One-Second-Lead Earthquake Warning and Impact Assessment at Campi Flegrei

One-Second-Lead Earthquake Warning and Impact Assessment at Campi Flegrei | 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 Article One-Second-Lead Earthquake Warning and Impact Assessment at Campi Flegrei Valeria Longobardi, Simona Colombelli, Aldo Zollo This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7619275/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 08 Apr, 2026 Read the published version in Scientific Reports → Version 1 posted 4 You are reading this latest preprint version Abstract Rapid dissemination of earthquake information during seismic crises is crucial for mitigating the impact of earthquakes, particularly in densely populated volcanic areas, such as the Campi Flegrei caldera (Southern Italy). This region is affected by recurrent bradyseism and shallow seismicity, with recent events up to Md 4.6, very well perceived by the local inhabitants and increasing concern among the local community. In such small-extent areas, conventional Earthquake Early Warning (EEW) systems face physical limitations, due to the short duration and relatively small magnitude of earthquakes. Here we propose and calibrate a hybrid, impact-based, on-site EEW system capable of providing rapid estimates of the earthquake size and expected impact (PGV, PGA), within one second after the P-wave detection. In addition, the proposed system extends the classical concept of an on-site EEW system, introducing the idea of an “area of competence” around each seismic station that can benefit from the local warning. The proposed methodology is easily transferable to other volcanic or seismically active regions and might represent a step toward low-latency, impact-based earthquake alert systems designed to enhance community resilience. Earth and environmental sciences/Natural hazards Earth and environmental sciences/Solid earth sciences Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction In volcanic and seismic regions, during the occurrence of seismic crises, the timely dissemination of earthquake information can significantly mitigate the earthquake impact by improving emergency safety actions [ 1 ]. A compelling case study is the current situation at the Campi Flegrei caldera in Southern Italy, where the ongoing volcanic unrest is raising increasing concern among local population and authorities in the area. During the past 75 years, the caldera has experienced repeated episodes of ground deformation, known as bradyseism (slow ground up- and downlift), documented since the 1950s [ 2 ]. The two most rapid phases of uplift occurred in 1970–72 and 1980–84 [ 3 ]. Since 2005, a new monotonic uplift episode has been underway, with a notable acceleration in seismicity observed from 2014 onwards [ 4 , 5 ]. During the last two years, the land instability with high uplift rates reaching 2–3 cm per month, is being accompanied by an intense and frequent seismic activity with the occurrence of several Md 4 + earthquakes originated at shallow depths (3 km) and causing strong ground shaking (up to several tenths of g) in restricted, few km radius, areas around the epicenter [ 7 ]. The metropolitan area surrounding the Campi Flegrei caldera is densely populated, with several hundred thousand residents [ 6 ]. The recent intense seismic activity—particularly the sudden uplift in the Pozzuoli area—has caused damage to infrastructure and buildings, increasing the urgency for enhanced seismic risk mitigation [ 7 ]. Automatic and near-real-time monitoring systems are currently in operation [ 1 , 8 ], but the time required to detect, process, and disseminate earthquake information (up to several tens of minutes) —especially for the lower magnitude but well perceived earthquakes—remains a critical limitation in real-time emergency management [ 9 , 10 ]. Since these earthquakes are generally small-to-moderate in magnitude, the area affected by strong shaking is also limited to a few kilometers around the epicenter and comparable with the “blind zone” of the warning systems, which is defined as the area where the strong shaking occurs before the alert can be issued. Because of this physical constraint the applicability of earthquake early warning (EEW) systems in such a small extent volcanic region, as the Campi Flegrei area, is limited. However, real-time data processing and methodologies can still be applied and locally calibrated to provide a P-wave-based rapid response within a short time window after the earthquake occurrence. This approach may be particularly useful for the largest magnitude, strong shaking causative events (M 3–4), which are clearly felt by residents [ 7 ] over a wider area (10–15 km). The rapid response might encompass different aspects of risk mitigation, including the evaluation of shake maps, the quick assessment of damage to guide emergency actions as well as the rapid dissemination of earthquake information to reduce the spread of misinformation and increase the awareness among the population. Within this frame, we tailored a standard EEW approach to the peculiar seismicity of the Campi Flegrei area, to provide early P-wave based shake and potential damage maps, that can become crucial tools to support Civil Protection Authorities (CPAs) and first responders during emergencies. EEW approaches are broadly classified into source-based and impact-based methods [ 1 ]. Source-based EEW systems rapidly estimate earthquake source parameters (e.g. location and magnitude) to predict ground shaking at distant sites before the arrival of damaging waves. While effective in tectonic regions with larger, longer-duration earthquakes [ 11 – 14 ], this approach faces severe limitations in areas like Campi Flegrei where seismic events are typically low-to-moderate in magnitude, very shallow, and short in duration [ 7 ]. Impact-based methods bypass instead the need for full source characterization [ 15 , 16 ]. They rely directly on ground motion observations at individual stations to estimate the potential earthquake impact in real time. Here, we propose a hybrid on-site (EEW) methodology adapting an existing method [ 17 ], to the Campi Flegrei case-study. The early P-displacement signal is used to predict peak ground motion at the site, while at the same time an approximate estimation of the magnitude is obtained to characterize the event size. We calibrate and test a rapid response system using a large dataset of earthquakes recorded during the recent unrest at Campi Flegrei. Additionally, show the example of its performance considering two significant earthquake scenarios: the Md 4.4 event of May 20, 2024, and the Md 4.6 event of March 13, 2025. Results To calibrate and test the system we used 700 events with duration magnitude Md between 1 and 4, which occurred during 2016-2024 at Campi Flegrei caldera (panel A of Figure 1). The preliminary dataset is split into two subsets: the training dataset is used for calibration of empirical relationships, and the test dataset is used for performance evaluation (see panel B of Figure 1). We focused here on the estimation, at a single station, with the onsite method of two quantities: the peak ground motion (PGV and PGA) and the earthquake magnitude. As for the PGV prediction, in panel A of Figure 1 we show the scaling of observed PGV with the P-peak displacement (Pd) measured within a one-second P-wave time window for the training dataset. To account for the unbalanced number of earthquakes in the different magnitude classes (see panel B of Figure 1), we calibrate the empirical scaling laws of PGV vs. Pd using 2-D binned data (represented as cyan points in panel A of Figure 2). The solid and dashed lines represent the resulting PGV–Pd scaling law and its standard error bounds, respectively. The results of the calibration and the test for the PGA are shown in the supplemental material (figure S4 and S5). As for magnitude, in panel B of Figure 2, we present the single-station magnitude estimate for a P-wave time window of 1 second, derived from the seismic moment (M₀). The seismic moment is estimated as the area under a triangular source time function, with Pd as the peak and tauc (period parameter) as the duration [19] (see Data & Methods section). We refer to this magnitude estimate as Mew. The Mew method relies on two assumptions: (1) the far-field displacement can be modeled as a triangle with amplitude Pd and duration τc, and (2) the hypocentral distance used in the estimate (see Data & Methods section) is an average value obtained from the distribution of all possible hypocentral distances for the given station (see Figure S1) determined from the recorded dataset, rather than its exact value. This latter assumption appears valid for a small area such as the Campi Flegrei caldera, where variations in hypocentral distance are limited due to the distribution of stations and event locations (see Figure 1). Nevertheless, the error on Mew determinations is related to the real distance and the empirical distribution of earthquake distances from the concerned station as inferred from the current seismicity. To allow a direct comparison, we computed the moment magnitude Mw, using the scaling relationship of Iervolino et al. (2024) [7] that links catalog duration magnitude Md with the moment magnitude Mw (shown on the x-axis of panel B in Figure 2). The prediction error, defined as the difference between Mew and moment magnitude (Mw), is shown in the inset histogram of panel B of Figure 2. The average error is centered around zero, with a standard deviation of 0.36 magnitude units. We also applied an alternative method to estimate the single-station magnitude by using the empirical relationship that relates log Px (x being acceleration, velocity and displacement) to the observed magnitude and observed log-hypocentral distance (see supplementary material text S1 and figure S1 for more details). We define this estimate as MPx. We