Advancing meteotsunami prediction: testing high-resolution modeling in harbor environments

Advancing meteotsunami prediction: testing high-resolution modeling in harbor environments | 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 Advancing meteotsunami prediction: testing high-resolution modeling in harbor environments Laura Nesteckytė, Vasily Titov, Loreta Kelpšaitė-Rimkienė, Jadranka Šepić This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6861009/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 04 Feb, 2026 Read the published version in Natural Hazards → Version 1 posted 4 You are reading this latest preprint version Abstract Meteotsunamis, often overlooked in global disaster discussions, pose significant threats to coastal communities and infrastructure, particularly in shallow coastlines, narrow bays, and harbors. Unlike seismic tsunamis, where source parameters can typically be determined quickly following an earthquake, forecasting meteotsunamis remains challenging due to difficulties in obtaining real-time high-resolution atmospheric pressure data that contain meteotsunami source parameters. Intense short-period atmospheric pressure disturbances are typical triggers of meteotsunamis, but these small-scale pressure anomalies are often missed in observations and numerical weather forecasts. This study explores the feasibility of using weather radar reflectivity as a proxy for atmospheric pressure anomalies to enhance meteotsunami forecasting capabilities. The study employs the Method of Splitting Tsunamis (MOST) model, commonly used for seismic tsunami forecasting, across three very different regions: the East Coast of the United States, the southeastern Baltic Sea, and the Adriatic Sea (the Mediterranean). The results demonstrate that radar-based modeling successfully captures meteotsunami generation and propagation dynamics, particularly for summer events driven by mesoscale convective systems. The findings suggest that real-time meteotsunami forecasting is achievable using weather radar inputs, providing a promising approach for coastal hazard mitigation and early warning systems. However, winter meteotsunamis which are often related to extratropical cyclones and wind-driven forces, and which are superimposed on ongoing storm surges require additional modeling refinements. Meteotsunami Numerical model Forecast Real-time data Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 1. Introduction The focus of global discussions on tsunami forecast and warning has expanded from the hazards of seismic tsunamis to include the risks posed by meteorologically-generated tsunamis (meteotsunamis) (Monserrat et al. 2006 ; Pattiaratchi and Wijeratne 2015 ; Pellikka 2020 ; Titov and Moore 2021 ; Wijeratne and Pattiaratchi 2024 ; Nesteckytė et al. 2024 ; Angove et al., 2025 ). Much like seismically-generated tsunamis, meteotsunamis pose a particular threat to populations and infrastructure located along shallow coastlines, in narrow bays, and in harbors (Monserrat et al. 2006 ; Rabinovich 2020 ). Seismic tsunami forecasting has advanced significantly, with the establishment of warning centers such as the National Tsunami Warning Center and the development of modeling platforms like SIFT, Tweb and ComMIT (Titov et al. 2011 ; Pattiaratchi and Wijeratne 2015 ), enabling more reliable predictions of tsunamis, once a tsunamigenic earthquake has occurred. In contrast, forecasting meteorological tsunamis and issuing reliable warnings remains a challenge. Meteotsunamis are a special group of tsunamis. They are generated by propagating atmospheric pressure (and wind) disturbances of limited spatial dimensions (tens of kilometers) and relatively short duration (tens of minutes to couple of hours) (Monserrat et al. 2006 ; Vilibić et al. 2016 ). To trigger a destructive meteotsunami, atmospheric disturbance needs to be characterized by a high rate of air pressure change (a couple of hPa over a few minutes), it must propagate over a shelf of slowly varying depth with a speed that is equal (or almost equal) to speed c of long ocean waves at that shelf ( \(\:c=\sqrt{gh}\) , where g is gravity acceleration, and h is ocean depth) or along a gently sloping bathymetry with a speed that is equal to speed of trapped edge waves. This equality of speeds is a prerequisite for resonant transfer of energy from atmospheric pressure disturbances to long ocean waves, i.e. for Proudman ( 1929 ) and Greenspan resonances (Greenspan 1956 ). Even a small offset of atmospheric disturbance from the “ideal” tsunamigenic properties (exact track, value of rate of air pressure change, speed) can result with an order of magnitude different height of a meteotsunami (e.g., Šepić et al. 2016 ). To track and forecast such tsunamigenic atmospheric disturbances we need to use high-resolution atmospheric pressure measurements and models – with high resolution in spatial and temporal dimensions - that can give the location, speed, amplitude, and propagation direction of the pressure disturbances that generate meteotsunamis (Pattiaratchi and Wijeratne 2015 ; Ha et al. 2018 ; Titov and Moore 2021 ; Villalonga et al. 2024 ; Wijeratne and Pattiaratchi 2024 ). However, even the most sophisticated weather models fail to capture exact properties of tsunamigenic atmospheric pressure disturbance (Denamiel et al. 2019a ; Mourre et al. 2021 ), which begs the question: Is real-time forecasting of meteotsunamis possible? One alternative is to use measurements of atmospheric pressure, which are assessed in real time. However, these measurements are often of low spatial resolution (distances between stations are of order of tens of kilometers, i.e., of order of spatial dimensions of atmospheric disturbances) and, thus, do not allow for capturing all relevant properties of small-scale short-lasting atmospheric pressure disturbances (Titov et al. 2011 ; Šepić et al. 2016 ). This study explores an alternative method for meteotsunami forecasting. Our method combines real-time measurements and numerical ocean modelling: we use precipitation reflection images obtained from weather radars to derive data which can be used to force a numerical ocean model. The radar reflectivity images allow us to simulate atmospheric pressure disturbance as a direct, time-dependent input for numerical ocean meteotsunami generation and propagation model. The governing idea is that higher radar reflectivity intensities correspond to relatively higher atmospheric pressure anomalies. While the absolute pressure level is unknown from the radar images, the pressure can be scaled by atmospheric pressure measurements available at the meteorological station near the prediction area (Titov and Moore 2021 ). Doppler radar data are widely available worldwide, from radars deployed for various purposes, particularly meteorology, aviation, and defence. A weather radar works by emitting pulses of electromagnetic energy at microwave frequencies into the atmosphere. Upon encountering various particles in the air, such as water and ice droplets, some of this electromagnetic energy is scattered back toward the radar—this phenomenon is commonly termed "reflection," which forms the basis of "reflectivity" (The University Corporation for Atmospheric Research, 2012 ). Reflectivity quantifies the efficiency of a radar target in intercepting and returning radar energy and is influenced by specific physical characteristics of the target, such as size, shape, orientation, and composition. The returned signal is processed by computer systems, enabling the assessment of precipitation location and intensity, along with wind speed and direction. Doppler radars deliver localized, high-resolution, timely, and three-dimensional observational capabilities that surpass those provided by other meteorological monitoring tools. They can detect changes in precipitation rates with a spatial resolution of a few square kilometers or finer and operate on temporal cycles of just a few minutes. This advanced capacity allows for the monitoring of rapidly evolving weather phenomena, which is crucial for issuing timely warnings regarding severe and hazardous weather conditions, including heavy rainfall, hail, strong winds (such as those associated with tornadoes and tropical and extratropical cyclones), and wind shear (National Weather Service). Furthermore, recent advancements in numerical weather prediction systems have facilitated the integration of continental-scale, radar-derived precipitation data into global models, thereby enhancing the accuracy of 4 - to 5-day precipitation forecasts for surrounding regions and continent (European Centre for Medium-Range Weather Forecasts). If radar data improves short-term weather forecasts, maybe it can also be used for forecasting meteotsunamis. Globally, networks of Doppler radars are operated by national meteorological agencies, such as the National Weather Service in the United States, Environment and Climate Change Canada, and various national meteorological services across Europe and around the Globe. In Europe, the EUMETNET OPERA programme supports the harmonization and exchange of radar data between national services, although it does not operate radar systems itself. These expansive networks ensure comprehensive coverage, providing timely and precise data essential for public safety and effective disaster management. The global standardization of weather radar allows seamless comparison of radar reflectivity data from different regions of the world, eliminating the need for an extensive and complicated data acquisition process. Consequently, a weather radar-based meteotsunami prediction model has the potential for widespread application. The advantage of using radar images for rapid meteotsunami source assessment lies in three key factors: (1) the data are available in near-real time; (2) the image sequence provides direct measurements of the speed, direction, and potential area of the pressure anomaly’s movement; (3) the spatial and temporal resolution of the radar imaging is sufficient to be used directly in tsunami generation and propagation models, eliminating the need for complex interpolation and interpretation — unlike sparse atmospheric pressure or sea-level measurements. It's important to note that real-time radar data is especially useful for tracking atmospheric disturbances that can generate tsunamis, particularly when those disturbances are part of convective systems rich in hydrometeors. This study examines the feasibility of broad geographic applicability of this method, focusing on three areas with different geographies, different meteotsunami generation mechanisms and different meteotsunami characteristics: (1) the East Coast of the United States (2) the Baltic Sea; and (3) the Adriatic Sea (the semi enclosed sea at the north of the Mediterranean Sea) (Fig. 1 ). Meteotsunamis are often classified as winter and summer events (Pattiaratchi and Wijeratne 2015 ; Dusek et al. 2019 ; Pellikka et al. 2020 , 2022 ; Nesteckytė et al. 2023 ; Wijeratne and Pattiaratchi 2024 ). On the East Coast of the U.S., wintertime meteotsunamis are associated with large storm systems that impact broad areas and are often accompanied by other hazardous oceanographic conditions, such as storm surges and large waves (Dusek et al. 2019 ). The three strongest known East Coast meteotsunamis are the Daytona Beach meteotsunami of 4 July 1992 (associated to passage of squall line (Churchill et al. 1995 ; Sallenger et al. 1995 ), the Boohtbay meteotsunami of 28 October 2008 (associated to cold front and atmospheric gravity waves (Vilibić et al. 2014 ; Whitmore and Knight 2014 ), and the New Jersey meteotsunami of 13 June 2013 (associated with derecho (Bailey et al. 2014 ; Wertman et al. 2014 ). Winter meteotsunamis in the Baltic Sea primarily occur between November and April (Pellikka et al. 2022 ; Nesteckytė et al. 2023 ). These events are closely linked to cold front passages and deep, large-scale extratropical cyclones moving from the west-northwest-north (Pellikka et al. 2022 ). The extratropical cyclones are characterized by strong surface winds, but also by significant smaller-scale short-period atmospheric pressure changes, creating the conditions necessary for meteotsunami generation. The dynamics of winter meteotsunamis reflect the harsh and energetic weather systems prevalent during the colder months. Summer meteotsunamis in the Baltic Sea are most common between late April and October and are linked to mesoscale convective systems (MCSs). Such systems frequently result in the occurrence of thunderstorms, rapid changes in atmospheric pressure, and, in certain instances, the formation of clear frontal structures, including gust fronts, squall lines, and hail. MCSs are most prevalent in July and August, with infrequent occurrences of intense MCSs during the winter months. Such events may be as severe as those classified as U.S. continental derechos (Punkka and Bister 2015 ; Pellikka et al. 2022 ). In the Adriatic Sea, meteotsunamis can also be classified into wintertime and summertime events (Ruić et al. 2023 ; 2025; Šepić et al. 2024). The Adriatic Sea wintertime meteotsunamis are associated with strong extratropical cyclones and appear simultaneously to storm surges, the summertime meteotsunamis are, on the other hand, usually associated with atmospheric gravity waves which usually appear jointly with convective storms or mesoscale convective systems (e.g., Belušić and Strelec Mahović 2007 ; Šepić et al. 2009 ). Conclusively, for all locations, summer meteotsunamis are typically driven by localized MCSs and rapid air pressure jumps which are very difficult to forecast precisely. And winter events are generally associated with larger-scale synoptic patterns, i.e., with extratropical cyclones, and occasionally tropical cyclones (only at the East Coast of the U.S.). Current synoptic models, such as those developed by the European Centre for Medium-Range Weather Forecasts (ECMWF) and the Global Forecast System (GFS), demonstrate robust capabilities in forecasting extratropical cyclones at mid-latitudes. This capability provides valuable context that helps identify conditions favorable for winter meteotsunami events. However, these large-scale models alone cannot resolve the small-scale atmospheric pressure disturbances embedded within cyclones, which directly generate meteotsunamis. Therefore, while accurate extratropical cyclone predictions improve meteotsunami forecasting potential, they do not entirely solve the prediction challenge. Furthermore, when dealing with rapidly evolving convective systems, such as thunderstorms, squall lines, or even derecho events, which are notoriously difficult to predict—even with advanced models like ECMWF or GFS — it becomes essential to have a real-time tool capable of assessing meteotsunami threats based on up-to-date meteorological data, including weather radar inputs. This work tested the hypothesis of using weather radar reflectivity as a proxy for atmospheric pressure at three different locations around the world, the East Coast of the United States, the southeast coast of the Baltic Sea, and the Adriatic Sea (Fig. 1 , 3 ). 