\u003e\n \u003cp\u003e3221 (2892–3552)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003eBC 800 ± 100\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e~2825\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.88\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eSegment 3\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 113px;\"\u003e\n \u003cp\u003e3046 (2148–3940)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003eBC 886 ± 350\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e~2900\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.95\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eSegment 4\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 113px;\"\u003e\n \u003cp\u003e3174 (2826–3523)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003eAD 750 ± 50\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e~1275\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e0.40\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003ch4\u003e6.3. Fault System Scale Relative Stress Accumulation Classification\u003c/h4\u003e\u003cp\u003eThe cumulative view of segment-based data at the fault system scale and the relative stress accumulation classification are summarized in Table 2. Analysis results point to an accumulation pattern that can be characterized as a late-stage in terms of the seismic cycle, particularly for the İnegöl West segment and certain sections of the Eskişehir Fault (especially Seg. 2 and Seg. 3). However, it must be strongly emphasized that this classification is not an earthquake timing prediction, but a data-based risk prioritization metric.\u003c/p\u003e\u003cp\u003eTable 2. Summary of relative stress accumulation at the fault scale along the BEFS.\u003c/p\u003e\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"624\"\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eFault\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNo. of Segments\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMedian Cycle Ratio\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMax. Cycle Ratio\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRelative Classification (Stress Accumulation)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eUlubat\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e0.44\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e0.50\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003eModerate\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eBursa\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e0.08\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e0.08\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003eLow\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eİnegöl\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e1.07\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e1.07\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003eHigh (Limit Exceeded)\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eEskişehir\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e0.66\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003e0.95\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 125px;\"\u003e\n \u003cp\u003eHigh\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003ch4\u003e6.4. Stress Transfer and Multi-Segment Rupture Dynamics\u003c/h4\u003e\u003cp\u003eIn multi-fault systems, seismic hazard cannot be reduced to the independent behavior of a single segment. As conceptually outlined by Field et al. (2014), the rupture probability of neighboring segments can change dynamically with stress transfer and inter-segment connectivity. Although paleoseismological data reduce uncertainty regarding the \"renewal timing\" of segments, geodesy (GNSS), seismotectonic modeling, and historical earthquake catalogs must be evaluated in an integrated manner to fully quantify stress transfer at the system scale. In this context, the BEFS example serves as a conducive pilot region for making the transition to \"probabilistic active fault mapping\" and the need for standardization visible in Türkiye, thanks to its data heterogeneity and high interaction with residential areas.