compare MPx with Mw in figure S2. The prediction error – the difference between MPx and Mw - shows a higher value of its standard deviation (0.41 magnitude units) with respect to what it is obtained for Mew. The Earthquake Early Warning and Rapid Response (EEWRR) system implemented at Campi Flegrei is designed to provide real-time estimates of earthquake impact within the first second after the arrival of the P-wave. As soon as the vertical component of ground motion is detected, key parameters (Pd, Pa, Pv, and τc) are extracted from the initial portion of the waveform. Among these, Pd and τc are used to derive a first-order estimate of the seismic moment and corresponding moment magnitude. In parallel, empirical ground-motion prediction equations (GMPEs) are applied to the observed P-wave peak amplitudes in order to forecast the expected peak ground velocity (PGV) and peak ground acceleration (PGA). Assuming a GMPE of the form log(PGV) = A + B·Mw + C log(R) (A,B,C are estimated from data), where M is the estimated magnitude and R the hypocentral distance, the system determines a circular area around the recording station within which PGV is expected to vary by ±X%, with X defined by the user. The final outputs consist of: (i) the predicted PGV and PGA at the site, which can be converted into macroseismic intensity using the empirical relations of Faenza and Michelini [20]; (ii) an approximate Mw estimate; (iii) the extent of the area surrounding the site where shaking levels are expected to fall within the chosen uncertainty range. In Figure 4 and 5, we show the test on two earthquake scenarios, the Md 4.4 occurred on May 20, 2024 and the Md 4.6 occurred on March 13, 2025 respectively. These two events – among the largest registered in the area- are not included in the dataset used for calibration and testing of the system. For both scenarios, we computed the predicted impact area around each station within which the predicted PGV (PGV pred ) may vary between +-50% of its value. Since each single station can provide the expected PGV (PGV pred ) and the magnitude (Mew) estimate, the couples (PGV pred - 50%PGV pred , Mew) and (PGV pred + 50%PGV pred , Mew) can be used to evaluate at first order the area within which this variation of the PGV will be experienced. In this way, each station defines its own “radius of competence,” that is, the area around the station in which we expect stable values of predicted PGV. The size of this radius depends on the GMPE adopted for the region. Here, we use the GMPE from Scala et al. (2025) [21], truncated to retain only the first term of hypocentral distance (see Data & Methods for details). Discussion In this study, we demonstrate the feasibility of implementing an on-site early warning system, tailored to the specific characteristics of the Campi Flegrei region. The Campi Flegrei caldera is a very densely populated area, with 500 thousand inhabitants distributed over a limited territory (200km 2 ) [ 6 ]. The presence of critical infrastructure elements (such as schools, hospitals, emergency response centers, and transportation hubs) makes the proposed real-time earthquake alert system relevant for risk mitigation. The system can predict earthquake impact at the site in terms of PGV (or PGA) and delineate a circular area within which this value is expected to vary by a user-defined percentage. This piece of information is compiled into an alert message that can be immediately broadcast to the exposed population, enabling the prompt activation of emergency safety measures—a form of local, real-time shakemap. The entire process is completed within roughly one second from the P-wave detection at the site (see Figs. 4 and 5 ). The proposed methodology represents the first attempt to integrate early warning and rapid response strategies for small-to-moderate earthquakes, whose effects are typically confined to limited areas—a particularly challenging scenario for any early warning system. This approach establishes a framework for low-latency, impact-based rapid response in contexts where small magnitudes, short source-to-site distances, and minimal lead times limit the effectiveness of conventional EEW applications. In the Campi Flegrei area, where seismicity often occurs close to buildings and critical infrastructure, traditional protective actions for individuals (e.g., “Drop, Cover, and Hold On”) are impractical, due to the very short warning times. By contrast, automated safety measures—such as elevator shutdown, unlocking of emergency exits, and activation of illuminated escape route signage—can still be effectively triggered. The system proposed here therefore represents a valuable tool to support emergency management and decision-making immediately after the occurrence of felt earthquakes. A key aspect of our method is the quantification of the earthquake impact zone and its “radius of competence” around each station, i.e., the area in which the predicted PGV is expected to vary within +- X% of the predicted value, with X a pre-defined threshold. This area is determined using an adapted, region-specific GMPE [ 21 ] and varies as a function of magnitude, predicted PGV and the selected threshold. The quantification of the “radius of competence” extends the classical concept of an on-site early warning system, by defining the area surrounding the station that can benefit from the warning. The proposed system works upon definition of a threshold on the predicted PGV value to declare the alert. In our considered earthquake scenarios, we evaluated the alerting performance of the system using different thresholds values on IMCS, from ‘weak shaking’ (IMCS = III) to light damage (IMCS = V) [ 20 ]. The performance is evaluated in terms of Successful Alert (SA), Successful No Alert (SNA), False Alert (FA) and Missed Alert (MA), following a matrix scheme already proposed in other studies [ 17 , 18 , 22 ] and described in Data and Methods section. We choose the optimal PGV (IMCS) threshold value that minimizes the issuance of false alert while enabling rapid response for most of the earthquakes that produced “felt” shaking (Fig. 3 ). The test performed on the two scenario events show that, although the system cannot provide useful warning times (lead-time), due to the proximity between the earthquake hypocenter and the station, the system provides rapid impact assessment of the shaking distribution around the recording station. A potential network of distributed single-station nodes, even without communication between them, would provide a realistic map of the earthquake impact across the area (Figs. 4 and 5 ). The methodology proposed here can be easily integrated into other platforms for alert dissemination and emergency management during natural, catastrophic events. As an example, the system could be integrated into a Multi-Risk Impact-based Early Warning Systems (MR-IEWS) within the framework of the seismic risk assessment. The MR-IEWSs aim at improving impact forecasting for geo and weather hazards, building accessible user-friendly platforms for the dissemination of high-resolution information at the occurrence of potential catastrophic events. (e.g. https://gobeyond-project.eu/ ). A natural evolution of the proposed method is its direct connection to mobile apps or other devices (i.e, tablet, smartwatches) for the dissemination of earthquake alerts and immediate post-event information. Equipped with geolocation sensors, these devices can be used to track the position of people, before, during and after the occurrence of an earthquake, providing essential information to guide and prioritize rescue operations [ 23 , 24 ]. Providing immediate, reliable information right after an earthquake is extremely valuable for the population, for emergency management, and especially to counteract the potential spread of misinformation or fake news, effectively helping to reassure the public [ 25 , 26 ]. Finally, the methodology developed here may be exported and adapted to other densely populated volcanic or seismically active regions worldwide, where conventional regional EEW systems face similar physical limitations. Data & methods The initial dataset consists of earthquakes with duration magnitudes (Md) ranging from 1 to 4, which occurred in the Campi Flegrei caldera between January 1, 2016, and May 20, 2024, as reported by the ONT (Osservatorio Nazionale Terremoti, the Italian National Earthquake Observatory, https://terremoti.ingv.it). The original data were downloaded from the FDSN web service (waveforms available at http://webservices.ingv.it, last accessed on September 2025). Waveforms were recorded by the INGV-OV (Istituto Nazionale di Geofisica e Vulcanologia, Osservatorio Vesuviano section) seismic network, which includes 15 stations equipped with accelerometer and velocimeter sensors (represented as triangles in Figure 1). The event locations were refined by Scotto di Uccio et al.