2. Materials and methods 2.1 Study sites The East Coast of the United States (Fig. 1 ) experiences an average of 25 meteotsunamis per year, according to Dusek et al. ( 2019 ). The statistics was developed by extracting meteotsunami signals from 22 years of sea level data using a station-dependent threshold-surpassing criteria (meteotsunamis were defined as those events during which energy of high-frequency sea level oscillations and their maximum peak-to-trough height surpassed pre-described thresholds) superimposed on a meteorological criteria (necessity of an atmospheric pressure disturbance). Depending on the station, events with peak-to-trough height as low as 0.2 m were classified as meteotsunamis. Nonetheless, some of these events exceeded 0.6 meters in height, making them capable of causing significant damage and posing risks to coastal communities Furthermore, even stronger meteotsunamis have been observed (but not measured in their full strength) along the East Coast of the U.S., e.g., the Daytona Beach meteotsunami with estimated height of 3–6 m (Churchill et al. 1995 ; Sallenger et al. 1995 ), and Boothbay meteotsunami with estimated height of 3.6 m (Whitmore and Knight 2014 ). The East Coast regularly experiences weather systems capable of producing meteotsunamis, including squall lines, thunderstorms, frontal systems, and derechos, a widespread, long-lived windstorm associated with a band of rapidly-moving showers or thunderstorms (Dusek et al. 2019 ). These weather systems produce sudden changes in atmospheric pressure and strong wind gusts that transfer energy to the ocean surface. In addition, many bays, inlets, and harbors along the coast have natural resonance periods that align with the periods of propagating atmospheric disturbances, amplifying the resulting waves and increasing the likelihood of strong meteotsunamis. Over the East Coast of the U. S., meteotsunamis are generated by offshore (eastward) propagating atmospheric disturbances (Pasquet and Vilibić 2013 ; Titov and Moore 2021 ). Atmospheric disturbances generally come to the East Coast from the west (i.e., from the continental U.S.), they cross the coast and advance towards the east over the western North Atlantic shelf. While propagating over the shelf, atmospheric disturbances can enhance long ocean waves with Proudman resonance, given that velocity of the atmospheric disturbance matches velocity of long ocean waves (Proudman 1929 ). Atmospheric disturbance and forced long ocean wave travel coupled as long as their velocities match. However, around 135 km from the coast, the coupled system reaches the shelf break. Due to a sudden change of ocean depth, long ocean waves speed abruptly changes, so part of the ocean wave reflects back from the shelf break towards the East Coast of the U.S., while atmospheric disturbance continues eastward. From this point on, long ocean waves propagate as free gravity-driven long waves – impacting the coast 4 to 6 hours after atmospheric disturbance has initially crossed it (Titov and Moore 2021 ). In the Baltic Sea (Fig. 1 ), there are historical accounts of very strong Baltic meteotsunamis (e.g., Renquist 1926 ; Piotrowski et al. 2017 ). These events were mostly observed over the northward facing Polish and German coasts. Atmospheric disturbances propagating towards these coasts (from the north to the south) pass over seas of relatively uniform depths – thus generating long ocean waves via Proudman resonance (Rauhala et al. 2014 ). The port of Klaipėda (our study location), however, faces the west. Hence, any atmospheric disturbance propagating towards the site would pass over the sea with significantly variable depth making the Proudman resonance unlikely, which appears to suggest low threat of strong meteotsunamis. Nonetheless, even weak meteotsunamis, can seriously affect the sea-going traffic and result in closure of the port (see Nesteckytė et al. 2024 ). The Port of Klaipėda is located in the Klaipėda Strait (Fig. 2 e). The Klaipėda strait serves as an important connection between the southern Baltic Sea and the largest lagoon in Europe, Curonian Lagoon. The port's quays are located within a 10 km long and 0,36–1,20 km wide inlet. Long ocean waves are usually intensified in the Port of Klaipėda due to its complex narrow and deep topography (Kelpšaitė-Rimkienė et al. 2018 ; (Kelpšaitė-Rimkienė et al. 2018 ; Rabinovich et al. 2021 ; Nesteckytė et al. 2023 , 2024 ) The Adriatic Sea (Fig. 1 ) is considered to be a meteotsunami hot spot (Šepić et al. 2015 ). According to Šepić and Orlić (2024), within the last ~ 90 years 36 meteotsunamis with crest-to-trough wave heights above 1 m, and 13 with crest-to-trough wave heights above 2 m were observed, including one of the strongest known World meteotsunamis, the Great Vela Luka flood of 21 June 1978 – an event when height of sea level oscillations reached astonishing 6 m (Orlić 1980 ; Orlić et al. 2010 ). In the Adriatic Sea, meteotsunamis are generated by atmospheric pressure disturbances propagating northeastward to eastward and directly towards the endangered coast (e.g., (Šepić et al. 2016 ). The Adriatic meteotsunamis reach particularly large wave heights within bays characterized by high amplification (Q) factors (Rabinovich 2009 ) such as Mali Lošinj, Široka Bay, Stari Grad and Vela Luka (Orlić 2015 ). These are funnel-like bays which narrow and become shallow at the closed end. Nowadays, tide gauges are located at two meteotsunami-prone locations in the Adriatic Sea, i.e., in Mali Lošinj and Stari Grad Bays, and events hitting these two locations will be studied in more detail in this paper. Mali Lošinj Bay (Fig. 2 f), a natural harbor situated in a deep, narrow bay on the island of Lošinj, provides excellent shelter from winds and rough seas. With multiple piers and quays, it accommodates fishing boats, yachts, and ferries. Similarly, Stari Grad Bay (Fig. 2 g), located on the northern coast of Hvar Island, is a long inlet whose natural configuration offers protection from most winds, making it a safe anchorage for boats and small ships. However, despite their well-sheltered appearances, both bays possess geographic features that favor amplification of incoming long ocean waves, energetic seiche oscillations and strong meteotsunami occurrences. Mali Lošinj has a relatively narrow but deep entrance, while Stari Grad extends inland in a fjord-like formation. These deep and narrow channels create ideal conditions for harbor resonance (Vilibić 2008 ; Šepić et al. 2009 ; Wijeratne and Pattiaratchi 2024 ), one of a key factor in meteotsunami enhancement. Meteotsunamis with wave heights > 3 m have been observed at both these locations (Vilibić et al. 2004 ; Belušić and Strelec Mahović 2007 ; Šepić and Orlić, 2024). According to the online version of the Adriatic meteotsunami catalogue (Šepić and Orlić 2024), Mali Lošinj has experienced two significant meteotsunamis, one of which reached a runup of 3.5 meters, causing damage to homes, vessels, and shops. In Stari Grad, seven meteotsunamis with crest-to-trough wave-heights above 1 m have been documented since the 1930s (including two with heights of 2.5 and 3.5, respectively), resulting in damage to houses, businesses, and infrastructure. 2.2 Meteotsunami modelling approach For meteotsunami simulations, the numerical model based on the Method of Splitting Tsunamis (MOST) formulation (Titov et al. 2011 , 2016 ) was employed. This model, widely recognized for its application in seismic tsunami modeling, has been extensively validated through laboratory experiments and historical tsunami simulations. Since 2012, NOAA has operationally used the MOST model for tsunami forecasting (Titov et al. 2016 ). Its adaptability extends to simulating time-dependent tsunami initiation mechanisms, such as those caused by landslides, or atmospheric pressure forces (with the latter relating to meteotsunami generation processes) (Titov and González 2000 ; Titov and Moore 2021 ; Boslough and Titov 2024 ). The MOST model incorporates atmospheric pressure forcing terms in Cartesian coordinates as follows: $$\:\frac{\partial\:u}{\partial\:t}+u\frac{\partial\:u}{\partial\:x}+v\frac{\partial\:u}{\partial\:y}+g\frac{\partial\:h}{\partial\:x}=g\frac{\partial\:d}{\partial\:x}-\frac{1}{\rho\:}\frac{\partial\:P}{\partial\:x};\frac{\partial\:v}{\partial\:t}+u\frac{\partial\:v}{\partial\:x}+v\frac{\partial\:v}{\partial\:y}+g\frac{\partial\:h}{\partial\:y}=g\frac{\partial\:d}{\partial\:y}-\frac{1}{\rho\:}\frac{\partial\:P}{\partial\:y};\frac{\partial\:h}{\partial\:t}+\frac{\partial\:\left(hu\right)}{\partial\:x}+\frac{\partial\:\left(hv\right)}{\partial\:y}=0;$$ 1 here \(\:h\left(x,y,t\right)=\zeta\:\left(x,y,t\right)+d\left(x,y\right),\:\zeta\:\left(x,y,t\right)\) is the amplitude, \(\:d\left(x,y\right)\) is the water depth, \(\:u\left(x,y,t\right)\) and \(\:v(x,y,t)\) are velocities in the \(\:x\) and \(\:y\) directions accordingly, \(\:P\) is atmospheric pressure, \(\:\rho\:\) is water density, \(\:g\) is the acceleration of gravity. The MOST numerically solves these shallow-water wave equations in spherical coordinates, incorporating Earth’s curvature, as detailed by Titov et al. ( 2016 ). The numerical approach we use in this study assumes several simplifications in the meteotsunami initiation process. When the input of atmospheric forcing in the model ends, the water surface disturbances created by atmospheric forcing continues propagating as free waves. This simplification can result in additional waves appearing behind the main meteotsunami wave as it reaches the coast. However, this approach represents the simplest real-time assumption, necessary when forecasts must be made before observational data becomes available. It is assumed that these artefactual waves at the meteotsunami's tail are of lesser magnitude and significance compared to the initial forced waves, as they are generated after the meteotsunami has decoupled from atmospheric forcing. In this formulation, wind stress forcing is omitted, leaving atmospheric pressure changes—associated with the moving pressure field derived from radar images—as the sole mechanism for wave generation. Although, simulations with added wind stress can produce improved results in meteotsunami modelling (e.g. Bechle and Wu 2014 ), effect of wind stress remains secondary to effect of atmospheric pressure (e.g. Orlić et al. 2010 ; Šepić et al. 2015 ), and wind stress is commonly omitted when modelling meteotsunamis. The meteotsunami modeling uses multi-grid simulation approach. The regional propagation of the meteotsunami in the basin-wide large area is performed using coarser resolution grid with simplified reflective boundaries along the coastline. Absorbing boundaries at the edges of the domain were applied to enable wave energy to exit the computational area without reflection. The coastal impact for smaller areas inside the regional grid around the tide gauges is modeled with nested grid computation that zooms the resolution of the model from the propagation grid into the fine-resolution grid of about 10–50 meter spacing. The moving boundary conditions are applied for these nested grids to account for inundation on land. The coastal impact modeling that uses three nested grids was performed with the ComMIT tool. ComMIT is a user-friendly graphical tool that connects the MOST model with a pre-computed tsunami propagation scenario database, which allows for modeling of seismic tsunami impact at any coastline around the world. Beyond enhancing modeling accessibility, ComMIT fosters collaboration within the tsunami research community by facilitating results sharing. Its intuitive design allows for efficient visualization of modeling outputs and integration of initial conditions from precomputed propagation databases, all while maintaining minimal hardware requirements. Together, the MOST model and ComMIT provide a robust framework for simulating and understanding meteotsunami propagation in both research and operational contexts. 3. Results 3.2 Numerical modelling of meteotsunamis using radar reflectivity images The study sites have very different geography and bathymetry settings. Nevertheless, in all cases the propagation speed of the weather systems we modeled were sufficiently close to the long wave celerity over at least part of the propagation path to sustain Proudman resonance amplification. This resonance increased the meteotsunami amplitude significantly beyond the estimates provided by the static “inverted barometer” approach for all events. An analysis of the atmospheric front along the East Coast of the United States reveals that it travelled approximately 80 to 110 km within an hour (Fig. 1 , Fig. 2 a). This corresponds to propagation speeds of around 22 to 30 m/s, which align with the long wave celerity in water depths ranging from 50 to 90 meters. In the case of Klaipėda (Fig. 2 b and e), the thunderstorm cloud traveled at speeds of 6 to 17 m/s. The average depth of the Port of Klaipėda is approximately 15 meters, and the average depth of Curonian lagoon is 3.8 m, while the depths near to the Lithuanian coast, where Proudman resonance should have taken place for meteotsunamis to occur, range between 40 and 50 meters (corresponding to meteotsunami celerity of ~ 19–22 m/s). In Croatia the average depth around the island of Lošinj (Fig. 2 f) ranges between 40 and 50 meters (corresponding to meteotsunami celerity of ~ 19 to 22 m/s), while the waters surrounding Hvar are deeper (Fig. 2 g), averaging 60 to 70 meters (corresponding to meteotsunami celerity of ~ 24 to 26 m/s). During the Adriatic meteotsunami events, storm clouds have been recorded moving at speeds of 17 to 19 m/s on July 19, 2023, and 15 to 17 m/s on July 22, 2023, aligning with the long-wave phase speed in the respective water depths, thus sustaining Proudman resonance and amplifying the waves. 3.2.1 The East Coast meteotsunami of June 13, 2013 (U.S.) For the East Coast of the U.S. location, meteotsunami of June 13, 2013 (Fig. 3 a), (hitting Atlantic City at 15:02 UTC) was simulated. This modeling was the extension of the model described by Titov and Moore ( 2021 ). The goal of this modeling effort was to test the multi-grid forecast approach for meteotsunami simulations. While the propagation results followed the previously published procedure, the simulation was extended to simulate areas around tide-gages with much higher resolution model that includes the flooding forecast capability, similar to the NOAA tsunami forecast tools (Titov et al. 2016 ). On June 12–13, 2013, a powerful derecho developed from severe thunderstorms initially forming over Iowa and Illinois (Fig. 1 , IA and IL). The storms began as discrete supercells in a highly unstable atmospheric environment, characterized by high CAPE values ranging from approximately 2000 to 4800 J/kg (Hersbach et al. 2020 ). These conditions, combined with strong mid-level wind shear and an upper-level atmospheric disturbance interacting with the warm and moist air mass over the northern Great Plains, led to rapid storm intensification and organization. The individual supercells merged into a well-organized squall line as they entered Indiana (Fig. 1 , IN), where embedded downbursts produced damaging winds with speeds up to approximately 45 m/s, causing widespread structural and vegetation damage. The derecho maintained its strength overnight, benefiting from persistent atmospheric instability, favorable wind shear, and near-unidirectional mid-tropospheric flow. As the system reached Ohio (Fig. 1 , OH), it produced further extensive wind damage and spawned several embedded tornadoes. By early morning on June 13, as the derecho moved across the eastern states and onto the Atlantic coastline, the associated rapid pressure disturbances generated significant sea-level oscillations recorded by some coastal tide gauges along the U.S. East Coast (e.g., at Lewes, Fig. 4 ). The generation, propagation and coast impact of the meteotsunami were modeled along the segment of the coast from 35° N to 41.6° N and 77° W to 70.4° W (Fig. 1 ). The propagation model utilized bathymetry with a resolution of 9 arc seconds (approximately 270 meters) and, a time step of 1 second. The atmospheric pressure forcing was applied for the first 2.8 hours of the simulation using the sequence of air pressure fields estimated from radar reflectivity data. The radar images correspond to time when the pressure disturbance was propagating over the continental shelf. The propagation simulation was used as the forcing function for the high-resolution models setup around tide gauges at Montauk, Atlantic City and Lewes. The models used three nested grids that zoom into the highest spatial grid resolutions of 2 arc seconds (less than 60 meters), where moving boundary conditions are applied to allow for predicting tsunami inundation estimate. The air pressure forcing was applied throughout the simulation for all numerical grids, to ensure the consistent dynamic tsunami generation. In spite of the fact that a maximum high-frequency atmospheric pressure change of 4.7 hPa per 18 min was measured at the Atlantic City airport (Rabinovich 2020 ), no significant sea-level oscillations were measured at the Atlantic City tide gauge station at time of atmospheric disturbance passage (Fig. 4 ). Similar holds for Montauk (Fig. 4 ). The model reproduced the height and period of initial sea level oscillations satisfactory at the three East Coast stations (Fig. 4 ; time from 0 to 3 hours). Lewes is situated at the southern side of the entrance to Delaware Bay. Since Delaware Bay is long and relatively shallow, the pressure disturbance had traveled for around an hour along the coast before reaching Lewes. Thus, it had enough time to affect the sea and generate a long ocean wave, as the one observed and modelled in Lewes (Fig. 4 ). On the contrary, the atmospheric disturbance came to Atlantic City from the land, and was not able to generate ocean wave on its passage over the station. However, as already shown by Titov and Moore ( 2021 ) and confirmed by our modelling results, the atmospheric pressure front continued travelling offshore (towards the east) across the continental shelf, and then over deeper waters of the Atlantic Ocean. While it traveled over the shelf, it was coupled to a long ocean wave which grew stronger through Proudman resonance. Upon reaching the shelf break, meteotsunami waves decoupled from the atmospheric forcing (which continued to propagate eastward) and started propagating as free waves. Because long ocean wave speeds (celerity) differ significantly between shallow continental shelf waters and the adjacent deeper ocean, the meteotsunami waves mostly refracted and turned back toward the U.S. East coast. As a result, most of the meteotsunami energy remained concentrated over the wide and gently sloping U.S. East Coast shelf—a region well-known to trap and amplify tsunami energy (Mofjeld et al. 2001 ). It was this reflected wave that reached considerable heights along the U. S: East Coast, including at Atlantic City and Montauk, as is both seen in measurements, and successfully reproduced by our model (Fig. 4 ). The results of our high-resolution modeling of the June 13, 2013 meteotsunami showed a better comparison with observations in comparison with the coarse propagation model (see Supplemental Material). These model results confirm that the multi-grid approach to meteotsunami simulation improve forecast accuracy and provides method for predicting potential flooding from meteotsunamis. To test the method further, we applied the same approach for several other meteotsunami events at different geographic settings. 