\u003c/p\u003e"},{"header":"7. Discussion: Transition from Deterministic Mapping to Parametric Active Fault Modeling","content":"\u003ch4\u003e7.1. Why \"Probabilistic Active Fault Map\"?\u003c/h4\u003e\u003cp\u003eThe existing Active Fault Map of Türkiye has filled a significant gap in literature and practice by providing a standard answer to the question of \"spatial distribution of source faults\" on a country scale. However, today, the requirements of Probabilistic Seismic Hazard Analysis (PSHA) and risk mitigation strategies have moved beyond merely knowing the geographic location of the fault. Modern seismotectonic approaches mandate the reporting of parameters such as fault geometry, segmentation structure, slip rate, recurrence interval (RI), and maximum magnitude (M\u003csub\u003emax\u003c/sub\u003e) together with their epistemic and aleatory uncertainties. This necessity means the active fault map must evolve from a static \"inventory product\" into a continuously updatable, dynamic \"model input\" for hazard analyses. Therefore, the transition from deterministic mapping to parametric and probabilistic mapping is not a preference but a scientific and practical necessity.\u003c/p\u003e\u003ch4\u003e7.2. From Multi-National Comparison to National Standard: Harmonization Parameters\u003c/h4\u003e\u003cp\u003eThe global inventory analysis presented in this study (Supplementary File 1) proposes a parameter set to ensure the interoperability of protocols in different countries rather than directly equating them. For Türkiye, a practicable harmonization core capable of meeting both regulatory zoning (surface faulting hazard) and PSHA-focused source modeling needs within a single national database can be gathered under the following headings:\u003c/p\u003e\u003cul\u003e\n \u003cli\u003eTime Window Label (time_bin): The chronological interval defining the fault's activity (e.g., QUAT, LQ, HOLO).\u003c/li\u003e\n \u003cli\u003eClassification Label (class_bin): Coding determining the quality of the product (e.g., AGE+MAP, AGE+EVID, SOURCE-MODEL).\u003c/li\u003e\n \u003cli\u003eEvidence Type and Quality (EVID/CONF): The source of data documenting activity (paleoseismology, geomorphology, historical/instrumental record, geodesy) and the confidence level attributed to this data.\u003c/li\u003e\n \u003cli\u003eGeometry and Segmentation Confidence: Precision of fault trace location, uncertainties regarding segment boundaries, and presence of distributed deformation zones instead of singular fault traces.\u003c/li\u003e\n \u003cli\u003eParametric Fields: Short/long-term slip rates, recurrence interval (RI), single event displacement, fault width/depth, $M_{max}$ values, and statistical uncertainty ranges for all these parameters.\u003c/li\u003e\n \u003cli\u003eModeling Compatibility: Direct usability of data in PSHA processes (moment balance, earthquake occurrence rate models, logic-tree branches).\u003c/li\u003e\n\u003c/ul\u003e\u003cp\u003eAdopting this parameter set as a common language will enable the multi-purpose use of the national database.\u003c/p\u003e\u003ch4\u003e7.3. Critical Requirement Pointed Out by the BEFS Example: Uncertainty Management\u003c/h4\u003e\u003cp\u003eSpecific to the BEFS, the scenarios produced particularly for the Ulubat East segment represent a practical application of the probabilistic and parametric mindset in Türkiye (Supplementary File 2). The fundamental philosophy here is not to ignore uncertainty, but to model it and make it visible. This approach creates value in two main dimensions: Scientifically, it prevents data deficiency or interpretation differences from being masked by a \"singular and definite value.\" Practically, it offers decision-makers (municipalities, infrastructure operators, insurance sector, and risk governance units) a more realistic risk projection containing \"ranges and probabilities\" instead of a deterministic judgment.\u003c/p\u003e\u003ch4\u003e7.4. Limitations and Manageable Uncertainty\u003c/h4\u003e\u003cp\u003eThis proposed transformation process requires the clear acceptance and management of certain limitations:\u003c/p\u003e\u003cul\u003e\n \u003cli\u003eData Constraints: The number of paleoseismological trenches and dating data is still limited in many fault segments; this may cause uncertainty ranges to widen.