(2024)[27] (see Figure 1). To calibrate and test the system, we divided the dataset into two subsets: 80% of the waveforms (n = 2600) was used to calibrate the empirical scaling laws, while the remaining 20% (n = 700 waveforms) was used to test the system performance (see panel B of Figure 1 for subsets distribution). We used the vertical component of recorded ground acceleration. On each record, we performed a pre-processing which includes linear trend and mean removal. Each single trace was then double integrated to obtain the displacement, and the resulting signal was high-pass filtered with a cutoff frequency of 1 Hz, to remove any potential artificial low-frequency contamination. Finally, the signal-to-noise ratio (SNR) was computed as: SNR = 20*log(Pd PTW=1s /Pd NOISE ) where Pd NOISE is the peak displacement averaged over a fixed pre-event time window of 5 seconds and Pd PTW=1s is the P-peak of displacement within a fixed 1 second P-time Window following the P-wave arrival. Traces having SNR below a fixed threshold (SNR < 8 from Caruso et al. 2017) were discarded. The final dataset after the selection consisted in 2000 waveforms. After the pre-processing, the following step was to calibrate the empirical scaling law that links the P-peak of displacement (Pd) to the Peak Ground Velocity (PGV) [28,29]. We selected a 1-second P-wave time window to measure Pd. To ensure data quality, we evaluated the theoretical S-wave arrival times and excluded waveforms that may have been contaminated by later phases. The observed PGV was computed as the maximum among horizontal components [21]. The resulting scaling law in PTW = 1 second was obtained from a linear regression on the calibration subset as: logPGV = A+BlogPd(PTW=1s) ±SE where SE is the Standard Error. For the magnitude estimate from a single station, here we propose and experimented two methodologies. The first method consists of the computation of the seismic moment M 0 as the area underneath the source time function. Assuming a triangular shape for the source time function [30], the seismic moment is computed as: Where , c p , F s , are fixed. P d is assumed to be the peak of the source time function with the duration of . The 𝜏𝑐 is the characteristic period of the chosen P-signal and it is computed according to the approach of Kanamori and Allen (2003) [31] in a time window, PTW=1 second. Here we use the properties that 𝜏𝑐 is equal to the reciprocal of the corner frequency [32] and it approximates the rupture duration, is the average of the hypocentral distances distribution of each station in the area. The seismic moment is then converted to moment magnitude. The second method requires the calibration of the attenuation law between P x (x being acceleration, velocity or displacement) versus moment magnitude and hypocentral distances, in the form of: logP x = A’+B’Mw + C’logR where Mw is obtained from the Mw-Md law of Iervolino et al. (2024). We then compute the moment magnitude on the test subset using: M x = ± M Where A’, B’, C’ are the coefficients from the corresponding attenuation law and is the mean value of the hypocentral distances distribution of the single station. M Px is obtained as the weighted average of the three measures of Mx. Hence, the error on M Px arises from the propagation of errors on the attenuation law and the single station hypo distances distribution (see Figure S2). The performance of the system is evaluated in terms of: Successful Alert (SA): IMCS pred >IMCS threshold & IMCS obs >IMCS threshold Successful No-Alert (SNA): IMCS pred IMCS threshold & IMCS obs IMCS threshold False Alert (FA): IMCS pred IMCS threshold & IMCS obs <IMCS threshold Missed Alert (MA): IMCS pred <IMCS threshold & IMCS obs IMCS threshold . where IMCS pred is predicted by the system and IMCS threshold is a threshold value which potential users can set. Finally, we evaluated the impacted area around the station where the predicted PGV varies within ±50% of its value, called PGV pred150 and PGV pred50 respectively. Indeed, combining the magnitude estimate and the predicted PGV at the site with the GMPEs in the area (Scala et al 2025), we can evaluate Rmin and Rmax as it follows: Where A’’, B’’, C’’ are the coefficients of the GMPEs of the area of interest. The difference ΔR (R max -R min ) represents the radius of the impacted area around the station. In Figure S3 we showed expected values for ΔR in the Campi Flegrei area at different ranges of magnitude and PGV. Declarations Author Contribution V.L. Conceptualization; Formal analysis; Investigation; Methodology; Writing – original draft Writing – review & editingS.C. Conceptualization; Funding acquisition; Investigation; Methodology; Validation; Writing – original draft Writing – review & editingA.Z. Conceptualization; Funding acquisition; Investigation; Methodology; Project administration; Resources; Supervision; Validation; Writing – original draft Writing – review & editing References Allen, R. M. & Melgar, D. Earthquake Early Warning: Advances, Scientic Challenges, and Societal Needs. (2019). 10.1146/annurev-earth-053018 Del Gaudio, C., Aquino, I., Ricciardi, G. P., Ricco, C. & Scandone, R. Unrest episodes at Campi Flegrei: A reconstruction of vertical ground movements during 1905–2009. J. Volcanol. Geoth. Res. 195 , 48–56 (2010). Scarpa, R. et al. in 257–282 (2022). 10.1007/978-3-642-37060-1_10 Bevilacqua, A. et al. Data analysis of the unsteadily accelerating GPS and seismic records at Campi Flegrei caldera from 2000 to 2020. Sci. Rep. 12 , 19175 (2022). De Landro, G. et al. 3D structure and dynamics of Campi Flegrei enhance multi-hazard assessment. Nat. Commun. 16 , 4814 (2025). Nadeau, B. ‘Be prepared for all outcomes’: Inside the saga of a supervolcano that’s waking up. (2023). Iervolino, I. et al. Seismic risk mitigation at Campi Flegrei in volcanic unrest. Nat. Commun. 15 , 10474 (2024). Kuyuk, H. S. & Allen, R. M. Optimal Seismic Network Density for Earthquake Early Warning: A Case Study from California. Seismol. Res. Lett. 84 , 946–954 (2013). Zollo, A. et al. in 117–137 (2014). 10.1007/978-3-642-12233-0_7 Meier, M. A. et al. How Often Can Earthquake Early Warning Systems Alert Sites With High-Intensity Ground Motion? J Geophys. Res. Solid Earth 125 , (2020). Satriano, C. et al. PRESTo, the earthquake early warning system for Southern Italy: Concepts, capabilities and future perspectives. Soil Dyn. Earthq. Eng. 31 , 137–153 (2011). Lux, A. I. et al. Status and Performance of the ShakeAlert Earthquake Early Warning System: 2019–2023. Bull. Seismol. Soc. Am. 114 , 3041–3062 (2024). Tunç, S. et al. Implementation of the EPIC Earthquake Early Warning System in the Bursa Province (Türkiye) and Its Surroundings. Appl. Sci. 13 , 4985 (2023). Ionescu, C., Marmureanu, A. & Marmureanu, G. in 441–449 (2016). 10.1007/978-3-319-29844-3_31 Kodera, Y. et al. The Propagation of Local Undamped Motion (PLUM) Method: A Simple and Robust Seismic Wavefield Estimation Approach for Earthquake Early Warning. Bull. Seismol. Soc. Am. 108 , 983–1003 (2018). Zollo, A., Colombelli, S., Caruso, A. & Elia, L. An Evolutionary Shaking-Forecast-Based Earthquake Early Warning Method. Earth Space Science 10 , (2023). Caruso, A., Colombelli, S., Elia, L., Picozzi, M. & Zollo, A. An on-site alert level early warning system for Italy. J. Geophys. Res. Solid Earth . 122 , 2106–2118 (2017). Colombelli, S., Caruso, A., Zollo, A., Festa, G. & Kanamori, H. A P wave-based, on‐site method for earthquake early warning. Geophys. Res. Lett. 42 , 1390–1398 (2015). Zollo, A., Nazeri, S. & Colombelli, S. Earthquake Seismic Moment, Rupture Radius, and Stress Drop from P-Wave Displacement Amplitude Versus Time Curves. IEEE Trans. Geoscience Remote Sensing 60 , (2022). Faenza, L. & Michelini, A. Regression analysis of MCS intensity and ground motion parameters in Italy and its application in ShakeMap. Geophys. J. Int. 180 , 1138–1152 (2010). Scala, A. et al. Ground Motion Prediction Equations for Campi Flegrei (Italy). (2025). 10.21203/rs.3.rs-5758860/v1 Rea, R., Colombelli, S., Elia, L. & Zollo, A. Retrospective performance analysis of a ground shaking early warning system for the 2023 Turkey–Syria earthquake. Commun. Earth Environ. 5 , 332 (2024). Colombelli, S., Carotenuto, F., Elia, L. & Zollo, A. Design and implementation of a mobile device app for network-based earthquake early warning systems (EEWSs): application to the PRESTo EEWS in southern Italy. Nat. Hazards Earth Syst. Sci. 20 , 921–931 (2020). Acampora, G., Carotenuto, F. & Zollo, A. A. Proposal for a Fuzzy-based Architecture for Rescue Operations in Earthquake Scenarios. IEEE International Conference on Fuzzy Systems (FUZZ) (FUZZ-IEEE 2025) 6 Preprint at (2025) 6 Preprint at (2025) (2025). Sutton, J., Wood, M. M., Crouch, S. & Waugh, N. Public perceptions of U.S. earthquake early warning post-alert messages: Findings from focus groups and interviews. Int. J. Disaster Risk Reduct. 84 , 103488 (2023). Fallou, L., Corradini, M., Bossu, R. & Cheny, J. M. Preventing and debunking earthquake misinformation: Insights into EMSC’s practices. Front Commun. (Lausanne) 7 , (2022). di Scotto, F. et al. Delineation and Fine-Scale Structure of Fault Zones Activated During the 2014–2024 Unrest at the Campi Flegrei Caldera (Southern Italy) From High‐Precision Earthquake Locations. Geophys Res. Lett 51 , (2024). Kanamori, H. Real-time seismology and earthquake damage mitigation. Annu. Rev. Earth Planet. Sci. 33 , 195–214 (2005). Zollo, A., Lancieri, M. & Nielsen, S. Earthquake magnitude estimation from peak amplitudes of very early seismic signals on strong motion records. Geophys Res. Lett 33 , (2006). Aki, K. & Richards, P. G. Quantitative seismology (University Science Books, 2009). Allen, R. M. & Kanamori, H. The Potential for Earthquake Early Warning in Southern California. Sci. (1979) . 