3.2.2 The Klaipeda meteotsunami of June 20, 2022, Lithuania The meteotsunami propagation throughout the SE Baltic Sea toward Klaipėda was modeled along the segment of the coast from 53° N to 59° N and 11° E to 31° E (Fig. 1 ), resolution was 13 arc seconds (approximately 400 meters), and time step was set to 0.3 seconds. We used 13 hours of the atmospheric pressure forcing, which means that radar reflectivity data, or direct atmospheric pressure forcing, was used throughout the simulation. Over 8 hours of meteotsunami dynamics was simulated following the first detection of cumulonimbus formation at 6:00 UTC over the western part of Lithuania (Fig. 3 b). The area around the Klaipėda port and the tide gauge was modeled with the multi-gird high-resolution inundation model. The highest spatial resolution of the model around the tide gauge location was 0.3 arc second (around 10 meters). The model was able to reproduce fairly small signal of June 20, 2022, meteotsunami event. The modeled event unfolded as follows: On 20 June 2020, high convective clouds began forming over western Lithuania and gradually moved northwest toward Klaipėda. As convection intensified, the first light rain shower was observed in Klaipėda at 07:50 UTC, accompanied by a weak atmospheric pressure disturbance and the onset of long ocean wave activity—both detected in measurements and replicated in the model (see Fig. 4 Klaipėda, ~ 2 hours into the simulation). Between 2.5 and 5 hours into the simulation, no significant atmospheric forcing was detected; however, sea level fluctuations persisted in the Port of Klaipėda, suggesting that seiches were likely triggered by the initial disturbance or incoming long waves. Around 10:50 UTC, intense cumulonimbus clouds reached Klaipėda, leading to renewed amplification of long ocean waves—again evident in both observations and model output (Fig. 4 Klaipėda, after 5 hours). The Klaipėda meteorological station recorded thunder and heavy rainfall, indicating strong convective activity. At the same time, the Convective Available Potential Energy (CAPE) in western Lithuania and Latvia reached 1400–1800 J/kg. For reference, CAPE values exceeding 1000 J/kg are typically sufficient to support the development of strong to severe storms (Pešice et al. 2003 ; Garner 2015 ; Punkka and Bister 2015 ). It can be observed that in the presence of stronger atmospheric disturbances, the model results are more accurate, successfully capturing both the period and the amplitudes of the oscillations, if not the exact shape (Fig. 4 ). 3.2.3 The Adriatic Sea meteotsunamis of July 20–22, 2023 The 19 July 2023 storm over the southeastern Europe was a textbook example of a highly organized convective system (Fig. 3 b), driven by a combination of strong atmospheric instability, mesoscale forcing, and upper-level dynamics. Its formation was associated with a pronounced upper-level trough moving eastward, providing the necessary dynamic lifting for deep convection (Mahovic et al. 2023 ). At the surface, a low-pressure system over northern Italy and the northern Adriatic Sea generated a strong pressure gradient, enhancing low-level convergence and moisture advection from the Mediterranean. During a period of extreme heat affecting southern Europe, several atmospheric factors contributed to the storm’s development. Strong upper-level westerly winds, with a jet streak positioned over Central Europe, brought cooler air aloft from northern Europe, creating steep lapse rates (Mahovic et al. 2023 ). At the same time, the low layers remained very humid, particularly around the Po Valley, creating an environment highly conducive to severe storm formation over multiple days in mid-July. One of the key drivers of the storm’s rapid intensification was high Convective Available Potential Energy (CAPE), exceeding 2000–5000 J/kg in some areas (ERA5 Reanalysis data Climate Data Store; Hersbach et al., 2020 ). This, combined with strong wind shear (with mid-level wind shear exceeding 20–25 m/s), facilitated the development of an intense mesoscale convective system (MCS) (Mahovic et al. 2023 ). The storm’s extreme wind speeds, localized downbursts, and sustained pressure fluctuations were direct consequences of deep convection and rapid vertical momentum transport. As the system moved eastward into Croatia, Slovenia, and Bosnia and Herzegovina, it caused widespread damage due to the combined effects of intense winds and heavy precipitation. However, this initial storm was not the direct trigger of the meteotsunamis. Instead, it was the second storm, which developed at night between 19 and 20 July, that played a crucial role in meteotsunami generation. Unlike the first one, this storm system moved not only over the land but also over the Adriatic Sea, propagating from the northwest to the southeast, along the Adriatic Sea main axis, and generating long ocean waves through Proudman resonance. Measured sea level oscillations at Mali Lošinj reached amplitudes of 67.75 cm. Three days later, on July 22, CAPE was again high over the Adriatic Sea (1500–3000 J/kg) (ERA5 Reanalysis data Climate Data Store; Hersbach et al., 2020 ). A similar weather pattern was unfolding, with the Mediterranean region experiencing a prolonged heatwave, a moisture-laden atmosphere, and a developing cyclonic cold front over the northern Italy, Switzerland, and Austria. As this system moved eastward, it triggered the formation of convective clouds. By noon of July 22, the first rain clouds emerged over the northern Adriatic, and by 12:20 UTC, a thunderstorm was detected at Venice Airport. As this storm cloud progressed over the Adriatic Sea (Fig. 3 c), it impacted the sea, generating a long wave with a peak amplitude of 44.95 cm in Mali Lošinj, and 26.31 cm in Stari Grad. It should be noted that Mali Lošinj and, in particular, Stari Grad tide gauges are not located at top of harbors where strongest meteotsunami oscillations occur but at relatively deep locations where lower sea level oscillations are observed during meteotsunamis (Fig. 2 f-g). The propagation of the Adriatic Sea meteotsunamis were modeled along the segment of the coast from 40° N to 47° N and 12° E to 20° E (encompassing both Mali Lošinj and Stari Grad) with the spacial resolution of 13 arc seconds (approximately 400 meters), with model time step of 0.3 seconds. High-resolution models were setup for the locations of Mali Lošinj and Stari Grad. These models used the propagation simulations as boundary input, with continuing air pressure forcing applied throughout all model grids. The model prediction for the tide gauge were obtained from the model grid with the highest model resolution of 1 arc second (about 30 meters). For the July 20, 2023 model (Fig. 3 c), forcing was initiated as soon as the first precipitation appeared in radar images over the northern Adriatic Sea at 19:10 UTC and the forcing continued until 03:40 UTC on the 20th. Model of the July 22, 2023 event used forcing from 12:00 UTC to 19:00 UTC. Over 8 h of meteotsunami dynamics were simulated. Figure 4 reveals that our model has successfully captured the July 20 event in Mali Lošinj. Onset of oscillations, their maximum height, and period are all well modelled. The July 22 event was modelled slightly less satisfactory. In Mali Lošinj, amplitude and period of oscillations are well reproduced, however the initial modelled meteotsunami wave lags for ~ 20 minutes in comparison with the measured one. In Stari Grad, amplitude and period are again well reproduced, but there is now a ~ 30 min lag of modelled wave in comparison to the measured one. Nonetheless, we find the results satisfactory, and suggestive of the idea that real-time modelling based on radar reflectivity images, could be used to forecast Adriatic meteotsunami events. Mali Lošinj Bay has a peculiar shape, a 5.4 km long bay is connected to the open sea with a 466-meter-wide strait which is perpendicular to the bay. This narrow entrance strait restricts the entry of the meteotsunamis from the open sea, delaying their penetration into the bay. However, once that the waves enter the bay, the bay's elongated and confined shape enhances harbor resonance, amplifying the meteotsunami wave. As a result, the wave is observed with greater amplitudes after the passage of the atmospheric front than at the time of its passage. In contrast, Stari Grad is situated in a bay which is connected to the open sea with a wide mouth, allowing meteotsunamis to be observed immediately as the front moves over the region, without significant delay. 4. Discussion Our study examines a meteotsunami forecast approach, which combines weather radar measurements with numerical ocean modelling. We tested the applicability of the MOST tsunami forecasting model in predicting meteotsunamis across various global regions, using weather radar data as the primary source of information on atmospheric forcing. We tested the method at three locations (the East Coast of the U.S., Baltic and the Adriatic Seas) and, at each of these locations, achieved a high success rate in timely modelling of meteotsunamis. The East Coast event has been examined before by Titov and Moore ( 2021 ) with similar approach, but without the high-resolution modeling of area close to Jtide gauges. In this work we expanded the methodology into use of the nested tsunami modeling approach to examine feasibility and accuracy of the meteotsunami forecast including the possible inundation predictions. We documented clear improvement of accuracy of this high-resolution approach in comparison with a simpler propagation modeling with coarse grids. In addition, we demonstrated the feasibility of using the forecast system that was developed for the seismic tsunami inundation forecast (Titov et al., 2016 ) in application to meteotsunamis. To further evaluate the performance of the meteotsunami forecast method, we examined two additional representative events—Klaipėda and Stari Grad—where modeled pressure fields were compared with observations (Fig. 5 ). In the Klaipėda case (June 20, 2020), a noticeable temporal mismatch is observed between the modeled and recorded atmospheric pressure fields: the model captures the pressure disturbance slightly earlier than the observations (Fig. 5 ). This discrepancy is likely due to the low temporal resolution of the measured atmospheric pressure data at Klaipėda, which were available only at hourly intervals, likely missing the precise timing and structure of the pressure anomaly. Despite this, the modeled sea level response aligns well with the observed timing of the water level fluctuations, indicating that the model successfully simulates the generation and propagation of long waves triggered by the atmospheric disturbance. The ability of the model to reproduce the observed sea level variation—even with a slightly misaligned pressure input—demonstrates its robustness. However, this case also highlights the importance of high-resolution atmospheric pressure data for more accurately constraining the timing and intensity of meteotsunami-generating disturbances. The current lack of fine-scale surface pressure observations at many coastal locations limits our ability to fully capture these events in real time. In this context, weather radar data provides a valuable complementary solution, offering higher temporal and spatial resolution that can significantly enhance meteotsunami detection and modeling efforts. In the Stari Grad case (July 22, 2023), the modeled and observed atmospheric pressure fields exhibit excellent agreement, both in timing and amplitude. The pressure disturbance is well represented in the model, and its arrival coincides precisely with the onset of long-wave sea level oscillations recorded at the tide gauge (Fig. 5 ). This strong correspondence confirms that the model effectively captures the coupling between atmospheric forcing and the resulting sea level response. However, the first wave observed in the tide gauge record is not present in the model output. This suggests that the initial meteotsunami wave was generated earlier in the Adriatic Sea, as the MCS propagated from the northwest to the southeast. In contrast, the second wave—successfully captured by the model — was likely generated locally, within the open bay near Stari Grad, as the pressure disturbance passed directly over the area. The effects of meteotsunamis are most pronounced in shallow waters and narrow bays (located at the end of wide shelves), where further wave amplification due to coastal geometry and bathymetry can occur. Therefore, to predict meteotsunamis in shallow waters, issue timely warnings to tourists and port authorities, and ensure forecast accuracy, it is essential to start the meteotsunamis model as soon as possible, and optimally before the weather system comes near the endangered coast. This approach helps capture the initial signal and enhances the precision of wave activity predictions. The study highlights the potential of using weather radar data to forecast meteotsunamis in real-time especially in situations where a sudden change of atmospheric pressure, driven by MCSs, is expected. Meteotsunami modeling can be initiated upon detecting a developing thunderstorm or heavy rainfall, allowing immediate adaptation in regions such as the eastern US coast, ports located in narrow bays, and tourist hotspots like the Adriatic Sea. Predicting meteotsunamis associated with storm surges may require incorporating a wind component. The current model results include only atmospheric pressure forcing. While they capture some of the observed wave activity, significant discrepancies remain — especially during the initial and peak phases of the event. This is because cold-season meteotsunamis are linked to cyclonic activity, where both atmospheric pressure jumps and wind intensification play critical roles (Monserrat et al. 2006 ; Šepić and Vilibić 2011 ; Pellikka et al. 2022 ; Nesteckytė et al. 2023 ). Consequently, the current model configuration lacks the necessary components to fully reproduce these events. To accurately simulate cold-season meteotsunamis, wind stress or direct wind forcing must be incorporated into the modeling framework. 