\u003c/li\u003e\n \u003cli\u003eGeomorphological Ambiguity: In areas where segment boundaries cannot be clearly traced geomorphologically or in distributed deformation zones, the concept of a mappable \"single trace\" may remain insufficient.\u003c/li\u003e\n \u003cli\u003eInterpretation Risk: Derived indicators like Cycle Ratio are not definite earthquake timing predictions, but merely tools for relative prioritization among segments. There is a risk that this data may be misinterpreted as \"an earthquake will definitely happen soon.\"\u003c/li\u003e\n \u003cli\u003eUse-case boundary: The cycle ratio based classes and the resulting Türkiye-scale map are intended as a national-scale, comparative prioritization tool for hazard and risk governance. They are not a substitute for site-specific surface-fault-rupture zoning, engineering microzonation, or legal setbacks, which require higher-resolution mapping, targeted paleoseismological/geomorphic investigations, and on-site verification in the project area. Misuse outside this intended scale and purpose may lead to erroneous “safe/unsafe” interpretations.\u003c/li\u003e\n\u003c/ul\u003e\u003cp\u003eConsequently, clearly coding and reporting these limitations is not a weakness of probabilistic mapping, but strictly its strongest aspect. This is because uncertainty is thereby removed from being an unknown and transformed into a manageable risk parameter.\u003c/p\u003e\u003ch4\u003e7.5. Transfer of Cycle Ratio Based Classification to Mapping at Türkiye Scale\u003c/h4\u003e\u003cp\u003eThe rapid increase in paleoseismological trench data and recurrence interval (RI) solutions based on these data in Türkiye makes it possible to go beyond producing active fault maps solely with \"trace/presence-absence\" logic. However, on a national scale, data scope and quality are heterogeneous among segments; while RI and/or time of last event are well-constrained in some segments, uncertainties are high or basic parameters are undefined in others. Therefore, a new mapping example requires a simple yet traceable classification scheme that (i) offers a comparable prioritization metric to decision-makers, and (ii) clearly shows data deficiency and uncertainty.\u003c/p\u003e\u003cp\u003eIn the approach proposed in this study, the seismic cycle position for each fault/segment is expressed by the normalized \"cycle ratio\" (CR = elapsed time / RImedian). In the map legend, CR values in the 0-1 range are shown with sequential color tones, establishing a readable visual hierarchy from the early stage of the cycle (lighter tone) to the late/critical stage (darker tone). In determining class intervals, the Fisher-Jenks \"natural breaks\" approach, which best represents the natural clustering of the Türkiye-wide CR distribution ((Supplementary File 1), Table 1), was used, and solutions minimizing intra-class variance were preferred in selecting the number of classes (Fisher, 1958; Jenks, 1967; Jenks and Caspall, 1971). Within this framework, the 0-1 range was divided into four main classes; the CR\u0026gt;1 condition was highlighted separately with hatching/thick contours instead of color to distinguish the segments with CR \u0026gt; 1 (cycle-limit exceedance) (Table\u0026nbsp;3).\u003c/p\u003e\u003cp\u003eThe critical component of the proposed classification is keeping undefined/uncertain records as a separate category on the map (C0). Thus, the map makes visible not only high or low-priority segments but also gaps in terms of national scale parametric data production; this also functions as a prioritization tool for future paleoseismological and geodetic studies. In Figure 2b, the Cycle Ratio (CR) based national fault segment classification (Table 3), determined from paleoseismological data obtained from active faults across the entire Turkish mainland, has been adapted for the BEFS and its immediate surroundings, which were selected as the key study area, in order to make first \u003cem\u003eParametric Active Fault Map\u003c/em\u003e representation concrete for practical use. Finally, it should be emphasized that CR classes are designed not to produce an earthquake-timing prediction, but to offer a relative cycle-prioritization scale among segments. Therefore, in transferring classes to planning/PSHA applications, it is essential to evaluate them together with additional parameters such as evidence type, data confidence level, slip rate, segment connectivity, and multi-segment rupture probability.