300 , 786–789 (2003). Lior, I., Ziv, A. & Madariaga, R. P. -Wave Attenuation with Implications for Earthquake Early Warning. Bull. Seismol. Soc. Am. 106 , 13–22 (2016). Additional Declarations No competing interests reported. Supplementary Files Supplementarymaterial.docx Cite Share Download PDF Status: Published Journal Publication published 08 Apr, 2026 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 17 Sep, 2025 Editor assigned by journal 17 Sep, 2025 Submission checks completed at journal 16 Sep, 2025 First submitted to journal 15 Sep, 2025 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-7619275","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":515226106,"identity":"898353dd-3803-4054-87a3-68819873bd06","order_by":0,"name":"Valeria Longobardi","email":"","orcid":"","institution":"University of Naples “Federico II”","correspondingAuthor":false,"prefix":"","firstName":"Valeria","middleName":"","lastName":"Longobardi","suffix":""},{"id":515226107,"identity":"052985cd-b394-429b-bb5a-47cfba17323a","order_by":1,"name":"Simona Colombelli","email":"","orcid":"","institution":"University of Naples “Federico II”","correspondingAuthor":false,"prefix":"","firstName":"Simona","middleName":"","lastName":"Colombelli","suffix":""},{"id":515226108,"identity":"3240effc-6fe1-4de8-8df8-1b2a9a73c2ca","order_by":2,"name":"Aldo Zollo","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAzUlEQVRIiWNgGAWjYDACdgYDECXDx8B8gAHMZiakhRmihYeNgS0BqoWQHoQWHgO4EF7A38y88QHDHxseNvaeb9IFBXfkGdj5D+DVInGYrdiAgSeNh43n7DbpGQbPDBsIOuwwj5kEUCMPm0TuNmkeg8OMBLXIg7UYgLTkPANpsSeoxQCsJQGshQ2kJZGgFkOQXxIOgPxyzNgaqCW5jZnZAK8WuePNGx98+GMjx8/e/PA2z5/Dtv38Bx/gtwYEEpA5bITVj4JRMApGwSggBAAEMjLq+MqAYAAAAABJRU5ErkJggg==","orcid":"","institution":"University of Naples “Federico II”","correspondingAuthor":true,"prefix":"","firstName":"Aldo","middleName":"","lastName":"Zollo","suffix":""}],"badges":[],"createdAt":"2025-09-15 10:08:48","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7619275/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7619275/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-026-42593-x","type":"published","date":"2026-04-08T15:58:24+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":91401529,"identity":"5e7e6fac-e8cd-4906-8c60-7f4560680b6f","added_by":"auto","created_at":"2025-09-16 07:07:58","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":861959,"visible":true,"origin":"","legend":"\u003cp\u003ePanel A shows the epicentral location of selected events as well as the position of used stations. As for earthquakes (circles), marker color follows the event depth and marker size follows the event magnitude (duration magnitude Md). Dark orange triangles are seismic stations. Panel B shows the distribution in magnitude and epicentral distance of waveforms from the calibration dataset (blue points and histograms) and from the test dataset (green points and histograms).\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7619275/v1/84d4bece6ffacf64cf8eb73b.jpeg"},{"id":91402601,"identity":"2bf9b8b4-a318-4258-b420-ff2f68b17306","added_by":"auto","created_at":"2025-09-16 07:15:58","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":665808,"visible":true,"origin":"","legend":"\u003cp\u003ePanel A shows the scaling of PGV vs PD in a one-second P-time window. Dark grey crosses represent single station measurements on the train dataset. Cyan points represent 2-d binned data (x-bin width = 1 cm and y-bin width = 0.5 cm/s). Dashed lines are the calibrated PGV vs Pd law and its standard error bounds for the 2-d binned data. Panel B shows the single station magnitude estimate from Pd x tauc method (Mew) in one-second P-time window versus observed moment magnitude (MW). Black solid line is the one-to-one MW line. Inset shows the histogram of the difference between Mew and MW. Mean and standard deviation of the distribution are reported at top left corner.\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7619275/v1/6eabb8641a300debcffb1aeb.jpeg"},{"id":91401525,"identity":"31fbabc7-95f7-4425-a4d6-f10365d83447","added_by":"auto","created_at":"2025-09-16 07:07:58","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":106923,"visible":true,"origin":"","legend":"\u003cp\u003eSystem performance on the test dataset in PTW= 1s. The performance is assessed in terms of Successful alert (SA) (dark green), Successful No-alert (SNA) (light green), Missed Alert (MA) (red) and False Alert (FA) (yellow) for different thresholds in IMCS. For each IMCS\u003csub\u003ethreshold\u003c/sub\u003e, the percentage of SA, SNA, MA, and FA is shown. Panel A shows the performance for IMCS\u003csub\u003ethreshold\u003c/sub\u003e=3; Panel B shows the performance for IMCS\u003csub\u003ethreshold\u003c/sub\u003e=4; Panel C shows the performance for IMCS\u003csub\u003ethreshold\u003c/sub\u003e=5.\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7619275/v1/9190048325a574bf365fb31e.jpeg"},{"id":91401527,"identity":"dce10f2e-4d63-4ff5-918d-ac66dfe6ab53","added_by":"auto","created_at":"2025-09-16 07:07:58","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":236914,"visible":true,"origin":"","legend":"\u003cp\u003eTest on the Md 4.4 earthquake of May 20, 2024. The map shows the result obtained in PTW=1 second. Magenta star represents the earthquake location (the moment magnitude obtained from Md-Mw scaling law by Iervolino et al 2024 is also shown). Each circle represents the area centered on the station within which the predicted PGV varies between +-50% of its value. Color follows PGV. In each circle, the estimated shaking intensity (IMCS) at the site is also reported.\u003c/p\u003e","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7619275/v1/6c00a22f9698612ccc760f48.jpeg"},{"id":91401526,"identity":"de5d683e-b227-415c-abc6-55133edd386b","added_by":"auto","created_at":"2025-09-16 07:07:58","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":270294,"visible":true,"origin":"","legend":"\u003cp\u003eTest on the Md 4.6 earthquake of March 13, 2025. The map shows the result obtained in PTW=1 second. Magenta star represents the earthquake location (the moment magnitude obtained from Md-Mw scaling law by Iervolino et al 2024 is also shown). Each circle represents the area centered on the station within which the predicted PGV varies between +-50% of its value. Color follows PGV. In each circle, the estimated shaking intensity (IMCS) at the site is also reported.\u003c/p\u003e","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7619275/v1/effbc30ddbafd2784d8e76f1.jpeg"},{"id":106808875,"identity":"80f39d8f-fb17-4988-aea9-686448a84e3c","added_by":"auto","created_at":"2026-04-13 16:04:11","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2593454,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7619275/v1/3bca6837-3ba8-4660-b1b2-73e92cc7edbb.pdf"},{"id":91401528,"identity":"5f9a0948-83ed-4a2b-8643-3e04e1552d05","added_by":"auto","created_at":"2025-09-16 07:07:58","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":2459981,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarymaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-7619275/v1/37d9278134ea22b94806e66b.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"One-Second-Lead Earthquake Warning and Impact Assessment at Campi Flegrei","fulltext":[{"header":"Introduction","content":"\u003cp\u003eIn volcanic and seismic regions, during the occurrence of seismic crises, the timely dissemination of earthquake information can significantly mitigate the earthquake impact by improving emergency safety actions [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eA compelling case study is the current situation at the Campi Flegrei caldera in Southern Italy, where the ongoing volcanic unrest is raising increasing concern among local population and authorities in the area.\u003c/p\u003e\u003cp\u003eDuring the past 75 years, the caldera has experienced repeated episodes of ground deformation, known as bradyseism (slow ground up- and downlift), documented since the 1950s [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. The two most rapid phases of uplift occurred in 1970\u0026ndash;72 and 1980\u0026ndash;84 [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Since 2005, a new monotonic uplift episode has been underway, with a notable acceleration in seismicity observed from 2014 onwards [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. During the last two years, the land instability with high uplift rates reaching 2\u0026ndash;3 cm per month, is being accompanied by an intense and frequent seismic activity with the occurrence of several Md 4\u0026thinsp;+\u0026thinsp;earthquakes originated at shallow depths (3 km) and causing strong ground shaking (up to several tenths of g) in restricted, few km radius, areas around the epicenter [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. The metropolitan area surrounding the Campi Flegrei caldera is densely populated, with several hundred thousand residents [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. The recent intense seismic activity\u0026mdash;particularly the sudden uplift in the Pozzuoli area\u0026mdash;has caused damage to infrastructure and buildings, increasing the urgency for enhanced seismic risk mitigation [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Automatic and near-real-time monitoring systems are currently in operation [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], but the time required to detect, process, and disseminate earthquake information (up to several tens of minutes) \u0026mdash;especially for the lower magnitude but well perceived earthquakes\u0026mdash;remains a critical limitation in real-time emergency management [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eSince these earthquakes are generally small-to-moderate in magnitude, the area affected by strong shaking is also limited to a few kilometers around the epicenter and comparable with the \u0026ldquo;blind zone\u0026rdquo; of the warning systems, which is defined as the area where the strong shaking occurs before the alert can be issued.\u003c/p\u003e\u003cp\u003eBecause of this physical constraint the applicability of earthquake early warning (EEW) systems in such a small extent volcanic region, as the Campi Flegrei area, is limited. However, real-time data processing and methodologies can still be applied and locally calibrated to provide a P-wave-based rapid response within a short time window after the earthquake occurrence. This approach may be particularly useful for the largest magnitude, strong shaking causative events (M 3\u0026ndash;4), which are clearly felt by residents [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e] over a wider area (10\u0026ndash;15 km). The rapid response might encompass different aspects of risk mitigation, including the evaluation of shake maps, the quick assessment of damage to guide emergency actions as well as the rapid dissemination of earthquake information to reduce the spread of misinformation and increase the awareness among the population.