5. Conclusions This study demonstrates the potential of using weather radar reflectivity data as a proxy for atmospheric pressure anomalies in meteotsunami forecasting. The application of the MOST tsunami propagation model, combined with real-time radar data, offers a practical and adaptable approach for predicting meteotsunami events across various geographic regions, including the East Coast of the United States, the southeastern Baltic Sea, and the Adriatic Sea. The results further highlight the importance of Proudman resonance in wave amplification and underscore the critical role of atmospheric pressure fluctuations in generating these hazardous events. Radar-based methods present several key advantages: (1) the data are available in near-real time; (2) radar image sequences provide direct information on the speed, direction, and affected area of the moving pressure anomaly; and (3) the spatial and temporal resolution is sufficient to be used directly in tsunami generation and propagation models—removing the need for complex interpolation, as is often required with sparse pressure or tide gauge measurements. Notably, real-time radar data are especially effective in tracking tsunamigenic atmospheric disturbances associated with convective systems rich in hydrometeors. Additional findings indicate that high-resolution models substantially improve meteotsunami simulation accuracy, with pseudo-R-squared values confirming better predictive performance. The model also proved sensitive enough to detect sea level oscillations triggered by even weak atmospheric disturbances, related to light rain showers, underscoring its responsiveness. Finally, the study emphasizes the importance of initiating model simulations before the weather system reaches the target location to enhance the precision of wave activity forecasts. Declarations Acknowledgement: We sincerely thank Lukas Sudvajus (LHMT) for his support, expertise, and data contribution, and Hrvoje Mihanović (Institute of Oceanography and Fisheries, Split, Croatia) for providing essential data. Their assistance has been invaluable to this research. Croatia radar data Data owner: Croatian Meteorological and Hydrological Service Ethical Approval Not applicable. Competing interests Not applicable. Authors' contributions Laura Nesteckytė performed the modeling and data analysis and wrote the initial manuscript draft. Vasily Titov guided the modeling approach and contributed to manuscript writing and editing. Jadranka Šepić analyzed atmospheric conditions and contributed to writing and editing. Loreta Kalpšaitė-Rimkienė supported the study design and manuscript revision. Funding Loreta Kelpšaitė-Rimkienė and Laura Nesteckytė were partly supported by the WaveWise project received funding from the Research Council of Lithuania (LMTLT), agreement No SMIP-24-140. Jadranka Šepić was supported by the ERC StG 853045 SHExtreme Grant, the ERC PoC 101213756 MeD-Track, the Croatian Science Foundation projects IP-2019-04-5875 StVar-Adri and IP-2022-10-4144 CroClimExtremes and STIM–REI project, Contract Number: KK.01.1.1.01.0003, a project funded by the European Union through the European Regional Development Fund – the Operational Programme Competitiveness and Cohesion 2014-2020 (KK.01.1.1.01). Measurement of sea level and atmospheric pressure in the Adriatic Sea was supported through the research project KLIMADRIA, funded by the European Union - NextGenerationEU. Availability of data and materials The sea level, sea surface atmospheric pressure, wind speed, and direction data from The Lithuanian Hydrometeorological Service can be obtained on request from the Lithuanian Hydrometeorological Service. References Angove M, Vilibić I, Heidarzadeh M, Monserrat S, Okal E, Rabinovich AB, Šepić J, Titov VV (2025) Meteotsunamis: definition, detection and alerting services investigation. UNESCO Publication. IOC Technical Series, IOC/2025/TS/200, 31 pp. Accessible at: https://unesdoc.unesco.org/ark:/48223/pf0000392653.locale=fr Bailey K, DiVeglio C, Welty A (2014) An Examination of the June 2013 East Coast Meteotsunami Captured By NOAA Observing Systems. NOAA, Silver Spring, Maryland Bechle AJ, Wu CH (2014) An entropy‐based surface velocity method for estuarine discharge measurement. Water Resources Research 50:6106–6128. https://doi.org/10.1002/2014WR015353 Belušić D, Strelec Mahović N (2007) Detecting and following atmospheric disturbances with a potential to generate meteotsunamis in the Adriatic. Physics and Chemistry of the Earth, Parts A/B/C 34:918–927. https://doi.org/10.1016/j.pce.2009.08.009 Boslough M, Titov V (2024) Air-coupled tsunamis generated from impacts and airbursts: Our understanding before Hunga-Tonga Hunga-Ha’apai. Acta Astronautica 222:641–646. https://doi.org/10.1016/j.actaastro.2024.06.041 Churchill DD, Houston SH, Bond NA (1995) The Daytona Beach wave of 3–4 July 1992: a shallow water gravity wave forced by a propagating squall line. Croatian National Tourist Board Visit Mali Lašinj. https://visitmali-losinj.croatia.hr/en-gb. Accessed 17 Mar 2025 Dalmatia Express Stari Grad, Croatia - a Complete Guide. https://dalmatia-express.com/blog/stari-grad-complete-guide/. Accessed 17 Mar 2025 Denamiel C, Šepić J, Huan X, et al (2019a) Stochastic Surrogate Model for Meteotsunami Early Warning System in the Eastern Adriatic Sea. JGR Oceans 124:8485–8499. https://doi.org/10.1029/2019JC015574 Denamiel C, Šepić J, Ivankovic D, Vilibić I (2019b) The Adriatic Sea and Coast modelling suite: Evaluation of the meteotsunami forecast component. Ocean Modelling 135:71–93. https://doi.org/10.1016/j.ocemod.2019.02.003 Dusek G, DiVeglio C, Licate L, et al (2019) A Meteotsunami Climatology along the U.S. East Coast. Bulletin of the American Meteorological Society 100:1329–1345. https://doi.org/10.1175/BAMS-D-18-0206.1 European Centre for Medium-Range Weather Forecasts ECMWF. https://www.ecmwf.int/. Accessed 15 Mar 2025 Garner S (2015) The Relationship between Hurricane Potential Intensity and CAPE. Journal of the Atmospheric Sciences 72:141–163. https://doi.org/10.1175/JAS-D-14-0008.1 Greenspan M (1956) Propagation of sound in five monatomic gases. The Journal of the AcousticalSociety of America 28:644–648. https://doi.org/10.1121/1.1908432 Ha T, Yoon J-S, Heo K-Y (2018) Numerical Simulations of a Meteotsunami using both Atmospheric and Phase-resolving Wave Models in the Yellow Sea. Journal of Coastal Research 85:786–790. https://doi.org/10.2112/SI85-158.1 Hersbach H, Bell B, Berrisford P, et al (2020) The ERA5 global reanalysis. Quart J Royal Meteoro Soc 146:1999–2049. https://doi.org/10.1002/qj.3803 Kelpšaitė-Rimkienė L, Soomere T, Bagdanavičiūtė I, et al (2018) Measurements of Long Waves in Port of Klaipėda, Lithuania. Journal of Coastal Research 85:761–765. https://doi.org/10.2112/SI85-153.1 Mahovic NS, Smiljanic I, Kerkmann J (2023) Destructive convective storms in SE Europe. In: Eumetsat. https://user.eumetsat.int/resources/case-studies/destructive-convective-storms-in-se-europe. Accessed 3 Mar 2025 Mofjeld HO, Titov VV, González FI, Newman JC (2001) Tsunami scattering provinces in the Pacific Ocean. Geophysical Research Letters 28:335–337. https://doi.org/10.1029/2000GL011710 Monserrat S, Vilibić I, Rabinovich AB (2006) Meteotsunamis: atmospherically induced destructive ocean waves in the tsunami frequency band. Nat Hazards Earth Syst Sci 6:1035–1051. https://doi.org/10.5194/nhess-6-1035-2006 Mourre B, Santana A, Buils A, et al (2021) On the potential of ensemble forecasting for the prediction of meteotsunamis in the Balearic Islands: sensitivity to atmospheric model parameterizations. Nat Hazards 106:1315–1336. https://doi.org/10.1007/s11069-020-03908-x National Weather Service When the National Weather Service is Watching Radar... In: National Weather Service. https://www.weather.gov/bmx/radar_aboutnwsradar. Accessed 25 Nov 2024 Nesteckytė L, Kelpšaitė-Rimkienė L, Rabinovich AB (2023) Hazardous meteotsunami-like sea-level oscillations in the Port of Klaipeda, the Baltic Sea. Nat Hazards. https://doi.org/10.1007/s11069-023-06311-4 Nesteckytė L, Stankūnavičius G, Kelpšaitė-Rimkienė L, Šepić J (2024) 2023 November Storm in The Port of Klaipeda; Why was The Port Closed? Pure Appl Geophys. https://doi.org/10.1007/s00024-024-03596-9 Orlić M (1980) About a possible occurrence of the Proudman resonancein the Adriatic. Thalassia Jugosl Orlić M (2015) The first attempt at cataloguing tsunami-like waves of meteorological origin in Croatian coastal waters Orlić M, Belušić D, Janeković I, Pasarić M (2010) Fresh evidence relating the great Adriatic surge of 21 June 1978 to mesoscale atmospheric forcing. J Geophys Res 115:2009JC005777. https://doi.org/10.1029/2009JC005777 Pasquet S, Vilibić I (2013) Shelf edge reflection of atmospherically generated long ocean waves along the central U.S. East Coast. Continental Shelf Research 1–8. https://doi.org/10.1016/j.csr.2013.06.007 Pattiaratchi CB, Wijeratne EMS (2015) Are meteotsunamis an underrated hazard? Phil Trans R Soc A 373:20140377. https://doi.org/10.1098/rsta.2014.0377 Pellikka H (2020) Dark-blue horizon: Sea level rise and meteotsunamis on the Finnish coast. Finnish Meteorological Institute Pellikka H, Laurila TK, Boman H, et al (2020) Meteotsunami occurrence in the Gulf of Finland over the past century. Nat Hazards Earth Syst Sci 20:2535–2546. https://doi.org/10.5194/nhess-20-2535-2020 Pellikka H, Šepić J, Lehtonen I, Vilibić I (2022) Meteotsunamis in the northern Baltic Sea and their relation to synoptic patterns. Weather and Climate Extremes 38:100527. https://doi.org/10.1016/j.wace.2022.100527 Pešice P, Sulan J, Řezáčová D (2003) Convection precursors in the Czech territory. Atmospheric Research 67–68:523–532. https://doi.org/10.1016/S0169-8095(03)00070-X Piotrowski A, Szczuciński W, Sydor P, et al (2017) Sedimentary evidence of extreme storm surge or tsunami events in the southern Baltic Sea (Rogowo area, NW Poland). GQ. https://doi.org/10.7306/gq.1385 Proudman J (1929) The Effects on the Sea of Changes in Atmospheric Pressure. Geophysical Journal International 2:197–209. https://doi.org/10.1111/j.1365-246X.1929.tb05408.x Punkka A-J, Bister M (2015) Mesoscale Convective Systems and Their Synoptic-Scale Environment in Finland. Weather and Forecasting 30:182–196. https://doi.org/10.1175/WAF-D-13-00146.1 Rabinovich A (2020) Twenty-seven years of progress in the science of meteorological tsunamis. Pure Appl Geophys Rabinovich AB (2009) Seiches and Harbor Oscillations. In: Handbook of Coastal and Ocean Engineering. World Scientific, p 1192 Rabinovich AB, Šepić J, Thomson RE (2021) The meteorological tsunami of 1 November 2010 in the southern Strait of Georgia: a case study. Nat Hazards 106:1503–1544. https://doi.org/10.1007/s11069-020-04203-5 Rauhala J, Kahma KK, Stipa T, et al (2014) Recent observations of meteotsunamis on the Finnish coast. Nat Hazards 74:197–215. https://doi.org/10.1007/s11069-014-1150-3 Renquist H (1926) Ein Seebär in Finnland. Zur Frage Nach der Entstehung der Seebären, Geogr Ann 230–236 Ruić K, Šepić J, Mlinar M, Međugorac I (2023) Contribution of high-frequency (T < 2 h) sea level oscillations to the Adriatic sea level maxima. Nat Hazards 116:3747–3777. https://doi.org/10.1007/s11069-023-05834-0 Sallenger AH, List JH, Gelfenbaum G, et al (1995) Large Wave at Daytona Beach, Florida, Explained as a Squall-line Surge. Journal of Coastal Research 11:1383–1388. https://www.jstor.org/stable/4298439 Šepić J, Metlicic N, Orlic M (2025) Synoptic conditions corresponding to the Adriatic meteorological tsunamis. https://meetingorganizer.copernicus.org/EGU24/EGU24-14855.html. Accessed 9 Apr 2025 Šepić J, Vilibić I (2011) The development and implementation of a real-time meteotsunami warning network for the Adriatic Sea. Nat Hazards Earth Syst Sci 11:83–91. https://doi.org/10.5194/nhess-11-83-2011 Šepić J, Vilibić I, Belušić D (2009) Source of the 2007 Ist meteotsunami (Adriatic Sea). J Geophys Res 114:C03016. https://doi.org/10.1029/2008JC005092 Šepić J, Vilibić I, Monserrat S (2016) Quantifying the probability of meteotsunami occurrence from synoptic atmospheric patterns: Meteotsunami Index. Geophys Res Lett 43:10,377-10,384. https://doi.org/10.1002/2016GL070754 Šepić J, Vilibić I, Rabinovich AB, Monserrat S (2015) Widespread tsunami-like waves of 23-27 June in the Mediterranean and Black Seas generated by high-altitude atmospheric forcing. Sci Rep 5:11682. https://doi.org/10.1038/srep11682 The University Corporation for Atmospheric Research (2012) Weather Radar Fundamentals. https://research.atmos.ucla.edu/weather/C110/Documents/tmp/basic_wxradar/index.htm. Accessed 8 Jan 2025 Titov V, González F (2000) Numerical study of the source of the 17 July 1998 PNG tsunami. https://doi.org/10.1007/978-94-017-3618-3_14 Titov V, Kânoğlu U, Synolakis C (2016) Development of MOST for Real-Time Tsunami Forecasting. J Waterway, Port, Coastal, Ocean Eng 142:03116004. https://doi.org/10.1061/(ASCE)WW.1943-5460.0000357 Titov V, Moore C (2021) Meteotsunami model forecast: can coastal hazard be quantified in real time? Nat Hazards 106:1545–1561. https://doi.org/10.1007/s11069-020-04450-6 Titov VV, Moore CW, Greenslade DJM, et al (2011) A New Tool for Inundation Modeling: Community Modeling Interface for Tsunamis (ComMIT). Pure Appl Geophys 168:2121–2131. https://doi.org/10.1007/s00024-011-0292-4 USA-Reisen.de (2025) Atlantic City Stadt, Strand & Spaß! Accessed 25 Nov 2024 Vilibić I (2008) Numerical simulations of the Proudman resonance. Continental Shelf Research 28:574–581. https://doi.org/10.1016/j.csr.2007.11.005 Vilibić I, Domijan N, Orlić M, et al (2004) Resonant coupling of a traveling air pressure disturbance with the east Adriatic coastal waters. J Geophys Res 109:2004JC002279. https://doi.org/10.1029/2004JC002279 Vilibić I, Horvath K, Strelec Mahović N, et al (2014) Atmospheric processes responsible for generation of the 2008 Boothbay meteotsunami. Nat Hazards 74:25–53. https://doi.org/10.1007/s11069-013-0811-y Vilibić I, Šepić J, Rabinovich AB, Monserrat S (2016) Modern Approaches in Meteotsunami Research and Early Warning. Front Mar Sci 3:. https://doi.org/10.3389/fmars.2016.00057 Villalonga J, Monserrat S, Gomis D, Jordà G (2024) Observational Characterization of Atmospheric Disturbances Generating Meteotsunamis in the Balearic Islands. JGR Oceans 129:e2024JC020910. https://doi.org/10.1029/2024JC020910 Wertman CA, Yablonsky RM, Shen Y, et al (2014) Mesoscale convective system surface pressure anomalies responsible for meteotsunamis along the U.S. East Coast on June 13th, 2013. Sci Rep 4:7143. https://doi.org/10.1038/srep07143 Whitmore P, Knight B (2014) Meteotsunami forecasting: sensitivities demonstrated by the 2008 Boothbay, Maine, event. Nat Hazards 74:11–23. https://doi.org/10.1007/s11069-014-1056-0 Wijeratne EMS, Pattiaratchi CB (2024) Meteotsunamis Generated by Thunderstorms. JGR Oceans 129:e2023JC020662. https://doi.org/10.1029/2023JC020662 Yano J-I, Ziemiański MZ, Cullen M, et al (2018) Scientific Challenges of Convective-Scale Numerical Weather Prediction. Bulletin of the American Meteorological Society 99:699–710. https://doi.org/10.1175/BAMS-D-17-0125.1 Port of Klaipėda. https://www.portofklaipeda.lt/. Accessed 25 Sep 2024 ERA5 Reanalysis. In: Copernicus Climate Change Service (C3S): ERA5: Fifth generation of ECMWF atmospheric reanalyses of the global climate. Copernicus Climate Change Service Climate Data Store (CDS). https://cds.climate.copernicus.eu/cdsapp#!/home Supplementary Files Nesteckytesupplementarymaterials.pdf Cite Share Download PDF Status: Published Journal Publication published 04 Feb, 2026 Read the published version in Natural Hazards → Version 1 posted Reviewers agreed at journal 13 Jun, 2025 Reviewers invited by journal 11 Jun, 2025 Editor assigned by journal 10 Jun, 2025 First submitted to journal 10 Jun, 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-6861009","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":469919336,"identity":"b5d48662-8077-4777-9717-5dcc7e098402","order_by":0,"name":"Laura Nesteckytė","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA8klEQVRIiWNgGAWjYBACPmQOYwOQ4OcBEg/waGFDZgG1GDBI9gA5CQmkaDE4Q0gL+xmzDz8Y7sjJz28++HFm2x854zNnDBgSf+DRwpNjPLOH4ZmxwTG2ZMmNbQbGZmd7DAg4LMeYgYfhcOIGNh4zxodtBonbzvMQ0ML/xpjxD8Ph+vlt/N9AWuo39xPSIpFjzAy0JYHhGA8bI9BhCQa8hBwm8ayYWcbgsOGGY2nGkjPOGRvOOHOs4EBCGm4t/PzJmxnfVByWl28+/PBjT5mcPH9P8sYHH2xwa4EAAyjNCI2mA4Q0IIE/JKgdBaNgFIyCEQMAx7tLA/r3WiMAAAAASUVORK5CYII=","orcid":"","institution":"Klaipedos universitetas","correspondingAuthor":true,"prefix":"","firstName":"Laura","middleName":"","lastName":"Nesteckytė","suffix":""},{"id":469919337,"identity":"399d8e30-05bc-4c4b-92fc-a905952b63e9","order_by":1,"name":"Vasily Titov","email":"","orcid":"","institution":"NOAA PMEL: NOAA Pacific Marine Environmental Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Vasily","middleName":"","lastName":"Titov","suffix":""},{"id":469919338,"identity":"39ef3388-829f-4828-898c-e11a21badf12","order_by":2,"name":"Loreta Kelpšaitė-Rimkienė","email":"","orcid":"","institution":"Klaipedos universitetas","correspondingAuthor":false,"prefix":"","firstName":"Loreta","middleName":"","lastName":"Kelpšaitė-Rimkienė","suffix":""},{"id":469919339,"identity":"30541767-2f7b-47f5-8ee6-1ae94901d13a","order_by":3,"name":"Jadranka Šepić","email":"","orcid":"","institution":"University of Split: Sveuciliste u Splitu","correspondingAuthor":false,"prefix":"","firstName":"Jadranka","middleName":"","lastName":"Šepić","suffix":""}],"badges":[],"createdAt":"2025-06-10 08:45:58","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6861009/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6861009/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s11069-025-07894-w","type":"published","date":"2026-02-04T15:57:22+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":84551532,"identity":"abd2def5-1ab7-45f9-af94-c06174fd0585","added_by":"auto","created_at":"2025-06-13 10:25:56","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1026279,"visible":true,"origin":"","legend":"\u003cp\u003eMap of study sites. Red diamonds mark Atlantic City, Montauk, Lewes, Klaipėda, Mali Lošinj, and Stari Grad, red rectangles indicate the modeling domains. The inset shows a radar image of the storm complex with timestamps indicating its progression (source: G. Carbin, NWS/SPC, NOAA). Zoomed-in views of study sites are shown in Fig. 2.