\u0026nbsp;\u003c/p\u003e\u003cp\u003eTable 3. Cycle Ratio (CR) based national fault segment classification and map representation (Adapted from Supplementary File 3, Table 2).\u003c/p\u003e\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"628\"\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eClass\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCycle Ratio (CR)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 241px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSeismic Cycle Position\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMap Representation\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eC0\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003eUncertain / Undefined\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 241px;\"\u003e\n \u003cp\u003eInsufficient data (No RI or last event age, or unreliable)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003eGray tone; dotted/thin line\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eC1\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003e0.00-0.21\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 241px;\"\u003e\n \u003cp\u003eVery early stage of cycle (post-event)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003eLight green tone\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eC2\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003e\u0026gt;0.21-0.40\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 241px;\"\u003e\n \u003cp\u003eEarly stage (stress accumulation begins)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003eYellow tone\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eC3\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003e\u0026gt;0.40-0.75\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 241px;\"\u003e\n \u003cp\u003eMiddle stage (accumulation evident)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003eOrange tone\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eC4\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003e\u0026gt;0.75-1.00\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 241px;\"\u003e\n \u003cp\u003eLate stage / critical period (approach to RI)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003eRed tone\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eC5\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003e\u0026gt;1.00\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 241px;\"\u003e\n \u003cp\u003eCycle limit exceedance (CR \u0026gt; 1; RI threshold exceeded)\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003eRed tone with surrounded hatching and/or thick contour\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e"},{"header":"8. Conclusions","content":"\u003cp\u003eThis paper reviews the practical diversity of “active fault” definitions and mapping standards worldwide, and uses that comparison to frame a set of minimum, transparent metadata fields for national fault inventories in Türkiye. The inventory of 35 protocols highlights that time window, activity criteria, and mapping purpose are commonly coupled, and that a single, universal “active” definition is rarely achievable across regulatory zoning, engineering applications, and PSHA-oriented source modeling.\u003c/p\u003e\u003cp\u003eWe then demonstrate the proposed workflow on the BEFS by compiling published paleoseismological constraints at the segment scale and deriving a Cycle Ratio (CR = elapsed time / RI_median) as a relative indicator of seismic-cycle position. The BEFS example shows strong along-strike variability, with low CR values for some segments (e.g., Bursa segment, ~0.08) and values approaching or exceeding unity in others (e.g., İnegöl West segment, ~1.07).