\u003c/p\u003e\u003cp\u003eWithin this frame, we tailored a standard EEW approach to the peculiar seismicity of the Campi Flegrei area, to provide early P-wave based shake and potential damage maps, that can become crucial tools to support Civil Protection Authorities (CPAs) and first responders during emergencies.\u003c/p\u003e\u003cp\u003eEEW approaches are broadly classified into source-based and impact-based methods [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Source-based EEW systems rapidly estimate earthquake source parameters (e.g. location and magnitude) to predict ground shaking at distant sites before the arrival of damaging waves. While effective in tectonic regions with larger, longer-duration earthquakes [\u003cspan additionalcitationids=\"CR12 CR13\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e], this approach faces severe limitations in areas like Campi Flegrei where seismic events are typically low-to-moderate in magnitude, very shallow, and short in duration [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eImpact-based methods bypass instead the need for full source characterization [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. They rely directly on ground motion observations at individual stations to estimate the potential earthquake impact in real time. Here, we propose a hybrid on-site (EEW) methodology adapting an existing method [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], to the Campi Flegrei case-study. The early P-displacement signal is used to predict peak ground motion at the site, while at the same time an approximate estimation of the magnitude is obtained to characterize the event size. We calibrate and test a rapid response system using a large dataset of earthquakes recorded during the recent unrest at Campi Flegrei. Additionally, show the example of its performance considering two significant earthquake scenarios: the Md 4.4 event of May 20, 2024, and the Md 4.6 event of March 13, 2025.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eTo calibrate and test the system we used 700 events with duration magnitude Md between 1 and 4, which occurred during 2016-2024 at Campi Flegrei caldera (panel A of Figure 1). The preliminary dataset is split into two subsets: the training dataset is used for calibration of empirical relationships, and the test dataset is used for performance evaluation (see panel B of Figure 1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe focused here on the estimation, at a single station, with the onsite method of two quantities: the peak ground motion (PGV and PGA) and the earthquake magnitude.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAs for the PGV prediction, in panel A of Figure 1 we show the scaling of observed PGV with the P-peak displacement (Pd) measured within a one-second P-wave time window for the training dataset. To account for the unbalanced number of earthquakes in the different magnitude classes (see panel B of Figure 1), we calibrate the empirical scaling laws of PGV vs. Pd using 2-D binned data (represented as cyan points in panel A of Figure 2). The solid and dashed lines represent the resulting PGV\u0026ndash;Pd scaling law and its standard error bounds, respectively. The results of the calibration and the test for the PGA are shown in the supplemental material (figure S4 and S5).\u003c/p\u003e\n\u003cp\u003eAs for magnitude, in panel B of Figure 2, we present the single-station magnitude estimate for a P-wave time window of 1 second, derived from the seismic moment (M₀). The seismic moment is estimated as the area under a triangular source time function, with Pd as the peak and tauc (period parameter) as the duration [19] (see Data \u0026amp; Methods section). We refer to this magnitude estimate as Mew. The Mew method relies on two assumptions: (1) the far-field displacement can be modeled as a triangle with amplitude Pd and duration \u0026nbsp;\u0026tau;c, and (2) the hypocentral distance used in the estimate (see Data \u0026amp; Methods section) is an average value obtained from the distribution of all possible hypocentral distances for the given station (see Figure S1) determined from the recorded dataset, rather than its exact value.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis latter assumption appears valid for a small area such as the Campi Flegrei caldera, where variations in hypocentral distance are limited due to the distribution of stations and event locations (see Figure 1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eNevertheless, the error on Mew determinations is related to the real distance and the empirical distribution of earthquake distances from the concerned station as inferred from the current seismicity. To allow a direct comparison, we computed the moment magnitude Mw, using the scaling relationship of Iervolino et al. (2024) [7] that links catalog duration magnitude Md with the moment magnitude Mw (shown on the x-axis of panel B in Figure 2).\u003c/p\u003e\n\u003cp\u003eThe prediction error, defined as the difference between Mew and moment magnitude (Mw), is shown in the inset histogram of panel B of Figure 2. The average error is centered around zero, with a standard deviation of 0.36 magnitude units.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe also applied an alternative method to estimate the single-station magnitude by using the empirical relationship that relates log Px (x being acceleration, velocity and displacement) to the observed magnitude and observed log-hypocentral distance (see supplementary material text S1 and figure S1 for more details). We define this estimate as MPx. We compare MPx with Mw in figure S2. The prediction error \u0026ndash; the difference between MPx and Mw - shows a higher value of its standard deviation (0.41 magnitude units) with respect to what it is obtained for Mew.\u003c/p\u003e\n\u003cp\u003eThe Earthquake Early Warning and Rapid Response (EEWRR) system implemented at Campi Flegrei is designed to provide real-time estimates of earthquake impact within the first second after the arrival of the P-wave. As soon as the vertical component of ground motion is detected, key parameters (Pd, Pa, Pv, and \u0026tau;c) are extracted from the initial portion of the waveform. Among these, Pd and \u0026tau;c are used to derive a first-order estimate of the seismic moment and corresponding moment magnitude. In parallel, empirical ground-motion prediction equations (GMPEs) are applied to the observed P-wave peak amplitudes in order to forecast the expected peak ground velocity (PGV) and peak ground acceleration (PGA).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAssuming a GMPE of the form log(PGV) = A + B\u0026middot;Mw + C log(R) (A,B,C are estimated from data), where M is the estimated magnitude and R the hypocentral distance, the system determines a circular area around the recording station within which PGV is expected to vary by \u0026plusmn;X%, with X defined by the user.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe final outputs consist of:\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e(i) the predicted PGV and PGA at the site, which can be converted into macroseismic intensity using the empirical relations of Faenza and Michelini [20];\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e(ii) an approximate Mw estimate;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e(iii) the extent of the area surrounding the site where shaking levels are \u0026nbsp;expected to fall within the chosen uncertainty range.\u003c/p\u003e\n\u003cp\u003eIn Figure 4 and 5, we show the test on two earthquake scenarios, the Md 4.4 occurred on May 20, 2024 and the Md 4.6 occurred on March 13, 2025 respectively. These two events \u0026ndash; among the largest registered in the area- are not included in the dataset used for calibration and testing of the system. For both scenarios, we computed the predicted impact area around each station within which the predicted PGV (PGV\u003csub\u003epred\u003c/sub\u003e) may vary between +-50% of its value. Since each single station can provide the expected PGV (PGV\u003csub\u003epred\u003c/sub\u003e) and the magnitude (Mew) estimate, the couples (PGV\u003csub\u003epred\u003c/sub\u003e - 50%PGV\u003csub\u003epred\u003c/sub\u003e, Mew) and (PGV\u003csub\u003epred\u003c/sub\u003e + 50%PGV\u003csub\u003epred\u003c/sub\u003e, Mew) can be used to evaluate at first order the area within which this variation of the PGV will be experienced. In this way, each station defines its own \u0026ldquo;radius of competence,\u0026rdquo; that is, the area around the station in which we expect stable values of predicted PGV. The size of this radius depends on the GMPE adopted for the region. Here, we use the GMPE from Scala et al. (2025) [21], truncated to retain only the first term of hypocentral distance (see Data \u0026amp; Methods for details).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we demonstrate the feasibility of implementing an on-site early warning system, tailored to the specific characteristics of the Campi Flegrei region. The Campi Flegrei caldera is a very densely populated area, with 500 thousand inhabitants distributed over a limited territory (200km\u003csup\u003e2\u003c/sup\u003e) [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. The presence of critical infrastructure elements (such as schools, hospitals, emergency response centers, and transportation hubs) makes the proposed real-time earthquake alert system relevant for risk mitigation.