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6861009/v1/7cb14abdbd8e886a2c8c00c8.png"},{"id":84551554,"identity":"e19c1978-2d83-46af-9d71-0930f20cbdbd","added_by":"auto","created_at":"2025-06-13 10:25:59","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1180869,"visible":true,"origin":"","legend":"\u003cp\u003eSituation maps, color bars indicate elevation in m, bathymetry maps and weather radar data (arrow indicates weather disturbance propagation direction), yellow buoys indicate tide gauges locations: a) derecho propagation over the East Coast shelf, b) thunderstorm propagation over Lithuania, c) thunderstorm propagation over Croatia, d) Atlantic City harbor (USA-Reisen.de), e) the Port of Klaipėda (portofklaipeda.lt), f) Mali Lošinj Bay (Croatian National Tourist Board), g) Stari Grad Bay (Dalmatia Express)\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6861009/v1/bd3eb33f5726d393ade2b6b7.png"},{"id":84551553,"identity":"4e4dc2e1-f106-4695-a6c5-d04b526d8c1c","added_by":"auto","created_at":"2025-06-13 10:25:58","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":524804,"visible":true,"origin":"","legend":"\u003cp\u003ea) picture of 23 June 2013 the East Coast Derecho (Buddy Denham), b) thunderstorm clouds in Lithuania during 20 June 2020 (lrytas.lt), c) thunderstorm front over Zagreb, Croatia 19 July 2023 (Darko Ćorluka)\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6861009/v1/46ddc87472feeee04efc1a86.png"},{"id":84551561,"identity":"7135e87b-824b-4dea-ae34-1e9af045464c","added_by":"auto","created_at":"2025-06-13 10:26:01","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":96857,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of the model (black line) and measured sea-level amplitudes (red line) at the East Coast of the United States (Montauk, Atlantic City, and Lewes), Baltic Sea, Klaipėda (20 June 2020), Croatia, Mali Lošinj (19 and 22 July 2023) and Stari Grad (22 July 2023)\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6861009/v1/bac7c4364baa68ff3ad269a0.png"},{"id":84551779,"identity":"c618c8cc-94a6-48a5-b3c7-b2217812cbd5","added_by":"auto","created_at":"2025-06-13 10:33:57","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":290760,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of modeled (red line) and observed (black line) sea level amplitudes and atmospheric pressure data (straight blue line) and simulated atmospheric pressure (dash green line) upper two for the Port of Klaipeda Lithuania (notice that observed pressure is measured with inadequate hourly resolution), lower two for the Stari Grad, Adriatic Sea\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-6861009/v1/74121a4a61b4685f0f49a89f.png"},{"id":102234242,"identity":"9a2d5ae5-3252-42c2-bf76-84fdb443d76a","added_by":"auto","created_at":"2026-02-09 16:08:23","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3937075,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6861009/v1/b1929e97-dca1-4add-98a9-066453aad04a.pdf"},{"id":84551535,"identity":"5eeb63e9-eb16-4035-92e7-66409e9f3f1c","added_by":"auto","created_at":"2025-06-13 10:25:56","extension":"pdf","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":637048,"visible":true,"origin":"","legend":"","description":"","filename":"Nesteckytesupplementarymaterials.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6861009/v1/b2be64c4e48319c0e3de6121.pdf"}],"financialInterests":"","formattedTitle":"Advancing meteotsunami prediction: testing high-resolution modeling in harbor environments","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eThe focus of global discussions on tsunami forecast and warning has expanded from the hazards of seismic tsunamis to include the risks posed by meteorologically-generated tsunamis (meteotsunamis) (Monserrat et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Pattiaratchi and Wijeratne \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Pellikka \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Titov and Moore \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Wijeratne and Pattiaratchi \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Nesteckytė et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Angove et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Much like seismically-generated tsunamis, meteotsunamis pose a particular threat to populations and infrastructure located along shallow coastlines, in narrow bays, and in harbors (Monserrat et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Rabinovich \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Seismic tsunami forecasting has advanced significantly, with the establishment of warning centers such as the National Tsunami Warning Center and the development of modeling platforms like SIFT, Tweb and ComMIT (Titov et al. \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Pattiaratchi and Wijeratne \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), enabling more reliable predictions of tsunamis, once a tsunamigenic earthquake has occurred. In contrast, forecasting meteorological tsunamis and issuing reliable warnings remains a challenge.\u003c/p\u003e \u003cp\u003eMeteotsunamis are a special group of tsunamis. They are generated by propagating atmospheric pressure (and wind) disturbances of limited spatial dimensions (tens of kilometers) and relatively short duration (tens of minutes to couple of hours) (Monserrat et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Vilibić et al. \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). To trigger a destructive meteotsunami, atmospheric disturbance needs to be characterized by a high rate of air pressure change (a couple of hPa over a few minutes), it must propagate over a shelf of slowly varying depth with a speed that is equal (or almost equal) to speed \u003cem\u003ec\u003c/em\u003e of long ocean waves at that shelf (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:c=\\sqrt{gh}\\)\u003c/span\u003e\u003c/span\u003e, where \u003cem\u003eg\u003c/em\u003e is gravity acceleration, and \u003cem\u003eh\u003c/em\u003e is ocean depth) or along a gently sloping bathymetry with a speed that is equal to speed of trapped edge waves. This equality of speeds is a prerequisite for resonant transfer of energy from atmospheric pressure disturbances to long ocean waves, i.e. for Proudman (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e1929\u003c/span\u003e) and Greenspan resonances (Greenspan \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e1956\u003c/span\u003e). Even a small offset of atmospheric disturbance from the \u0026ldquo;ideal\u0026rdquo; tsunamigenic properties (exact track, value of rate of air pressure change, speed) can result with an order of magnitude different height of a meteotsunami (e.g., Šepić et al. \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTo track and forecast such tsunamigenic atmospheric disturbances we need to use high-resolution atmospheric pressure measurements and models \u0026ndash; with high resolution in spatial and temporal dimensions - that can give the location, speed, amplitude, and propagation direction of the pressure disturbances that generate meteotsunamis (Pattiaratchi and Wijeratne \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Ha et al. \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Titov and Moore \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Villalonga et al. \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Wijeratne and Pattiaratchi \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). However, even the most sophisticated weather models fail to capture exact properties of tsunamigenic atmospheric pressure disturbance (Denamiel et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2019a\u003c/span\u003e; Mourre et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), which begs the question: Is real-time forecasting of meteotsunamis possible? One alternative is to use measurements of atmospheric pressure, which are assessed in real time. However, these measurements are often of low spatial resolution (distances between stations are of order of tens of kilometers, i.e., of order of spatial dimensions of atmospheric disturbances) and, thus, do not allow for capturing all relevant properties of small-scale short-lasting atmospheric pressure disturbances (Titov et al. \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Šepić et al. \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). This study explores an alternative method for meteotsunami forecasting. Our method combines real-time measurements and numerical ocean modelling: we use precipitation reflection images obtained from weather radars to derive data which can be used to force a numerical ocean model. The radar reflectivity images allow us to simulate atmospheric pressure disturbance as a direct, time-dependent input for numerical ocean meteotsunami generation and propagation model. The governing idea is that higher radar reflectivity intensities correspond to relatively higher atmospheric pressure anomalies. While the absolute pressure level is unknown from the radar images, the pressure can be scaled by atmospheric pressure measurements available at the meteorological station near the prediction area (Titov and Moore \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDoppler radar data are widely available worldwide, from radars deployed for various purposes, particularly meteorology, aviation, and defence. A weather radar works by emitting pulses of electromagnetic energy at microwave frequencies into the atmosphere. Upon encountering various particles in the air, such as water and ice droplets, some of this electromagnetic energy is scattered back toward the radar\u0026mdash;this phenomenon is commonly termed \"reflection,\" which forms the basis of \"reflectivity\" (The University Corporation for Atmospheric Research, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Reflectivity quantifies the efficiency of a radar target in intercepting and returning radar energy and is influenced by specific physical characteristics of the target, such as size, shape, orientation, and composition. The returned signal is processed by computer systems, enabling the assessment of precipitation location and intensity, along with wind speed and direction. Doppler radars deliver localized, high-resolution, timely, and three-dimensional observational capabilities that surpass those provided by other meteorological monitoring tools. They can detect changes in precipitation rates with a spatial resolution of a few square kilometers or finer and operate on temporal cycles of just a few minutes. This advanced capacity allows for the monitoring of rapidly evolving weather phenomena, which is crucial for issuing timely warnings regarding severe and hazardous weather conditions, including heavy rainfall, hail, strong winds (such as those associated with tornadoes and tropical and extratropical cyclones), and wind shear (National Weather Service). Furthermore, recent advancements in numerical weather prediction systems have facilitated the integration of continental-scale, radar-derived precipitation data into global models, thereby enhancing the accuracy of 4 - to 5-day precipitation forecasts for surrounding regions and continent (European Centre for Medium-Range Weather Forecasts). If radar data improves short-term weather forecasts, maybe it can also be used for forecasting meteotsunamis.\u003c/p\u003e \u003cp\u003eGlobally, networks of Doppler radars are operated by national meteorological agencies, such as the National Weather Service in the United States, Environment and Climate Change Canada, and various national meteorological services across Europe and around the Globe. In Europe, the EUMETNET OPERA programme supports the harmonization and exchange of radar data between national services, although it does not operate radar systems itself. These expansive networks ensure comprehensive coverage, providing timely and precise data essential for public safety and effective disaster management. The global standardization of weather radar allows seamless comparison of radar reflectivity data from different regions of the world, eliminating the need for an extensive and complicated data acquisition process. Consequently, a weather radar-based meteotsunami prediction model has the potential for widespread application. The advantage of using radar images for rapid meteotsunami source assessment lies in three key factors: (1) the data are available in near-real time; (2) the image sequence provides direct measurements of the speed, direction, and potential area of the pressure anomaly\u0026rsquo;s movement; (3) the spatial and temporal resolution of the radar imaging is sufficient to be used directly in tsunami generation and propagation models, eliminating the need for complex interpolation and interpretation \u0026mdash; unlike sparse atmospheric pressure or sea-level measurements. It's important to note that real-time radar data is especially useful for tracking atmospheric disturbances that can generate tsunamis, particularly when those disturbances are part of convective systems rich in hydrometeors.\u003c/p\u003e \u003cp\u003eThis study examines the feasibility of broad geographic applicability of this method, focusing on three areas with different geographies, different meteotsunami generation mechanisms and different meteotsunami characteristics: (1) the East Coast of the United States (2) the Baltic Sea; and (3) the Adriatic Sea (the semi enclosed sea at the north of the Mediterranean Sea) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eMeteotsunamis are often classified as winter and summer events (Pattiaratchi and Wijeratne \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Dusek et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Pellikka et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2020\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Nesteckytė et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Wijeratne and Pattiaratchi \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). On the East Coast of the U.S., wintertime meteotsunamis are associated with large storm systems that impact broad areas and are often accompanied by other hazardous oceanographic conditions, such as storm surges and large waves (Dusek et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). The three strongest known East Coast meteotsunamis are the Daytona Beach meteotsunami of 4 July 1992 (associated to passage of squall line (Churchill et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e1995\u003c/span\u003e; Sallenger et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e1995\u003c/span\u003e), the Boohtbay meteotsunami of 28 October 2008 (associated to cold front and atmospheric gravity waves (Vilibić et al. \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Whitmore and Knight \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), and the New Jersey meteotsunami of 13 June 2013 (associated with derecho (Bailey et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Wertman et al. \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eWinter meteotsunamis in the Baltic Sea primarily occur between November and April (Pellikka et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Nesteckytė et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). These events are closely linked to cold front passages and deep, large-scale extratropical cyclones moving from the west-northwest-north (Pellikka et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The extratropical cyclones are characterized by strong surface winds, but also by significant smaller-scale short-period atmospheric pressure changes, creating the conditions necessary for meteotsunami generation. The dynamics of winter meteotsunamis reflect the harsh and energetic weather systems prevalent during the colder months. Summer meteotsunamis in the Baltic Sea are most common between late April and October and are linked to mesoscale convective systems (MCSs). Such systems frequently result in the occurrence of thunderstorms, rapid changes in atmospheric pressure, and, in certain instances, the formation of clear frontal structures, including gust fronts, squall lines, and hail. MCSs are most prevalent in July and August, with infrequent occurrences of intense MCSs during the winter months. Such events may be as severe as those classified as U.S. continental derechos (Punkka and Bister \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Pellikka et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn the Adriatic Sea, meteotsunamis can also be classified into wintertime and summertime events (Ruić et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; 2025; Šepić et al. 2024). The Adriatic Sea wintertime meteotsunamis are associated with strong extratropical cyclones and appear simultaneously to storm surges, the summertime meteotsunamis are, on the other hand, usually associated with atmospheric gravity waves which usually appear jointly with convective storms or mesoscale convective systems (e.g., Belušić and Strelec Mahović \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Šepić et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2009\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eConclusively, for all locations, summer meteotsunamis are typically driven by localized MCSs and rapid air pressure jumps which are very difficult to forecast precisely. And winter events are generally associated with larger-scale synoptic patterns, i.e., with extratropical cyclones, and occasionally tropical cyclones (only at the East Coast of the U.S.). Current synoptic models, such as those developed by the European Centre for Medium-Range Weather Forecasts (ECMWF) and the Global Forecast System (GFS), demonstrate robust capabilities in forecasting extratropical cyclones at mid-latitudes. This capability provides valuable context that helps identify conditions favorable for winter meteotsunami events. However, these large-scale models alone cannot resolve the small-scale atmospheric pressure disturbances embedded within cyclones, which directly generate meteotsunamis. Therefore, while accurate extratropical cyclone predictions improve meteotsunami forecasting potential, they do not entirely solve the prediction challenge. Furthermore, when dealing with rapidly evolving convective systems, such as thunderstorms, squall lines, or even derecho events, which are notoriously difficult to predict\u0026mdash;even with advanced models like ECMWF or GFS \u0026mdash; it becomes essential to have a real-time tool capable of assessing meteotsunami threats based on up-to-date meteorological data, including weather radar inputs.