\u003c/p\u003e\u003cp\u003eCR should not be interpreted as a deterministic earthquake-timing statement. Its meaning depends on the quality of event chronologies, the representativeness of estimated recurrence intervals, and the adopted renewal-model assumptions. Accordingly, CR is best used as an auditable prioritization layer to identify where additional data, alternative segment scenarios, or sensitivity testing would most improve downstream hazard models.\u003c/p\u003e\u003cp\u003eFor practical transfer to seismogenic source databases, we recommend reporting (at minimum) the adopted time window and classification logic, evidence type and confidence, geometry/segmentation uncertainty, and parametric fields relevant to time-dependent modeling (e.g., RI distributions and, where available, slip-rate constraints). In this sense, the existing deterministic Active Fault Map of Türkiye remains a critical geometric baseline, while an uncertainty-coded parametric layer such as the one illustrated here can be maintained as a complementary product within the same evolving database.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eEthics approval and consent to participate\u003c/p\u003e\n\u003cp\u003eNot applicable. This study does not involve human participants, human data, or animals.\u003c/p\u003e\n\u003cp\u003eConsent for publication\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003eAvailability of data and materials\u003c/p\u003e\n\u003cp\u003eAll data and materials supporting the findings of this study are included within the manuscript and its Supplementary Files. Additional information can be provided by the corresponding author upon reasonable request.\u003c/p\u003e\n\u003cp\u003eCompeting interests\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003eFunding\u003c/p\u003e\n\u003cp\u003eThis research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.\u003c/p\u003e\n\u003cp\u003eAuthors\u0026rsquo; contributions\u003c/p\u003e\n\u003cp\u003eVK conceived the study, developed the methodological framework, compiled and analyzed the datasets, and wrote the first draft of the manuscript. Co-authors \u0026Ccedil;\u0026Ouml; and \u0026Ouml;S compiled and analyzed the datasets, contributed to data interpretation, figure preparation, critical revision of the manuscript, and approved the final version.\u003c/p\u003e\n\u003cp\u003eAcknowledgements\u003c/p\u003e\n\u003cp\u003eThe authors thank colleagues and institutions that provided access to published datasets and cartographic resources used in this study. We also thank the anonymous reviewers in advance for their constructive comments that will help improve the manuscript.\u003c/p\u003e\n\u003cp\u003eSupplementary information\u003c/p\u003e\n\u003cp\u003eSupplementary information accompanies this paper and is provided as Supplementary Files 1\u0026ndash;3.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e● \u0026nbsp; \u0026nbsp;Supplementary File 1: Global \u0026quot;Active Fault Mapping\u0026quot; protocols\u003c/p\u003e\n\u003cp\u003e● \u0026nbsp; \u0026nbsp;Supplementary File 2: Bursa\u0026ndash;Eskişehir Fault System (BEFS) paleoseismological analysis\u003c/p\u003e\n\u003cp\u003e● \u0026nbsp; \u0026nbsp;Supplementary File 3: Paleoseismological based earthquake cycle analysis of T\u0026uuml;rkiye\u0026apos;s Active Faults\u003c/p\u003e\n\u003cp\u003eUse of AI and AI-assisted technologies\u003c/p\u003e\n\u003cp\u003eThe authors declare that no generative AI or AI-assisted technologies were used to generate scientific content, interpret results, or draw conclusions. Digital tools were used for language editing and formatting only, and the authors take full responsibility for the integrity and originality of the work.