\u003c/p\u003e\u003cp\u003eThe system can predict earthquake impact at the site in terms of PGV (or PGA) and delineate a circular area within which this value is expected to vary by a user-defined percentage. This piece of information is compiled into an alert message that can be immediately broadcast to the exposed population, enabling the prompt activation of emergency safety measures\u0026mdash;a form of local, real-time shakemap. The entire process is completed within roughly one second from the P-wave detection at the site (see Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e4\u003c/span\u003e and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e5\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe proposed methodology represents the first attempt to integrate early warning and rapid response strategies for small-to-moderate earthquakes, whose effects are typically confined to limited areas\u0026mdash;a particularly challenging scenario for any early warning system. This approach establishes a framework for low-latency, impact-based rapid response in contexts where small magnitudes, short source-to-site distances, and minimal lead times limit the effectiveness of conventional EEW applications. In the Campi Flegrei area, where seismicity often occurs close to buildings and critical infrastructure, traditional protective actions for individuals (e.g., \u0026ldquo;Drop, Cover, and Hold On\u0026rdquo;) are impractical, due to the very short warning times. By contrast, automated safety measures\u0026mdash;such as elevator shutdown, unlocking of emergency exits, and activation of illuminated escape route signage\u0026mdash;can still be effectively triggered. The system proposed here therefore represents a valuable tool to support emergency management and decision-making immediately after the occurrence of felt earthquakes.\u003c/p\u003e\u003cp\u003eA key aspect of our method is the quantification of the earthquake impact zone and its \u0026ldquo;radius of competence\u0026rdquo; around each station, i.e., the area in which the predicted PGV is expected to vary within +- X% of the predicted value, with X a pre-defined threshold. This area is determined using an adapted, region-specific GMPE [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e] and varies as a function of magnitude, predicted PGV and the selected threshold. The quantification of the \u0026ldquo;radius of competence\u0026rdquo; extends the classical concept of an on-site early warning system, by defining the area surrounding the station that can benefit from the warning.\u003c/p\u003e\u003cp\u003eThe proposed system works upon definition of a threshold on the predicted PGV value to declare the alert. In our considered earthquake scenarios, we evaluated the alerting performance of the system using different thresholds values on IMCS, from \u0026lsquo;weak shaking\u0026rsquo; (IMCS\u0026thinsp;=\u0026thinsp;III) to light damage (IMCS\u0026thinsp;=\u0026thinsp;V) [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. The performance is evaluated in terms of Successful Alert (SA), Successful No Alert (SNA), False Alert (FA) and Missed Alert (MA), following a matrix scheme already proposed in other studies [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e] and described in Data and Methods section.\u003c/p\u003e\u003cp\u003eWe choose the optimal PGV (IMCS) threshold value that minimizes the issuance of false alert while enabling rapid response for most of the earthquakes that produced \u0026ldquo;felt\u0026rdquo; shaking (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe test performed on the two scenario events show that, although the system cannot provide useful warning times (lead-time), due to the proximity between the earthquake hypocenter and the station, the system provides rapid impact assessment of the shaking distribution around the recording station. A potential network of distributed single-station nodes, even without communication between them, would provide a realistic map of the earthquake impact across the area (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e4\u003c/span\u003e and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e5\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe methodology proposed here can be easily integrated into other platforms for alert dissemination and emergency management during natural, catastrophic events. As an example, the system could be integrated into a Multi-Risk Impact-based Early Warning Systems (MR-IEWS) within the framework of the seismic risk assessment. The MR-IEWSs aim at improving impact forecasting for geo and weather hazards, building accessible user-friendly platforms for the dissemination of high-resolution information at the occurrence of potential catastrophic events. (e.g. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://gobeyond-project.eu/\u003c/span\u003e\u003cspan address=\"https://gobeyond-project.eu/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e ).\u003c/p\u003e\u003cp\u003eA natural evolution of the proposed method is its direct connection to mobile apps or other devices (i.e, tablet, smartwatches) for the dissemination of earthquake alerts and immediate post-event information. Equipped with geolocation sensors, these devices can be used to track the position of people, before, during and after the occurrence of an earthquake, providing essential information to guide and prioritize rescue operations [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Providing immediate, reliable information right after an earthquake is extremely valuable for the population, for emergency management, and especially to counteract the potential spread of misinformation or fake news, effectively helping to reassure the public [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eFinally, the methodology developed here may be exported and adapted to other densely populated volcanic or seismically active regions worldwide, where conventional regional EEW systems face similar physical limitations.\u003c/p\u003e"},{"header":"Data \u0026 methods","content":"\u003cp\u003eThe initial dataset consists of earthquakes with duration magnitudes (Md) ranging from 1 to 4, which occurred in the Campi Flegrei caldera between January 1, 2016, and May 20, 2024, as reported by the ONT (Osservatorio Nazionale Terremoti, the Italian National Earthquake Observatory, https://terremoti.ingv.it). The original data were downloaded from the FDSN web service (waveforms available at http://webservices.ingv.it, last accessed on September 2025). Waveforms were recorded by the INGV-OV (Istituto Nazionale di Geofisica e Vulcanologia, Osservatorio Vesuviano section) seismic network, which includes 15 stations equipped with accelerometer and velocimeter sensors (represented as triangles in Figure 1). The event locations were refined by Scotto di Uccio et al.(2024)[27] (see Figure 1).\u003cbr\u003eTo calibrate and test the system, we divided the dataset into two subsets: 80% of the waveforms (n = 2600) was used to calibrate the empirical scaling laws, while the remaining 20% (n = 700 waveforms) was used to test the system performance (see panel B of Figure 1 for subsets distribution).\u003c/p\u003e\n\u003cp\u003eWe used the vertical component of recorded ground acceleration. On each record, we performed a pre-processing which includes linear trend and mean removal. Each single trace was then double integrated to obtain the displacement, and the resulting signal was high-pass filtered with a cutoff frequency of 1 Hz, to remove any potential artificial low-frequency contamination.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFinally, the signal-to-noise ratio (SNR) was computed as:\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSNR = 20*log(Pd\u003csub\u003ePTW=1s\u003c/sub\u003e/Pd\u003csub\u003eNOISE\u003c/sub\u003e)\u003c/p\u003e\n\u003cp\u003ewhere Pd\u003csub\u003eNOISE\u003c/sub\u003e is the peak displacement averaged over a fixed pre-event time window of 5 seconds and Pd\u003csub\u003ePTW=1s\u003c/sub\u003e is the P-peak of displacement within a fixed 1 second P-time Window following the P-wave arrival. Traces having SNR below a fixed threshold (SNR \u0026lt; 8 from Caruso et al. 2017) were discarded.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe final dataset after the selection consisted in 2000 waveforms.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAfter the pre-processing, the following step was to calibrate the empirical scaling law that links the P-peak of displacement (Pd) to the Peak Ground Velocity (PGV) [28,29].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe selected a 1-second P-wave time window to measure Pd. To ensure data quality, we evaluated the theoretical S-wave arrival times and excluded waveforms that may have been contaminated by later phases.\u003c/p\u003e\n\u003cp\u003eThe observed PGV was computed as the maximum among horizontal components [21]. The resulting scaling law in PTW = 1 second was obtained from a linear regression on the calibration subset as:\u003c/p\u003e\n\u003cp\u003elogPGV = A+BlogPd(PTW=1s) \u0026plusmn;SE\u003c/p\u003e\n\u003cp\u003ewhere SE is the Standard Error.\u003c/p\u003e\n\u003cp\u003eFor the magnitude estimate from a single station, here we propose and experimented \u0026nbsp;two methodologies.