\u003c/p\u003e \u003cp\u003eThis work tested the hypothesis of using weather radar reflectivity as a proxy for atmospheric pressure at three different locations around the world, the East Coast of the United States, the southeast coast of the Baltic Sea, and the Adriatic Sea (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e"},{"header":"2. Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Study sites\u003c/h2\u003e \u003cp\u003eThe East Coast of the United States (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) experiences an average of 25 meteotsunamis per year, according to Dusek et al. (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). The statistics was developed by extracting meteotsunami signals from 22 years of sea level data using a station-dependent threshold-surpassing criteria (meteotsunamis were defined as those events during which energy of high-frequency sea level oscillations and their maximum peak-to-trough height surpassed pre-described thresholds) superimposed on a meteorological criteria (necessity of an atmospheric pressure disturbance). Depending on the station, events with peak-to-trough height as low as 0.2 m were classified as meteotsunamis. Nonetheless, some of these events exceeded 0.6 meters in height, making them capable of causing significant damage and posing risks to coastal communities Furthermore, even stronger meteotsunamis have been observed (but not measured in their full strength) along the East Coast of the U.S., e.g., the Daytona Beach meteotsunami with estimated height of 3\u0026ndash;6 m (Churchill et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e1995\u003c/span\u003e; Sallenger et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e1995\u003c/span\u003e), and Boothbay meteotsunami with estimated height of 3.6 m (Whitmore and Knight \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). The East Coast regularly experiences weather systems capable of producing meteotsunamis, including squall lines, thunderstorms, frontal systems, and derechos, a widespread, long-lived windstorm associated with a band of rapidly-moving showers or thunderstorms (Dusek et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). These weather systems produce sudden changes in atmospheric pressure and strong wind gusts that transfer energy to the ocean surface. In addition, many bays, inlets, and harbors along the coast have natural resonance periods that align with the periods of propagating atmospheric disturbances, amplifying the resulting waves and increasing the likelihood of strong meteotsunamis. Over the East Coast of the U. S., meteotsunamis are generated by offshore (eastward) propagating atmospheric disturbances (Pasquet and Vilibić \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Titov and Moore \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Atmospheric disturbances generally come to the East Coast from the west (i.e., from the continental U.S.), they cross the coast and advance towards the east over the western North Atlantic shelf. While propagating over the shelf, atmospheric disturbances can enhance long ocean waves with Proudman resonance, given that velocity of the atmospheric disturbance matches velocity of long ocean waves (Proudman \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e1929\u003c/span\u003e). Atmospheric disturbance and forced long ocean wave travel coupled as long as their velocities match. However, around 135 km from the coast, the coupled system reaches the shelf break. Due to a sudden change of ocean depth, long ocean waves speed abruptly changes, so part of the ocean wave reflects back from the shelf break towards the East Coast of the U.S., while atmospheric disturbance continues eastward. From this point on, long ocean waves propagate as free gravity-driven long waves \u0026ndash; impacting the coast 4 to 6 hours after atmospheric disturbance has initially crossed it (Titov and Moore \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn the Baltic Sea (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), there are historical accounts of very strong Baltic meteotsunamis (e.g., Renquist \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e1926\u003c/span\u003e; Piotrowski et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). These events were mostly observed over the northward facing Polish and German coasts. Atmospheric disturbances propagating towards these coasts (from the north to the south) pass over seas of relatively uniform depths \u0026ndash; thus generating long ocean waves via Proudman resonance (Rauhala et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). The port of Klaipėda (our study location), however, faces the west. Hence, any atmospheric disturbance propagating towards the site would pass over the sea with significantly variable depth making the Proudman resonance unlikely, which appears to suggest low threat of strong meteotsunamis. Nonetheless, even weak meteotsunamis, can seriously affect the sea-going traffic and result in closure of the port (see Nesteckytė et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The Port of Klaipėda is located in the Klaipėda Strait (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee). The Klaipėda strait serves as an important connection between the southern Baltic Sea and the largest lagoon in Europe, Curonian Lagoon. The port's quays are located within a 10 km long and 0,36\u0026ndash;1,20 km wide inlet. Long ocean waves are usually intensified in the Port of Klaipėda due to its complex narrow and deep topography (Kelpšaitė-Rimkienė et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; (Kelpšaitė-Rimkienė et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Rabinovich et al. \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Nesteckytė et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2023\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe Adriatic Sea (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) is considered to be a meteotsunami hot spot (Šepić et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). According to Šepić and Orlić (2024), within the last\u0026thinsp;~\u0026thinsp;90 years 36 meteotsunamis with crest-to-trough wave heights above 1 m, and 13 with crest-to-trough wave heights above 2 m were observed, including one of the strongest known World meteotsunamis, the Great Vela Luka flood of 21 June 1978 \u0026ndash; an event when height of sea level oscillations reached astonishing 6 m (Orlić \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e1980\u003c/span\u003e; Orlić et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). In the Adriatic Sea, meteotsunamis are generated by atmospheric pressure disturbances propagating northeastward to eastward and directly towards the endangered coast (e.g., (Šepić et al. \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). The Adriatic meteotsunamis reach particularly large wave heights within bays characterized by high amplification (Q) factors (Rabinovich \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2009\u003c/span\u003e) such as Mali Lošinj, Široka Bay, Stari Grad and Vela Luka (Orlić \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). These are funnel-like bays which narrow and become shallow at the closed end. Nowadays, tide gauges are located at two meteotsunami-prone locations in the Adriatic Sea, i.e., in Mali Lošinj and Stari Grad Bays, and events hitting these two locations will be studied in more detail in this paper. Mali Lošinj Bay (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef), a natural harbor situated in a deep, narrow bay on the island of Lošinj, provides excellent shelter from winds and rough seas. With multiple piers and quays, it accommodates fishing boats, yachts, and ferries. Similarly, Stari Grad Bay (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eg), located on the northern coast of Hvar Island, is a long inlet whose natural configuration offers protection from most winds, making it a safe anchorage for boats and small ships. However, despite their well-sheltered appearances, both bays possess geographic features that favor amplification of incoming long ocean waves, energetic seiche oscillations and strong meteotsunami occurrences. Mali Lošinj has a relatively narrow but deep entrance, while Stari Grad extends inland in a fjord-like formation. These deep and narrow channels create ideal conditions for harbor resonance (Vilibić \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Šepić et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Wijeratne and Pattiaratchi \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), one of a key factor in meteotsunami enhancement. Meteotsunamis with wave heights\u0026thinsp;\u0026gt;\u0026thinsp;3 m have been observed at both these locations (Vilibić et al. \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Belušić and Strelec Mahović \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Šepić and Orlić, 2024). According to the online version of the Adriatic meteotsunami catalogue (Šepić and Orlić 2024), Mali Lošinj has experienced two significant meteotsunamis, one of which reached a runup of 3.5 meters, causing damage to homes, vessels, and shops. In Stari Grad, seven meteotsunamis with crest-to-trough wave-heights above 1 m have been documented since the 1930s (including two with heights of 2.5 and 3.5, respectively), resulting in damage to houses, businesses, and infrastructure.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Meteotsunami modelling approach\u003c/h2\u003e \u003cp\u003eFor meteotsunami simulations, the numerical model based on the Method of Splitting Tsunamis (MOST) formulation (Titov et al. \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2011\u003c/span\u003e, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) was employed. This model, widely recognized for its application in seismic tsunami modeling, has been extensively validated through laboratory experiments and historical tsunami simulations. Since 2012, NOAA has operationally used the MOST model for tsunami forecasting (Titov et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Its adaptability extends to simulating time-dependent tsunami initiation mechanisms, such as those caused by landslides, or atmospheric pressure forces (with the latter relating to meteotsunami generation processes) (Titov and Gonz\u0026aacute;lez \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Titov and Moore \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Boslough and Titov \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The MOST model incorporates atmospheric pressure forcing terms in Cartesian coordinates as follows:\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:\\frac{\\partial\\:u}{\\partial\\:t}+u\\frac{\\partial\\:u}{\\partial\\:x}+v\\frac{\\partial\\:u}{\\partial\\:y}+g\\frac{\\partial\\:h}{\\partial\\:x}=g\\frac{\\partial\\:d}{\\partial\\:x}-\\frac{1}{\\rho\\:}\\frac{\\partial\\:P}{\\partial\\:x};\\frac{\\partial\\:v}{\\partial\\:t}+u\\frac{\\partial\\:v}{\\partial\\:x}+v\\frac{\\partial\\:v}{\\partial\\:y}+g\\frac{\\partial\\:h}{\\partial\\:y}=g\\frac{\\partial\\:d}{\\partial\\:y}-\\frac{1}{\\rho\\:}\\frac{\\partial\\:P}{\\partial\\:y};\\frac{\\partial\\:h}{\\partial\\:t}+\\frac{\\partial\\:\\left(hu\\right)}{\\partial\\:x}+\\frac{\\partial\\:\\left(hv\\right)}{\\partial\\:y}=0;$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ehere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:h\\left(x,y,t\\right)=\\zeta\\:\\left(x,y,t\\right)+d\\left(x,y\\right),\\:\\zeta\\:\\left(x,y,t\\right)\\)\u003c/span\u003e\u003c/span\u003e is the amplitude, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:d\\left(x,y\\right)\\)\u003c/span\u003e\u003c/span\u003e is the water depth, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:u\\left(x,y,t\\right)\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:v(x,y,t)\\)\u003c/span\u003e\u003c/span\u003e are velocities in the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:x\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:y\\)\u003c/span\u003e\u003c/span\u003e directions accordingly, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:P\\)\u003c/span\u003e\u003c/span\u003e is atmospheric pressure, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\rho\\:\\)\u003c/span\u003e\u003c/span\u003e is water density, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:g\\)\u003c/span\u003e\u003c/span\u003e is the acceleration of gravity. The MOST numerically solves these shallow-water wave equations in spherical coordinates, incorporating Earth\u0026rsquo;s curvature, as detailed by Titov et al. (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe numerical approach we use in this study assumes several simplifications in the meteotsunami initiation process. When the input of atmospheric forcing in the model ends, the water surface disturbances created by atmospheric forcing continues propagating as free waves. This simplification can result in additional waves appearing behind the main meteotsunami wave as it reaches the coast. However, this approach represents the simplest real-time assumption, necessary when forecasts must be made before observational data becomes available. It is assumed that these artefactual waves at the meteotsunami's tail are of lesser magnitude and significance compared to the initial forced waves, as they are generated after the meteotsunami has decoupled from atmospheric forcing. In this formulation, wind stress forcing is omitted, leaving atmospheric pressure changes\u0026mdash;associated with the moving pressure field derived from radar images\u0026mdash;as the sole mechanism for wave generation. Although, simulations with added wind stress can produce improved results in meteotsunami modelling (e.g. Bechle and Wu \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), effect of wind stress remains secondary to effect of atmospheric pressure (e.g. Orlić et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Šepić et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), and wind stress is commonly omitted when modelling meteotsunamis.\u003c/p\u003e \u003cp\u003eThe meteotsunami modeling uses multi-grid simulation approach. The regional propagation of the meteotsunami in the basin-wide large area is performed using coarser resolution grid with simplified reflective boundaries along the coastline. Absorbing boundaries at the edges of the domain were applied to enable wave energy to exit the computational area without reflection. The coastal impact for smaller areas inside the regional grid around the tide gauges is modeled with nested grid computation that zooms the resolution of the model from the propagation grid into the fine-resolution grid of about 10\u0026ndash;50 meter spacing. The moving boundary conditions are applied for these nested grids to account for inundation on land.\u003c/p\u003e \u003cp\u003eThe coastal impact modeling that uses three nested grids was performed with the ComMIT tool. ComMIT is a user-friendly graphical tool that connects the MOST model with a pre-computed tsunami propagation scenario database, which allows for modeling of seismic tsunami impact at any coastline around the world. Beyond enhancing modeling accessibility, ComMIT fosters collaboration within the tsunami research community by facilitating results sharing. Its intuitive design allows for efficient visualization of modeling outputs and integration of initial conditions from precomputed propagation databases, all while maintaining minimal hardware requirements. Together, the MOST model and ComMIT provide a robust framework for simulating and understanding meteotsunami propagation in both research and operational contexts.