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eBasili, R., Danciu, L., Beauval, C., Sesetyan, K., Vilanova, S. P., Adamia, S.,... \u0026amp; Zupančič, P. (2022). European Fault-Source Model 2020 (EFSM20): online data on fault geometry and activity parameters. Istituto Nazionale di Geofisica e Vulcanologia (INGV). https://doi.org/10.13127/efsm20\u003c/li\u003e\n \u003cli\u003eBegg, J.G., Mouslopoulou , V., Heron, D., Nicol, A., 2025. AFG - Active Faults Greece: a comprehensive geomorphologybased 1:25,000 fault database. Scientific Data, 12:1853 | https://doi.org/10.1038/s41597-025-06283-z\u003c/li\u003e\n \u003cli\u003eCalifornia Geological Survey (CGS). (2018). Earthquake Fault Zones: A Guide for Government Agencies, Property Owners / Developers, and Geoscience Practitioners for Assessing Fault Rupture Hazards in California (Special Publication 42, Revised 2018). Sacramento, CA: California Department of Conservation.\u003c/li\u003e\n \u003cli\u003eCaputo R., Chatzipetros A., Pavlides S. and Sboras S. (2012): The Greek Database of Seismogenic Sources (GreDaSS): state-of-the-art for northern Greece. Ann. Geophys., 55(5), 859-894.\u003c/li\u003e\n \u003cli\u003eCornell, C. A. (1968). Engineering seismic risk analysis. Bulletin of the Seismological Society of America, 58(5), 1583\u0026ndash;1606. https://doi.org/10.1785/BSSA0580051583\u003c/li\u003e\n \u003cli\u003eD\u0026rsquo;Agostino, N., M\u0026eacute;tois, M., Koci, R., Duni, L., Kuka, N., Ganas, A., et al. (2020). Active crustal deformation and rotations in the southwestern Balkans from continuous GPS measurements. Earth and Planetary Science Letters, 539, 116246. https://doi.org/10.1016/j.epsl.2020.116246\u003c/li\u003e\n \u003cli\u003eDuman TY, \u0026Ccedil;an T, Emre \u0026Ouml;, Kadirioğlu FT, Başarır Bast\u0026uuml;rk N, Kılıc\u0026cedil; T, Arslan S, \u0026Ouml;zalp S, Kartal RF, Kalafat D, et al. 2018. Seismotectonic database of Turkey. Bull Earthquake Eng. 16(8):3277\u0026ndash;3316. doi: 10.1007/s10518-016-9965-9.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eElma, İ., \u0026Ouml;z\u0026ccedil;elik, B., Karabacak, V., \u0026Ouml;zkaymak, \u0026Ccedil;. ve S\u0026uuml;mer, \u0026Ouml;. (2024). First paleoseismological evidence from the İn\u0026ouml;n\u0026uuml;\u0026ndash;Oklubalı segment of the Eskişehir Fault. \u003cem\u003eTurkish Journal of Earthquake Research\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e, 349\u0026ndash;368.\u003c/li\u003e\n \u003cli\u003eElma, İ., Safarov, M. A., Karabacak, V., \u0026hellip; \u0026Ouml;zkaymak, \u0026Ccedil;. (2025). Paleosismolojik Veriler Işığında Eskişehir Fayı\u0026rsquo;nın Batı Segmentinin Deprem Davranışı / Earthquake Behaviour of the Western Segment of the Eskişehir Fault in Light of Paleoseismological Data. T\u0026uuml;rkiye Jeoloji B\u0026uuml;lteni, 68(3), 333-356. https://doi.org/10.25288/tjb.1647559\u003c/li\u003e\n \u003cli\u003eEmre \u0026Ouml;, Duman TY, \u0026Ouml;zalp S, Şaroğlu F, Olgun S\u0026cedil;, Elmacı H, \u0026Ccedil;an T. 2018. Active fault database of Turkey. Bull Earthquake Eng. 16(8):3229\u0026ndash;3275. doi: 10.1007/s10518-016-0041-2.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eEmre, \u0026Ouml;., Duman, T. Y., \u0026Ouml;zalp, S., Elmacı, H., Olgun, Ş., \u0026amp; Şaroğlu, F. (2013). Active Fault Map of T\u0026uuml;rkiye with Explanatory Text (Scale 1:1,250,000). Special Publication Series-30. Ankara: General Directorate of Mineral Research and Exploration (MTA).\u003c/li\u003e\n \u003cli\u003eEmre, \u0026Ouml;., Duman, T. Y., \u0026Ouml;zalp, S., Şaroğlu, F., Olgun, Ş., Elmacı, H., \u0026amp; \u0026Ccedil;an, T. (2018). Active fault database of T\u0026uuml;rkiye. Bulletin of Earthquake Engineering, 16(8), 3229\u0026ndash;3275. https://doi.org/10.1007/s10518-016-0041-2\u003c/li\u003e\n \u003cli\u003eField, E. H., Arrowsmith, R. J., Biasi, G. P., Bird, P., Dawson, T. E., Felzer, K. R.,... \u0026amp; Zeng, Y. (2014). Uniform California Earthquake Rupture Forecast, Version 3 (UCERF3)\u0026mdash;The Time-Independent Model. Bulletin of the Seismological Society of America, 104(3), 1122\u0026ndash;1180. https://doi.org/10.1785/0120130164\u003c/li\u003e\n \u003cli\u003eFisher, W.D. (1958). On grouping for maximum homogeneity. Journal of the American Statistical Association, 53(284), 789-798.