\u003c/p\u003e\n\u003col\u003e\n \u003cli\u003eThe first method consists of the computation of the seismic moment M\u003csub\u003e0\u003c/sub\u003e as the area underneath the source time function. Assuming a triangular shape for the source time function [30], the\u0026nbsp;seismic moment is computed as:\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003e\u003cimg width=\"130\" height=\"44\" src=\"data:image/png;base64,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\" alt=\"image\"\u003e\u003c/p\u003e\n\u003cp\u003eWhere\u0026nbsp;\u003cimg width=\"11\" height=\"15\" src=\"data:image/png;base64,R0lGODlhEAAXAHcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAAACgALAA0AhAAAAAAAAAAAOgAAZgA6OgA6ZgA6kABmtjoAADpmtjqQtjqQ22YAAGY6AGY6Oma222a2/5A6AJDb/7ZmALbb/7b//9uQOtu2Ztv///+2Zv/btv//tv//2wECAwECAwECAwVEICACWhMEBTUCWaBgXDRU4sYcoyVIos73AQggZsCMJjuAbTGyFVnBXBRgCTABF0RCFCs4ToRHk3FdiVpCc2+mFk2e7RAAOw==\" alt=\"image\"\u003e, c\u003csub\u003ep\u003c/sub\u003e, F\u003csub\u003es\u003c/sub\u003e,\u0026nbsp;\u003cimg width=\"24\" height=\"17\" src=\"data:image/png;base64,R0lGODlhJAAaAHcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAAABQAiABUAhQAAAAAAAAAAOgAAZgA6OgA6ZgA6kABmtjoAADoAOjo6ADo6Ojo6kDpmtjqQ22YAAGYAOmY6AGaQ22a222a2/5A6AJA6OpBmOpDb/7ZmALZmOraQZraQkLbbkLbb/7b//9uQOtu2Ztu2kNvbkNv/ttv/29v///+2Zv/bkP/btv//tv//2wECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwbzQMBKEwgUMAChRoBMOp/QKEAzAaQizKsiwJR6oSiGKXkKEBpjUODwbQNAXQAKwU6qHnW3tCyo3g1jAHcUel8hCEx/Y0NVhV9zFH8lGwiAjl8pEBh3AQ0lFQGEl1JUgg+WanlJV0ZNjooAZapzVRmqend1sk4rFXUglm4pDxQoCU27ZHHAgUOREa1y0AILrmUGHRIAGcG2SSsWE70O2wcrHCOVgXYPAYC9RfHk5bEDH/RJG/dQJ2Km8yo0yTn2po63V8SSgKhjDMM4OQhEFeqlLYS9b75IaBKBoMEoiEY8PElxIR6BRh/bnAiW0s3Blm0ePgkCADs=\" alt=\"image\"\u003e\u0026nbsp;are fixed. P\u003csub\u003ed\u003c/sub\u003e is assumed to be the peak of the source time function with the duration of\u0026nbsp;\u003cimg width=\"12\" height=\"15\" src=\"data:image/png;base64,R0lGODlhEgAXAHcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAAACgARAA0AhAAAAAAAAAAAOgA6kABmtjoAADoAOjo6kDpmkDqQtjqQ22YAAGY6kGaQtma222a2/5A6AJBmOpDb/7ZmANuQOtuQZtu2kNv///+2Zv/bkP/btv//tv//2wECAwECAwECAwVKIFAVQVkOF6CuYnNti8LOc1Y8dA5QgqTPHAjqx4IRiCybbKaJnFIATOChYUBhiQvmqAoGEFDAZIgEwHBlQMbgS8McAMshrBt9oSEAOw==\" alt=\"image\"\u003e. \u0026nbsp;The\u0026nbsp;𝜏𝑐\u0026nbsp;is the characteristic period of the chosen P-signal and it is computed according to the approach of Kanamori and Allen (2003) [31] in a time window, PTW=1 second. Here we use the properties that\u0026nbsp;𝜏𝑐\u0026nbsp;is equal to the reciprocal of the corner frequency [32] and it approximates the rupture duration, \u0026lt;R\u0026gt; is the average of the hypocentral distances distribution of each station in the area. The seismic moment is then converted to moment magnitude.\u0026nbsp;\u003c/p\u003e\n\u003col start=\"2\"\u003e\n \u003cli\u003eThe second method requires the calibration of the attenuation law between P\u003csub\u003ex\u003c/sub\u003e (x being acceleration, velocity or displacement) versus moment magnitude and hypocentral distances, in the form of:\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003elogP\u003csub\u003ex\u003c/sub\u003e = A\u0026rsquo;+B\u0026rsquo;Mw + C\u0026rsquo;logR\u003c/p\u003e\n\u003cp\u003ewhere Mw is obtained from the Mw-Md law of Iervolino et al. (2024). We then compute the moment magnitude on the test subset using:\u003c/p\u003e\n\u003cp\u003eM\u003csub\u003ex\u003c/sub\u003e = \u003cimg width=\"105\" height=\"29\" src=\"data:image/png;base64,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\" alt=\"image\"\u003e\u0026plusmn;\u003cimg width=\"10\" height=\"19\" src=\"data:image/png;base64,R0lGODlhDwAcAHcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAAADAAPAAsAhAAAAAAAAAAAOgAAZgA6OgA6ZgA6kABmtjoAADo6ADqQ22YAAGY6AGZmOmaQtma2/5A6AJC225Db/7b//9uQOtu2Ztu2kNv///+2Zv/bkP//tv//2wECAwECAwECAwECAwVJICCKVpMEKGpcY8akcHC0SFBEADVMo6gtMhYgg5j1KIEVTSnaQAKPHpEJcO56mKQQ8KM6A4qelSfKXrFgUQUhkPRGllOA4NiOQgA7\" alt=\"image\"\u003eM\u003c/p\u003e\n\u003cp\u003eWhere A\u0026rsquo;, B\u0026rsquo;, C\u0026rsquo; are the coefficients from the corresponding attenuation law and \u0026lt;R\u0026gt; is the mean value of the hypocentral distances distribution of the single station. M\u003csub\u003ePx\u003c/sub\u003e is obtained as the weighted average of the three measures of Mx. Hence, the error on M\u003csub\u003ePx\u003c/sub\u003e arises from the propagation of errors on the attenuation law and the single station hypo distances distribution (see Figure S2).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe performance of the system is evaluated in terms of:\u003c/p\u003e\n\u003cp\u003eSuccessful Alert (SA): IMCS\u003csub\u003epred\u003c/sub\u003e\u0026gt;IMCS\u003csub\u003ethreshold\u0026nbsp;\u003c/sub\u003e\u0026amp; IMCS\u003csub\u003eobs\u003c/sub\u003e\u0026gt;IMCS\u003csub\u003ethreshold\u003c/sub\u003e\u003c/p\u003e\n\u003cp\u003eSuccessful No-Alert (SNA): IMCS\u003csub\u003epred\u003c/sub\u003e\u003cimg width=\"12\" height=\"19\" src=\"data:image/png;base64,R0lGODlhEgAcAHcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAEACQAQAA4AhAAAAAAAAAAAOgA6ZjoAADo6ADo6ZjpmkGY6OmaQtpBmOpC225Db/7aQZrbb/9u2Ztu2kNvbttv///+2Zv/bkP/btv/b2///2wECAwECAwECAwECAwECAwECAwECAwECAwVCICCO5AghTFlGSnA4qlgphLHEgNUQwyLFl0dBkPjhJoEernRpDItLko7ni45mtZtVxHLBtoBTCkwmUQiBtHqdFjBCADs=\" alt=\"image\"\u003eIMCS\u003csub\u003ethreshold\u0026nbsp;\u003c/sub\u003e\u0026amp; IMCS\u003csub\u003eobs\u003c/sub\u003e\u003cimg width=\"12\" height=\"19\" src=\"data:image/png;base64,R0lGODlhEgAcAHcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAEACQAQAA4AhAAAAAAAAAAAOgA6ZjoAADo6ADo6ZjpmkGY6OmaQtpBmOpC225Db/7aQZrbb/9u2Ztu2kNvbttv///+2Zv/bkP/btv/b2///2wECAwECAwECAwECAwECAwECAwECAwECAwVCICCO5AghTFlGSnA4qlgphLHEgNUQwyLFl0dBkPjhJoEernRpDItLko7ni45mtZtVxHLBtoBTCkwmUQiBtHqdFjBCADs=\" alt=\"image\"\u003eIMCS\u003csub\u003ethreshold\u003c/sub\u003e\u003c/p\u003e\n\u003cp\u003eFalse Alert (FA): IMCS\u003csub\u003epred\u003c/sub\u003e\u003cimg width=\"12\" height=\"19\" src=\"data:image/png;base64,R0lGODlhEgAcAHcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAEACQAQAA4AhAAAAAAAAAAAOgA6OjoAADo6ADo6ZjpmkDpmtmY6AGY6OmaQtma225Db/7ZmOrbb27bb/9u2Ztu2kNvbttvb/9v///+2Zv/b2///2wECAwECAwECAwECAwECAwECAwECAwU8oCUgFGCe6IlFxcBUaWxeDmE8sjwpY5mjq9YC9jtJEoLGLzgY5naCAyRHs+FkzNdPRCp6v15LYEwuk5MhADs=\" alt=\"image\"\u003eIMCS\u003csub\u003ethreshold\u0026nbsp;\u003c/sub\u003e\u0026amp; IMCS\u003csub\u003eobs\u003c/sub\u003e\u0026lt;IMCS\u003csub\u003ethreshold\u003c/sub\u003e\u003c/p\u003e\n\u003cp\u003eMissed Alert (MA): IMCS\u003csub\u003epred\u003c/sub\u003e\u0026lt;IMCS\u003csub\u003ethreshold\u0026nbsp;\u003c/sub\u003e\u0026amp; IMCS\u003csub\u003eobs\u003c/sub\u003e\u003cimg width=\"12\" height=\"19\" src=\"data:image/png;base64,R0lGODlhEgAcAHcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAEACQAQAA4AhAAAAAAAAAAAOgA6OjoAADo6ADo6ZjpmkDpmtmY6AGY6OmaQtma225Db/7ZmOrbb27bb/9u2Ztu2kNvbttvb/9v///+2Zv/b2///2wECAwECAwECAwECAwECAwECAwECAwU8oCUgFGCe6IlFxcBUaWxeDmE8sjwpY5mjq9YC9jtJEoLGLzgY5naCAyRHs+FkzNdPRCp6v15LYEwuk5MhADs=\" alt=\"image\"\u003eIMCS\u003csub\u003ethreshold\u003c/sub\u003e.\u003c/p\u003e\n\u003cp\u003ewhere IMCS\u003csub\u003epred\u0026nbsp;\u003c/sub\u003eis predicted by the system and IMCS\u003csub\u003ethreshold\u0026nbsp;\u003c/sub\u003eis a threshold value which potential users can set.\u003c/p\u003e\n\u003cp\u003eFinally, we evaluated the impacted area around the station where the predicted PGV varies within \u0026plusmn;50% of its value, called PGV\u003csub\u003epred150\u003c/sub\u003e and PGV\u003csub\u003epred50\u003c/sub\u003e respectively. Indeed, combining the magnitude estimate and the predicted PGV at the site with the GMPEs in the area (Scala et al 2025), we can evaluate Rmin and Rmax as it follows:\u003c/p\u003e\n\u003cp\u003e\u003cimg width=\"252\" height=\"31\" src=\"data:image/png;base64,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\" alt=\"image\"\u003e\u003c/p\u003e\n\u003cp\u003e\u003cimg width=\"239\" height=\"31\" src=\"data:image/png;base64,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\" alt=\"image\"\u003e\u003c/p\u003e\n\u003cp\u003eWhere A\u0026rsquo;\u0026rsquo;, B\u0026rsquo;\u0026rsquo;, C\u0026rsquo;\u0026rsquo; are the coefficients of the GMPEs of the area of interest. The difference \u0026Delta;R (R\u003csub\u003emax\u003c/sub\u003e-R\u003csub\u003emin\u003c/sub\u003e) represents the radius of the impacted area around the station. In Figure S3 we showed expected values for \u0026Delta;R in the Campi Flegrei area at different ranges of magnitude and PGV.\u0026nbsp;\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eV.L. Conceptualization; Formal analysis; Investigation; Methodology; Writing \u0026ndash; original draft Writing \u0026ndash; review \u0026amp; editingS.C. Conceptualization; Funding acquisition; Investigation; Methodology; Validation; Writing \u0026ndash; original draft Writing \u0026ndash; review \u0026amp; editingA.Z. Conceptualization; Funding acquisition; Investigation; Methodology; Project administration; Resources; Supervision; Validation; Writing \u0026ndash; original draft Writing \u0026ndash; review \u0026amp; editing\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAllen, R. M. \u0026amp; Melgar, D. Earthquake Early Warning: Advances, Scientic Challenges, and Societal Needs. (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1146/annurev-earth-053018\u003c/span\u003e\u003cspan address=\"10.1146/annurev-earth-053018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDel Gaudio, C., Aquino, I., Ricciardi, G. P., Ricco, C. \u0026amp; Scandone, R. Unrest episodes at Campi Flegrei: A reconstruction of vertical ground movements during 1905\u0026ndash;2009. \u003cem\u003eJ. Volcanol. Geoth. Res.\u003c/em\u003e \u003cb\u003e195\u003c/b\u003e, 48\u0026ndash;56 (2010).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eScarpa, R. et al. in 257\u0026ndash;282 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/978-3-642-37060-1_10\u003c/span\u003e\u003cspan address=\"10.1007/978-3-642-37060-1_10\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBevilacqua, A. et al. Data analysis of the unsteadily accelerating GPS and seismic records at Campi Flegrei caldera from 2000 to 2020. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e, 19175 (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDe Landro, G. et al. 3D structure and dynamics of Campi Flegrei enhance multi-hazard assessment. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e, 4814 (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNadeau, B. \u0026lsquo;Be prepared for all outcomes\u0026rsquo;: Inside the saga of a supervolcano that\u0026rsquo;s waking up. (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eIervolino, I. et al. Seismic risk mitigation at Campi Flegrei in volcanic unrest. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 10474 (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKuyuk, H. S. \u0026amp; Allen, R. M. Optimal Seismic Network Density for Earthquake Early Warning: A Case Study from California. \u003cem\u003eSeismol. Res. Lett.