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Numerical modelling of meteotsunamis using radar reflectivity images\u003c/h2\u003e \u003cp\u003eThe study sites have very different geography and bathymetry settings. Nevertheless, in all cases the propagation speed of the weather systems we modeled were sufficiently close to the long wave celerity over at least part of the propagation path to sustain Proudman resonance amplification. This resonance increased the meteotsunami amplitude significantly beyond the estimates provided by the static \u0026ldquo;inverted barometer\u0026rdquo; approach for all events. An analysis of the atmospheric front along the East Coast of the United States reveals that it travelled approximately 80 to 110 km within an hour (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). This corresponds to propagation speeds of around 22 to 30 m/s, which align with the long wave celerity in water depths ranging from 50 to 90 meters. In the case of Klaipėda (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb and e), the thunderstorm cloud traveled at speeds of 6 to 17 m/s. The average depth of the Port of Klaipėda is approximately 15 meters, and the average depth of Curonian lagoon is 3.8 m, while the depths near to the Lithuanian coast, where Proudman resonance should have taken place for meteotsunamis to occur, range between 40 and 50 meters (corresponding to meteotsunami celerity of ~\u0026thinsp;19\u0026ndash;22 m/s). In Croatia the average depth around the island of Lošinj (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef) ranges between 40 and 50 meters (corresponding to meteotsunami celerity of ~\u0026thinsp;19 to 22 m/s), while the waters surrounding Hvar are deeper (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eg), averaging 60 to 70 meters (corresponding to meteotsunami celerity of ~\u0026thinsp;24 to 26 m/s). During the Adriatic meteotsunami events, storm clouds have been recorded moving at speeds of 17 to 19 m/s on July 19, 2023, and 15 to 17 m/s on July 22, 2023, aligning with the long-wave phase speed in the respective water depths, thus sustaining Proudman resonance and amplifying the waves.\u003c/p\u003e \u003cdiv id=\"Sec7\" class=\"Section3\"\u003e \u003ch2\u003e3.2.1 The East Coast meteotsunami of June 13, 2013 (U.S.)\u003c/h2\u003e \u003cp\u003eFor the East Coast of the U.S. location, meteotsunami of June 13, 2013 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea), (hitting Atlantic City at 15:02 UTC) was simulated. This modeling was the extension of the model described by Titov and Moore (\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). The goal of this modeling effort was to test the multi-grid forecast approach for meteotsunami simulations. While the propagation results followed the previously published procedure, the simulation was extended to simulate areas around tide-gages with much higher resolution model that includes the flooding forecast capability, similar to the NOAA tsunami forecast tools (Titov et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOn June 12\u0026ndash;13, 2013, a powerful derecho developed from severe thunderstorms initially forming over Iowa and Illinois (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, IA and IL). The storms began as discrete supercells in a highly unstable atmospheric environment, characterized by high CAPE values ranging from approximately 2000 to 4800 J/kg (Hersbach et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). These conditions, combined with strong mid-level wind shear and an upper-level atmospheric disturbance interacting with the warm and moist air mass over the northern Great Plains, led to rapid storm intensification and organization. The individual supercells merged into a well-organized squall line as they entered Indiana (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, IN), where embedded downbursts produced damaging winds with speeds up to approximately 45 m/s, causing widespread structural and vegetation damage. The derecho maintained its strength overnight, benefiting from persistent atmospheric instability, favorable wind shear, and near-unidirectional mid-tropospheric flow. As the system reached Ohio (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, OH), it produced further extensive wind damage and spawned several embedded tornadoes. By early morning on June 13, as the derecho moved across the eastern states and onto the Atlantic coastline, the associated rapid pressure disturbances generated significant sea-level oscillations recorded by some coastal tide gauges along the U.S. East Coast (e.g., at Lewes, Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe generation, propagation and coast impact of the meteotsunami were modeled along the segment of the coast from 35\u0026deg; N to 41.6\u0026deg; N and 77\u0026deg; W to 70.4\u0026deg; W (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The propagation model utilized bathymetry with a resolution of 9 arc seconds (approximately 270 meters) and, a time step of 1 second. The atmospheric pressure forcing was applied for the first 2.8 hours of the simulation using the sequence of air pressure fields estimated from radar reflectivity data. The radar images correspond to time when the pressure disturbance was propagating over the continental shelf. The propagation simulation was used as the forcing function for the high-resolution models setup around tide gauges at Montauk, Atlantic City and Lewes. The models used three nested grids that zoom into the highest spatial grid resolutions of 2 arc seconds (less than 60 meters), where moving boundary conditions are applied to allow for predicting tsunami inundation estimate. The air pressure forcing was applied throughout the simulation for all numerical grids, to ensure the consistent dynamic tsunami generation.\u003c/p\u003e \u003cp\u003eIn spite of the fact that a maximum high-frequency atmospheric pressure change of 4.7 hPa per 18 min was measured at the Atlantic City airport (Rabinovich \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), no significant sea-level oscillations were measured at the Atlantic City tide gauge station at time of atmospheric disturbance passage (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Similar holds for Montauk (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The model reproduced the height and period of initial sea level oscillations satisfactory at the three East Coast stations (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e; time from 0 to 3 hours). Lewes is situated at the southern side of the entrance to Delaware Bay. Since Delaware Bay is long and relatively shallow, the pressure disturbance had traveled for around an hour along the coast before reaching Lewes. Thus, it had enough time to affect the sea and generate a long ocean wave, as the one observed and modelled in Lewes (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). On the contrary, the atmospheric disturbance came to Atlantic City from the land, and was not able to generate ocean wave on its passage over the station. However, as already shown by Titov and Moore (\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) and confirmed by our modelling results, the atmospheric pressure front continued travelling offshore (towards the east) across the continental shelf, and then over deeper waters of the Atlantic Ocean. While it traveled over the shelf, it was coupled to a long ocean wave which grew stronger through Proudman resonance. Upon reaching the shelf break, meteotsunami waves decoupled from the atmospheric forcing (which continued to propagate eastward) and started propagating as free waves. Because long ocean wave speeds (celerity) differ significantly between shallow continental shelf waters and the adjacent deeper ocean, the meteotsunami waves mostly refracted and turned back toward the U.S. East coast. As a result, most of the meteotsunami energy remained concentrated over the wide and gently sloping U.S. East Coast shelf\u0026mdash;a region well-known to trap and amplify tsunami energy (Mofjeld et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2001\u003c/span\u003e). It was this reflected wave that reached considerable heights along the U. S: East Coast, including at Atlantic City and Montauk, as is both seen in measurements, and successfully reproduced by our model (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe results of our high-resolution modeling of the June 13, 2013 meteotsunami showed a better comparison with observations in comparison with the coarse propagation model (see Supplemental Material). These model results confirm that the multi-grid approach to meteotsunami simulation improve forecast accuracy and provides method for predicting potential flooding from meteotsunamis. To test the method further, we applied the same approach for several other meteotsunami events at different geographic settings.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section3\"\u003e \u003ch2\u003e3.2.2 The Klaipeda meteotsunami of June 20, 2022, Lithuania\u003c/h2\u003e \u003cp\u003eThe meteotsunami propagation throughout the SE Baltic Sea toward Klaipėda was modeled along the segment of the coast from 53\u0026deg; N to 59\u0026deg; N and 11\u0026deg; E to 31\u0026deg; E (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), resolution was 13 arc seconds (approximately 400 meters), and time step was set to 0.3 seconds. We used 13 hours of the atmospheric pressure forcing, which means that radar reflectivity data, or direct atmospheric pressure forcing, was used throughout the simulation. Over 8 hours of meteotsunami dynamics was simulated following the first detection of cumulonimbus formation at 6:00 UTC over the western part of Lithuania (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb). The area around the Klaipėda port and the tide gauge was modeled with the multi-gird high-resolution inundation model. The highest spatial resolution of the model around the tide gauge location was 0.3 arc second (around 10 meters). The model was able to reproduce fairly small signal of June 20, 2022, meteotsunami event.\u003c/p\u003e \u003cp\u003eThe modeled event unfolded as follows: On 20 June 2020, high convective clouds began forming over western Lithuania and gradually moved northwest toward Klaipėda. As convection intensified, the first light rain shower was observed in Klaipėda at 07:50 UTC, accompanied by a weak atmospheric pressure disturbance and the onset of long ocean wave activity\u0026mdash;both detected in measurements and replicated in the model (see Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e Klaipėda, ~\u0026thinsp;2 hours into the simulation). Between 2.5 and 5 hours into the simulation, no significant atmospheric forcing was detected; however, sea level fluctuations persisted in the Port of Klaipėda, suggesting that seiches were likely triggered by the initial disturbance or incoming long waves.\u003c/p\u003e \u003cp\u003eAround 10:50 UTC, intense cumulonimbus clouds reached Klaipėda, leading to renewed amplification of long ocean waves\u0026mdash;again evident in both observations and model output (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e Klaipėda, after 5 hours). The Klaipėda meteorological station recorded thunder and heavy rainfall, indicating strong convective activity. At the same time, the Convective Available Potential Energy (CAPE) in western Lithuania and Latvia reached 1400\u0026ndash;1800 J/kg. For reference, CAPE values exceeding 1000 J/kg are typically sufficient to support the development of strong to severe storms (Pešice et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Garner \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Punkka and Bister \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). It can be observed that in the presence of stronger atmospheric disturbances, the model results are more accurate, successfully capturing both the period and the amplitudes of the oscillations, if not the exact shape (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section3\"\u003e \u003ch2\u003e3.2.3 The Adriatic Sea meteotsunamis of July 20\u0026ndash;22, 2023\u003c/h2\u003e \u003cp\u003eThe 19 July 2023 storm over the southeastern Europe was a textbook example of a highly organized convective system (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb), driven by a combination of strong atmospheric instability, mesoscale forcing, and upper-level dynamics. Its formation was associated with a pronounced upper-level trough moving eastward, providing the necessary dynamic lifting for deep convection (Mahovic et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). At the surface, a low-pressure system over northern Italy and the northern Adriatic Sea generated a strong pressure gradient, enhancing low-level convergence and moisture advection from the Mediterranean.\u003c/p\u003e \u003cp\u003eDuring a period of extreme heat affecting southern Europe, several atmospheric factors contributed to the storm\u0026rsquo;s development. Strong upper-level westerly winds, with a jet streak positioned over Central Europe, brought cooler air aloft from northern Europe, creating steep lapse rates (Mahovic et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). At the same time, the low layers remained very humid, particularly around the Po Valley, creating an environment highly conducive to severe storm formation over multiple days in mid-July.\u003c/p\u003e \u003cp\u003eOne of the key drivers of the storm\u0026rsquo;s rapid intensification was high Convective Available Potential Energy (CAPE), exceeding 2000\u0026ndash;5000 J/kg in some areas (ERA5 Reanalysis data Climate Data Store; Hersbach et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). This, combined with strong wind shear (with mid-level wind shear exceeding 20\u0026ndash;25 m/s), facilitated the development of an intense mesoscale convective system (MCS) (Mahovic et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The storm\u0026rsquo;s extreme wind speeds, localized downbursts, and sustained pressure fluctuations were direct consequences of deep convection and rapid vertical momentum transport. As the system moved eastward into Croatia, Slovenia, and Bosnia and Herzegovina, it caused widespread damage due to the combined effects of intense winds and heavy precipitation. However, this initial storm was not the direct trigger of the meteotsunamis. Instead, it was the second storm, which developed at night between 19 and 20 July, that played a crucial role in meteotsunami generation. Unlike the first one, this storm system moved not only over the land but also over the Adriatic Sea, propagating from the northwest to the southeast, along the Adriatic Sea main axis, and generating long ocean waves through Proudman resonance. Measured sea level oscillations at Mali Lošinj reached amplitudes of 67.75 cm.\u003c/p\u003e \u003cp\u003eThree days later, on July 22, CAPE was again high over the Adriatic Sea (1500\u0026ndash;3000 J/kg) (ERA5 Reanalysis data Climate Data Store; Hersbach et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). A similar weather pattern was unfolding, with the Mediterranean region experiencing a prolonged heatwave, a moisture-laden atmosphere, and a developing cyclonic cold front over the northern Italy, Switzerland, and Austria. As this system moved eastward, it triggered the formation of convective clouds.\u003c/p\u003e \u003cp\u003eBy noon of July 22, the first rain clouds emerged over the northern Adriatic, and by 12:20 UTC, a thunderstorm was detected at Venice Airport. As this storm cloud progressed over the Adriatic Sea (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec), it impacted the sea, generating a long wave with a peak amplitude of 44.95 cm in Mali Lošinj, and 26.31 cm in Stari Grad. It should be noted that Mali Lošinj and, in particular, Stari Grad tide gauges are not located at top of harbors where strongest meteotsunami oscillations occur but at relatively deep locations where lower sea level oscillations are observed during meteotsunamis (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef-g).