\u003c/li\u003e\n \u003cli\u003eGanas, A., Oikonomou, I. A., \u0026amp; Tsimi, C. (2013). NOAfaults: a digital database for active faults in Greece. Bulletin of the Geological Society of Greece, 47(2), 518\u0026ndash;530. https://doi.org/10.12681/bgsg.11079\u003c/li\u003e\n \u003cli\u003eITHACA Working Group. (2019). ITHACA (ITaly HAzard from CApable faulting), A database of active capable faults of the Italian territory (Version Dec 2019). ISPRA Geological Survey of Italy. http://sgi.isprambiente.it/ithacaweb/\u003c/li\u003e\n \u003cli\u003eJenks, G.F. (1967). The data model concept in statistical mapping. International Yearbook of Cartography, 7, 186-190.\u003c/li\u003e\n \u003cli\u003eJenks, G.F., \u0026amp; Caspall, F.C. (1971). Error on choroplethic maps: definition, measurement, reduction. Annals of the Association of American Geographers, 61(2), 217-244.\u003c/li\u003e\n \u003cli\u003eKarabacak, V., \u0026amp; San\u0026ccedil;ar, T. (2025). Bursa İli Yerleşim Alanlarının Deprem Tehlikesi: Ulubat, Bursa ve İneg\u0026ouml;l Diri Faylarının Paleosismolojik \u0026Ouml;zellikleri / Earthquake Hazard in Urban Areas of Bursa Province: Paleosismological Properties of the Ulubat, Bursa and İneg\u0026ouml;l Active Faults. T\u0026uuml;rkiye Jeoloji B\u0026uuml;lteni, 68(2), 151-176.\u0026nbsp;https://doi.org/10.25288/tjb.1642150\u003c/li\u003e\n \u003cli\u003eKarabacak, V., \u0026Ouml;zkaymak, \u0026Ccedil;., S\u0026uuml;mer, \u0026Ouml;. (2024). Determination of Paleoseismological Characteristics of the Eskişehir Fault and the Dodurga Fault, T\u0026Uuml;BİTAK Project No. 123G010, Progress Report.\u003c/li\u003e\n \u003cli\u003eKarabacak, V., San\u0026ccedil;ar, T., Sağlam Sel\u0026ccedil;uk, A., \u0026amp; B\u0026uuml;y\u0026uuml;kdeniz, Y. (2021). Paleoseismicity of the Ulubat Fault: Inferences on Seismic Behaviour of the Southern Branch of the North Anatolian Fault Zone, South Marmara. Turkish Journal of Earthquake Research, 3(1), 1\u0026ndash;19. https://doi.org/10.46464/tdad.909358\u003c/li\u003e\n \u003cli\u003eKerr, J., Nathan, S., Van Dissen, R., Webb, P., Brunsdon, D., \u0026amp; King, A. (2003). Planning for Development of Land on or Close to Active Faults: A guideline to assist resource management planners in New Zealand. Ministry for the Environment (MfE).\u003c/li\u003e\n \u003cli\u003eMachette, M. N. (2000). Active, capable, and potentially active faults\u0026mdash;a paleoseismic perspective. Journal of Geodynamics, 29(3\u0026ndash;5), 387\u0026ndash;392. https://doi.org/10.1016/S0264-3707(99)00060-5\u003c/li\u003e\n \u003cli\u003eMachette, M. N., Haller, K. M., Dart, R. L., \u0026amp; Rhea, S. B. (2004). Quaternary fault and fold database of the United States. U.S. Geological Survey Open-File Report 03-417. https://doi.org/10.3133/ofr03417\u003c/li\u003e\n \u003cli\u003eMcGuire, R. K. (2004). Seismic Hazard and Risk Analysis (Monograph MNO-10). Oakland, CA: Earthquake Engineering Research Institute (EERI).\u003c/li\u003e\n \u003cli\u003e\u0026Ouml;zkaymak \u0026Ccedil; (2015). Tectonic analysis of the Honaz Fault (western Anatolia) using geomorphic indices and the regional implications. Geodinamica Acta 27 (2-3): 110-129.\u003c/li\u003e\n \u003cli\u003eP/SISMO-T\u0026Uuml;RK. (2023\u0026ndash;2026). \u003cem\u003eDetermination of the paleoseismological characteristics of T\u0026uuml;rkiye\u0026rsquo;s active faults\u003c/em\u003e (T\u0026Uuml;BİTAK Project No. 107G001), Research Project, T\u0026Uuml;BİTAK\u003c/li\u003e\n \u003cli\u003ePavlides S, Caputo R, Sboras S, Chatzipetros A, Papathanasiou G, Valkaniotis S. The Greek catalogue of active faults and database of seismogenic sources. Bull Geol Soc. Greece 2010;43(1):486\u0026ndash;94.\u003c/li\u003e\n \u003cli\u003ePavlides, S., Drakatos, G., Zouros, N. (2024). Active Tectonics and Seismicity in Greece. In: Darques, R., Sidiropoulos, G., Kalabokidis, K. (eds) The Geography of Greece. World Regional Geography Book Series. Springer, Cham. https://doi.org/10.1007/978-3-031-29819-6_25.\u003c/li\u003e\n \u003cli\u003eResearch Group for Active Faults of Japan. (1991). Active Faults in Japan: Sheet Maps and Inventories (Revised ed.). Tokyo: University of Tokyo Press.