\u003c/em\u003e \u003cb\u003e84\u003c/b\u003e, 946\u0026ndash;954 (2013).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZollo, A. et al. in 117\u0026ndash;137 (2014). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/978-3-642-12233-0_7\u003c/span\u003e\u003cspan address=\"10.1007/978-3-642-12233-0_7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMeier, M. A. et al. How Often Can Earthquake Early Warning Systems Alert Sites With High-Intensity Ground Motion? \u003cem\u003eJ Geophys. Res. Solid Earth\u003c/em\u003e \u003cb\u003e125\u003c/b\u003e, (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSatriano, C. et al. PRESTo, the earthquake early warning system for Southern Italy: Concepts, capabilities and future perspectives. \u003cem\u003eSoil Dyn. Earthq. Eng.\u003c/em\u003e \u003cb\u003e31\u003c/b\u003e, 137\u0026ndash;153 (2011).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLux, A. I. et al. Status and Performance of the ShakeAlert Earthquake Early Warning System: 2019\u0026ndash;2023. \u003cem\u003eBull. Seismol. Soc. Am.\u003c/em\u003e \u003cb\u003e114\u003c/b\u003e, 3041\u0026ndash;3062 (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTun\u0026ccedil;, S. et al. Implementation of the EPIC Earthquake Early Warning System in the Bursa Province (T\u0026uuml;rkiye) and Its Surroundings. \u003cem\u003eAppl. Sci.\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e, 4985 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eIonescu, C., Marmureanu, A. \u0026amp; Marmureanu, G. in 441\u0026ndash;449 (2016). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/978-3-319-29844-3_31\u003c/span\u003e\u003cspan address=\"10.1007/978-3-319-29844-3_31\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKodera, Y. et al. The Propagation of Local Undamped Motion (PLUM) Method: A Simple and Robust Seismic Wavefield Estimation Approach for Earthquake Early Warning. \u003cem\u003eBull. Seismol. Soc. Am.\u003c/em\u003e \u003cb\u003e108\u003c/b\u003e, 983\u0026ndash;1003 (2018).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZollo, A., Colombelli, S., Caruso, A. \u0026amp; Elia, L. An Evolutionary Shaking-Forecast-Based Earthquake Early Warning Method. \u003cem\u003eEarth Space Science\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e, (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCaruso, A., Colombelli, S., Elia, L., Picozzi, M. \u0026amp; Zollo, A. An on-site alert level early warning system for Italy. \u003cem\u003eJ. Geophys. Res. Solid Earth\u003c/em\u003e. \u003cb\u003e122\u003c/b\u003e, 2106\u0026ndash;2118 (2017).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eColombelli, S., Caruso, A., Zollo, A., Festa, G. \u0026amp; Kanamori, H. A \u003cem\u003eP\u003c/em\u003e wave-based, on‐site method for earthquake early warning. \u003cem\u003eGeophys. Res. Lett.\u003c/em\u003e \u003cb\u003e42\u003c/b\u003e, 1390\u0026ndash;1398 (2015).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZollo, A., Nazeri, S. \u0026amp; Colombelli, S. Earthquake Seismic Moment, Rupture Radius, and Stress Drop from P-Wave Displacement Amplitude Versus Time Curves. \u003cem\u003eIEEE Trans. Geoscience Remote Sensing\u003c/em\u003e \u003cb\u003e60\u003c/b\u003e, (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFaenza, L. \u0026amp; Michelini, A. Regression analysis of MCS intensity and ground motion parameters in Italy and its application in ShakeMap. \u003cem\u003eGeophys. J. Int.\u003c/em\u003e \u003cb\u003e180\u003c/b\u003e, 1138\u0026ndash;1152 (2010).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eScala, A. et al. Ground Motion Prediction Equations for Campi Flegrei (Italy). (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.21203/rs.3.rs-5758860/v1\u003c/span\u003e\u003cspan address=\"10.21203/rs.3.rs-5758860/v1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRea, R., Colombelli, S., Elia, L. \u0026amp; Zollo, A. Retrospective performance analysis of a ground shaking early warning system for the 2023 Turkey\u0026ndash;Syria earthquake. \u003cem\u003eCommun. Earth Environ.\u003c/em\u003e \u003cb\u003e5\u003c/b\u003e, 332 (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eColombelli, S., Carotenuto, F., Elia, L. \u0026amp; Zollo, A. Design and implementation of a mobile device app for network-based earthquake early warning systems (EEWSs): application to the PRESTo EEWS in southern Italy. \u003cem\u003eNat. Hazards Earth Syst. Sci.\u003c/em\u003e \u003cb\u003e20\u003c/b\u003e, 921\u0026ndash;931 (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAcampora, G., Carotenuto, F. \u0026amp; Zollo, A. A. Proposal for a Fuzzy-based Architecture for Rescue Operations in Earthquake Scenarios. \u003cem\u003eIEEE International Conference on Fuzzy Systems (FUZZ) (FUZZ-IEEE 2025)\u003c/em\u003e 6 Preprint at (2025) 6 Preprint at (2025) (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSutton, J., Wood, M. M., Crouch, S. \u0026amp; Waugh, N. Public perceptions of U.S. earthquake early warning post-alert messages: Findings from focus groups and interviews. \u003cem\u003eInt. J. Disaster Risk Reduct.\u003c/em\u003e \u003cb\u003e84\u003c/b\u003e, 103488 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFallou, L., Corradini, M., Bossu, R. \u0026amp; Cheny, J. M. Preventing and debunking earthquake misinformation: Insights into EMSC\u0026rsquo;s practices. \u003cem\u003eFront Commun. (Lausanne)\u003c/em\u003e \u003cb\u003e7\u003c/b\u003e, (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003edi Scotto, F. et al. Delineation and Fine-Scale Structure of Fault Zones Activated During the 2014\u0026ndash;2024 Unrest at the Campi Flegrei Caldera (Southern Italy) From High‐Precision Earthquake Locations. \u003cem\u003eGeophys Res. Lett\u003c/em\u003e \u003cb\u003e51\u003c/b\u003e, (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKanamori, H. Real-time seismology and earthquake damage mitigation. \u003cem\u003eAnnu. Rev. Earth Planet. Sci.\u003c/em\u003e \u003cb\u003e33\u003c/b\u003e, 195\u0026ndash;214 (2005).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZollo, A., Lancieri, M. \u0026amp; Nielsen, S. Earthquake magnitude estimation from peak amplitudes of very early seismic signals on strong motion records. \u003cem\u003eGeophys Res. Lett\u003c/em\u003e \u003cb\u003e33\u003c/b\u003e, (2006).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAki, K. \u0026amp; Richards, P. G. \u003cem\u003eQuantitative seismology\u003c/em\u003e (University Science Books, 2009).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAllen, R. M. \u0026amp; Kanamori, H. The Potential for Earthquake Early Warning in Southern California. \u003cem\u003eSci. (1979)\u003c/em\u003e. \u003cb\u003e300\u003c/b\u003e, 786\u0026ndash;789 (2003).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLior, I., Ziv, A. \u0026amp; Madariaga, R. P. -Wave Attenuation with Implications for Earthquake Early Warning. \u003cem\u003eBull. Seismol. Soc. Am.\u003c/em\u003e \u003cb\u003e106\u003c/b\u003e, 13\u0026ndash;22 (2016).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-7619275/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7619275/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eRapid dissemination of earthquake information during seismic crises is crucial for mitigating the impact of earthquakes, particularly in densely populated volcanic areas, such as the Campi Flegrei caldera (Southern Italy). This region is affected by recurrent bradyseism and shallow seismicity, with recent events up to Md 4.6, very well perceived by the local inhabitants and increasing concern among the local community. In such small-extent areas, conventional Earthquake Early Warning (EEW) systems face physical limitations, due to the short duration and relatively small magnitude of earthquakes. Here we propose and calibrate a hybrid, impact-based, on-site EEW system capable of providing rapid estimates of the earthquake size and expected impact (PGV, PGA), within one second after the P-wave detection. In addition, the proposed system extends the classical concept of an on-site EEW system, introducing the idea of an \u0026ldquo;area of competence\u0026rdquo; around each seismic station that can benefit from the local warning.\u003c/p\u003e\u003cp\u003eThe proposed methodology is easily transferable to other volcanic or seismically active regions and might represent a step toward low-latency, impact-based earthquake alert systems designed to enhance community resilience.\u003c/p\u003e","manuscriptTitle":"One-Second-Lead Earthquake Warning and Impact Assessment at Campi Flegrei","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-16 07:07:53","doi":"10.21203/rs.3.rs-7619275/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-09-17T09:35:51+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-09-17T09:28:57+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-09-17T01:46:47+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-09-15T10:01:14+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"17d1119e-4018-411e-95d1-f824fbd63e39","owner":[],"postedDate":"September 16th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":54726349,"name":"Earth and environmental sciences/Natural hazards"},{"id":54726350,"name":"Earth and environmental sciences/Solid earth sciences"}],"tags":[],"updatedAt":"2026-04-13T16:01:20+00:00","versionOfRecord":{"articleIdentity":"rs-7619275","link":"https://doi.org/10.1038/s41598-026-42593-x","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2026-04-08 15:58:24","publishedOnDateReadable":"April 8th, 2026"},"versionCreatedAt":"2025-09-16 07:07:53","video":"","vorDoi":"10.1038/s41598-026-42593-x","vorDoiUrl":"https://doi.org/10.1038/s41598-026-42593-x","workflowStages":[]},"version":"v1","identity":"rs-7619275","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7619275","identity":"rs-7619275","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}