\u003c/li\u003e\n \u003cli\u003eSozbilir, H., Ozkaymak, C., Uzel, B., \u0026amp; Sumer, O. (2018). Criteria for surface rupture microzonation of active faults for earthquake hazards in urban areas. In Handbook of Research on Trends and Digital Advances in Engineering Geology (pp. 187\u0026ndash;230). Hershey, PA: IGI Global. https://doi.org/10.4018/978-1-5225-2709-1.ch006\u003c/li\u003e\n \u003cli\u003eStyron, R., \u0026amp; Pagani, M. (2020). The GEM Global Active Faults Database. Earthquake Spectra, 36(1_suppl), 160\u0026ndash;180. https://doi.org/10.1177/8755293020944182\u003c/li\u003e\n \u003cli\u003eŞafak Yaşar L., Tiryakioğlu İ., Aktuğ B., Erdoğan H. and \u0026Ouml;zkaymak \u0026Ccedil;., (2025) Determination of Anatolian Plate\u0026rsquo;s tectonic block boundaries with clustering analysis using GNSS sites velocities, Geomatics, Natural Hazards and Risk, 16:1, 2446588, DOI: 10.1080/19475705.2024.2446588\u003c/li\u003e\n \u003cli\u003eTaymaz, T., Yılmaz, Y., Dilek, Y., 2007. The geodynamics of the Aegean and Anatolia: introduction. Geological Society, London, Special Publications, 291, 1\u0026ndash;16.\u003c/li\u003e\n \u003cli\u003eViltres, R., J\u0026oacute;nsson, S., Alothman, A. O., Liu, S., Leroy, S., Masson, F., Doubre, C., Reilinger R., (2022). Present-day motion of the Arabian plate. Tectonics, 41, e2021TC007013. https://doi.org/10.1029/2021TC007013\u003c/li\u003e\n \u003cli\u003eZelenin, E., Bachmanov, D., Garipova, S., Trifonov, V., \u0026amp; Kozhurin, A. (2022). The Active Faults of Eurasia Database (AFEAD): the ontology and design behind the continental-scale dataset. Earth System Science Data, 14(10), 4489\u0026ndash;4503. https://doi.org/10.5194/essd-14-4489-2022\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Active fault mapping, Probabilistic Seismic Hazard Assessment (PSHA), Seismic cycle, Bursa–Eskişehir Fault System, Time-dependent hazard, Türkiye","lastPublishedDoi":"10.21203/rs.3.rs-8404630/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8404630/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Active fault maps provide the geometric backbone for probabilistic seismic hazard assessment (PSHA) and surface-fault-rupture considerations, yet what qualifies as an “active fault” varies markedly among countries and use-cases. Here we compile and compare 35 widely used mapping protocols and databases, evaluate current practice in Türkiye, and propose a harmonization framework that (i) makes the selected time window and classification logic explicit and (ii) encodes evidence quality and geometric/segmentation uncertainty in a traceable way. As a case study we apply the workflow to the multi-segment Bursa–Eskişehir Fault System (BEFS) and calculate a segment-based Cycle Ratio, CR = elapsed time since the most recent event / median recurrence interval, as a relative indicator of seismic-cycle position. In BEFS, CR spans from very low values (e.g., the Bursa segment, ~0.08) to values near or above unity (e.g., the İnegöl West segment, ~1.07), highlighting strong along-strike variability in inferred stress accumulation. We emphasize that CR is not an earthquake-timing forecast; rather, it provides an auditable prioritization layer that can be combined with slip rate, segmentation uncertainty, and event-history constraints when building time-dependent source models. The proposed workflow offers a reproducible path for incorporating Türkiye’s rapidly growing paleoseismological record into next-generation seismogenic source databases.","manuscriptTitle":"Transitioning from Deterministic Active Fault Maps to Probabilistic Models for Seismic Hazard Assessment: A Global Review and the Bursa-Eskişehir Fault System (Türkiye) Case Study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-19 13:44:52","doi":"10.21203/rs.3.rs-8404630/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"b3a5d868-4918-4ebc-aaff-0794d3389f0d","owner":[],"postedDate":"January 19th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-04-24T00:03:54+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-19 13:44:52","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8404630","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8404630","identity":"rs-8404630","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}