\u003c/p\u003e \u003cp\u003eThe propagation of the Adriatic Sea meteotsunamis were modeled along the segment of the coast from 40\u0026deg; N to 47\u0026deg; N and 12\u0026deg; E to 20\u0026deg; E (encompassing both Mali Lošinj and Stari Grad) with the spacial resolution of 13 arc seconds (approximately 400 meters), with model time step of 0.3 seconds. High-resolution models were setup for the locations of Mali Lošinj and Stari Grad. These models used the propagation simulations as boundary input, with continuing air pressure forcing applied throughout all model grids. The model prediction for the tide gauge were obtained from the model grid with the highest model resolution of 1 arc second (about 30 meters).\u003c/p\u003e \u003cp\u003eFor the July 20, 2023 model (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec), forcing was initiated as soon as the first precipitation appeared in radar images over the northern Adriatic Sea at 19:10 UTC and the forcing continued until 03:40 UTC on the 20th. Model of the July 22, 2023 event used forcing from 12:00 UTC to 19:00 UTC. Over 8 h of meteotsunami dynamics were simulated.\u003c/p\u003e \u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e reveals that our model has successfully captured the July 20 event in Mali Lošinj. Onset of oscillations, their maximum height, and period are all well modelled. The July 22 event was modelled slightly less satisfactory. In Mali Lošinj, amplitude and period of oscillations are well reproduced, however the initial modelled meteotsunami wave lags for ~\u0026thinsp;20 minutes in comparison with the measured one. In Stari Grad, amplitude and period are again well reproduced, but there is now a\u0026thinsp;~\u0026thinsp;30 min lag of modelled wave in comparison to the measured one. Nonetheless, we find the results satisfactory, and suggestive of the idea that real-time modelling based on radar reflectivity images, could be used to forecast Adriatic meteotsunami events. Mali Lošinj Bay has a peculiar shape, a 5.4 km long bay is connected to the open sea with a 466-meter-wide strait which is perpendicular to the bay. This narrow entrance strait restricts the entry of the meteotsunamis from the open sea, delaying their penetration into the bay. However, once that the waves enter the bay, the bay's elongated and confined shape enhances harbor resonance, amplifying the meteotsunami wave. As a result, the wave is observed with greater amplitudes after the passage of the atmospheric front than at the time of its passage. In contrast, Stari Grad is situated in a bay which is connected to the open sea with a wide mouth, allowing meteotsunamis to be observed immediately as the front moves over the region, without significant delay.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eOur study examines a meteotsunami forecast approach, which combines weather radar measurements with numerical ocean modelling. We tested the applicability of the MOST tsunami forecasting model in predicting meteotsunamis across various global regions, using weather radar data as the primary source of information on atmospheric forcing. We tested the method at three locations (the East Coast of the U.S., Baltic and the Adriatic Seas) and, at each of these locations, achieved a high success rate in timely modelling of meteotsunamis.\u003c/p\u003e \u003cp\u003eThe East Coast event has been examined before by Titov and Moore (\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) with similar approach, but without the high-resolution modeling of area close to Jtide gauges. In this work we expanded the methodology into use of the nested tsunami modeling approach to examine feasibility and accuracy of the meteotsunami forecast including the possible inundation predictions. We documented clear improvement of accuracy of this high-resolution approach in comparison with a simpler propagation modeling with coarse grids. In addition, we demonstrated the feasibility of using the forecast system that was developed for the seismic tsunami inundation forecast (Titov et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) in application to meteotsunamis.\u003c/p\u003e \u003cp\u003eTo further evaluate the performance of the meteotsunami forecast method, we examined two additional representative events\u0026mdash;Klaipėda and Stari Grad\u0026mdash;where modeled pressure fields were compared with observations (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). In the Klaipėda case (June 20, 2020), a noticeable temporal mismatch is observed between the modeled and recorded atmospheric pressure fields: the model captures the pressure disturbance slightly earlier than the observations (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). This discrepancy is likely due to the low temporal resolution of the measured atmospheric pressure data at Klaipėda, which were available only at hourly intervals, likely missing the precise timing and structure of the pressure anomaly.\u003c/p\u003e \u003cp\u003eDespite this, the modeled sea level response aligns well with the observed timing of the water level fluctuations, indicating that the model successfully simulates the generation and propagation of long waves triggered by the atmospheric disturbance. The ability of the model to reproduce the observed sea level variation\u0026mdash;even with a slightly misaligned pressure input\u0026mdash;demonstrates its robustness. However, this case also highlights the importance of high-resolution atmospheric pressure data for more accurately constraining the timing and intensity of meteotsunami-generating disturbances. The current lack of fine-scale surface pressure observations at many coastal locations limits our ability to fully capture these events in real time. In this context, weather radar data provides a valuable complementary solution, offering higher temporal and spatial resolution that can significantly enhance meteotsunami detection and modeling efforts.\u003c/p\u003e \u003cp\u003eIn the Stari Grad case (July 22, 2023), the modeled and observed atmospheric pressure fields exhibit excellent agreement, both in timing and amplitude. The pressure disturbance is well represented in the model, and its arrival coincides precisely with the onset of long-wave sea level oscillations recorded at the tide gauge (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). This strong correspondence confirms that the model effectively captures the coupling between atmospheric forcing and the resulting sea level response.\u003c/p\u003e \u003cp\u003eHowever, the first wave observed in the tide gauge record is not present in the model output. This suggests that the initial meteotsunami wave was generated earlier in the Adriatic Sea, as the MCS propagated from the northwest to the southeast. In contrast, the second wave\u0026mdash;successfully captured by the model \u0026mdash; was likely generated locally, within the open bay near Stari Grad, as the pressure disturbance passed directly over the area. The effects of meteotsunamis are most pronounced in shallow waters and narrow bays (located at the end of wide shelves), where further wave amplification due to coastal geometry and bathymetry can occur. Therefore, to predict meteotsunamis in shallow waters, issue timely warnings to tourists and port authorities, and ensure forecast accuracy, it is essential to start the meteotsunamis model as soon as possible, and optimally before the weather system comes near the endangered coast. This approach helps capture the initial signal and enhances the precision of wave activity predictions.\u003c/p\u003e \u003cp\u003eThe study highlights the potential of using weather radar data to forecast meteotsunamis in real-time especially in situations where a sudden change of atmospheric pressure, driven by MCSs, is expected. Meteotsunami modeling can be initiated upon detecting a developing thunderstorm or heavy rainfall, allowing immediate adaptation in regions such as the eastern US coast, ports located in narrow bays, and tourist hotspots like the Adriatic Sea. Predicting meteotsunamis associated with storm surges may require incorporating a wind component. The current model results include only atmospheric pressure forcing. While they capture some of the observed wave activity, significant discrepancies remain \u0026mdash; especially during the initial and peak phases of the event. This is because cold-season meteotsunamis are linked to cyclonic activity, where both atmospheric pressure jumps and wind intensification play critical roles (Monserrat et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Šepić and Vilibić \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Pellikka et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Nesteckytė et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Consequently, the current model configuration lacks the necessary components to fully reproduce these events. To accurately simulate cold-season meteotsunamis, wind stress or direct wind forcing must be incorporated into the modeling framework.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"5. Conclusions","content":"\u003cp\u003eThis study demonstrates the potential of using weather radar reflectivity data as a proxy for atmospheric pressure anomalies in meteotsunami forecasting. The application of the MOST tsunami propagation model, combined with real-time radar data, offers a practical and adaptable approach for predicting meteotsunami events across various geographic regions, including the East Coast of the United States, the southeastern Baltic Sea, and the Adriatic Sea. The results further highlight the importance of Proudman resonance in wave amplification and underscore the critical role of atmospheric pressure fluctuations in generating these hazardous events.\u003c/p\u003e \u003cp\u003eRadar-based methods present several key advantages: (1) the data are available in near-real time; (2) radar image sequences provide direct information on the speed, direction, and affected area of the moving pressure anomaly; and (3) the spatial and temporal resolution is sufficient to be used directly in tsunami generation and propagation models\u0026mdash;removing the need for complex interpolation, as is often required with sparse pressure or tide gauge measurements. Notably, real-time radar data are especially effective in tracking tsunamigenic atmospheric disturbances associated with convective systems rich in hydrometeors.\u003c/p\u003e \u003cp\u003eAdditional findings indicate that high-resolution models substantially improve meteotsunami simulation accuracy, with pseudo-R-squared values confirming better predictive performance. The model also proved sensitive enough to detect sea level oscillations triggered by even weak atmospheric disturbances, related to light rain showers, underscoring its responsiveness. Finally, the study emphasizes the importance of initiating model simulations before the weather system reaches the target location to enhance the precision of wave activity forecasts.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgement:\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe sincerely thank Lukas Sudvajus (LHMT) for his support, expertise, and data contribution, and Hrvoje Mihanović (Institute of Oceanography and Fisheries, Split, Croatia) for providing essential data. Their assistance has been invaluable to this research. Croatia radar data Data owner: Croatian Meteorological and Hydrological Service\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthical Approval\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eLaura Nesteckytė performed the modeling and data analysis and wrote the initial manuscript draft. Vasily Titov guided the modeling approach and contributed to manuscript writing and editing. Jadranka \u0026Scaron;epić analyzed atmospheric conditions and contributed to writing and editing. Loreta Kalp\u0026scaron;aitė-Rimkienė supported the study design and manuscript revision.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eLoreta Kelp\u0026scaron;aitė-Rimkienė and Laura Nesteckytė were partly supported by the WaveWise project received funding from the Research Council of Lithuania (LMTLT), agreement No SMIP-24-140. Jadranka \u0026Scaron;epić was supported by the ERC StG 853045 SHExtreme Grant, the ERC PoC 101213756 MeD-Track, the Croatian Science Foundation projects IP-2019-04-5875 StVar-Adri and IP-2022-10-4144 CroClimExtremes and STIM\u0026ndash;REI project, Contract Number: KK.01.1.1.01.0003, a project funded by the European Union through the European Regional Development Fund \u0026ndash; the Operational Programme Competitiveness and Cohesion 2014-2020 (KK.01.1.1.01). Measurement of sea level and atmospheric pressure in the Adriatic Sea was supported through the research project KLIMADRIA, funded by the European Union - NextGenerationEU.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe sea level, sea surface atmospheric pressure, wind speed, and direction data from The Lithuanian Hydrometeorological Service can be obtained on request from the Lithuanian Hydrometeorological Service. \u003cstrong\u003e\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eAngove M, Vilibić I, Heidarzadeh M, Monserrat S, Okal E, Rabinovich AB, \u0026Scaron;epić J, Titov VV (2025) Meteotsunamis: definition, detection and alerting services investigation. UNESCO Publication. IOC Technical Series, IOC/2025/TS/200, 31 pp. Accessible at: https://unesdoc.unesco.org/ark:/48223/pf0000392653.locale=fr\u003c/li\u003e\n \u003cli\u003eBailey K, DiVeglio C, Welty A (2014) An Examination of the June 2013 East Coast Meteotsunami Captured By NOAA Observing Systems. NOAA, Silver Spring, Maryland\u003c/li\u003e\n \u003cli\u003eBechle AJ, Wu CH (2014) An entropy‐based surface velocity method for estuarine discharge measurement. Water Resources Research 50:6106\u0026ndash;6128. https://doi.org/10.1002/2014WR015353\u003c/li\u003e\n \u003cli\u003eBelu\u0026scaron;ić D, Strelec Mahović N (2007) Detecting and following atmospheric disturbances with a potential to generate meteotsunamis in the Adriatic. Physics and Chemistry of the Earth, Parts A/B/C 34:918\u0026ndash;927. https://doi.org/10.1016/j.pce.2009.08.009\u003c/li\u003e\n \u003cli\u003eBoslough M, Titov V (2024) Air-coupled tsunamis generated from impacts and airbursts: Our understanding before Hunga-Tonga Hunga-Ha\u0026rsquo;apai. Acta Astronautica 222:641\u0026ndash;646. https://doi.org/10.1016/j.actaastro.2024.06.041\u003c/li\u003e\n \u003cli\u003eChurchill DD, Houston SH, Bond NA (1995) The Daytona Beach wave of 3\u0026ndash;4 July 1992: a shallow water gravity wave forced by a propagating squall line.\u003c/li\u003e\n \u003cli\u003eCroatian National Tourist Board Visit Mali La\u0026scaron;inj. https://visitmali-losinj.croatia.hr/en-gb. Accessed 17 Mar 2025\u003c/li\u003e\n \u003cli\u003eDalmatia Express Stari Grad, Croatia - a Complete Guide. https://dalmatia-express.com/blog/stari-grad-complete-guide/. Accessed 17 Mar 2025\u003c/li\u003e\n \u003cli\u003eDenamiel C, \u0026Scaron;epić J, Huan X, et al (2019a) Stochastic Surrogate Model for Meteotsunami Early Warning System in the Eastern Adriatic Sea. JGR Oceans 124:8485\u0026ndash;8499. https://doi.org/10.1029/2019JC015574\u003c/li\u003e\n \u003cli\u003eDenamiel C, \u0026Scaron;epić J, Ivankovic D, Vilibić I (2019b) The Adriatic Sea and Coast modelling suite: Evaluation of the meteotsunami forecast component. Ocean Modelling 135:71\u0026ndash;93. https://doi.org/10.1016/j.ocemod.2019.02.003\u003c/li\u003e\n \u003cli\u003eDusek G, DiVeglio C, Licate L, et al (2019) A Meteotsunami Climatology along the U.S. East Coast. Bulletin of the American Meteorological Society 100:1329\u0026ndash;1345. https://doi.org/10.1175/BAMS-D-18-0206.1\u003c/li\u003e\n \u003cli\u003eEuropean Centre for Medium-Range Weather Forecasts ECMWF. https://www.ecmwf.int/. Accessed 15 Mar 2025\u003c/li\u003e\n \u003cli\u003eGarner S (2015) The Relationship between Hurricane Potential Intensity and CAPE. Journal of the Atmospheric Sciences 72:141\u0026ndash;163. https://doi.org/10.1175/JAS-D-14-0008.1\u003c/li\u003e\n \u003cli\u003eGreenspan M (1956) Propagation of sound in five monatomic gases. The Journal of the AcousticalSociety of America 28:644\u0026ndash;648. https://doi.org/10.1121/1.1908432\u003c/li\u003e\n \u003cli\u003eHa T, Yoon J-S, Heo K-Y (2018) Numerical Simulations of a Meteotsunami using both Atmospheric and Phase-resolving Wave Models in the Yellow Sea. Journal of Coastal Research 85:786\u0026ndash;790. https://doi.org/10.2112/SI85-158.1\u003c/li\u003e\n \u003cli\u003eHersbach H, Bell B, Berrisford P, et al (2020) The ERA5 global reanalysis. Quart J Royal Meteoro Soc 146:1999\u0026ndash;2049. https://doi.org/10.1002/qj.3803\u003c/li\u003e\n \u003cli\u003eKelp\u0026scaron;aitė-Rimkienė L, Soomere T, Bagdanavičiūtė I, et al (2018) Measurements of Long Waves in Port of Klaipėda, Lithuania. Journal of Coastal Research 85:761\u0026ndash;765. https://doi.org/10.2112/SI85-153.1\u003c/li\u003e\n \u003cli\u003eMahovic NS, Smiljanic I, Kerkmann J (2023) Destructive convective storms in SE Europe. In: Eumetsat. https://user.eumetsat.int/resources/case-studies/destructive-convective-storms-in-se-europe. Accessed 3 Mar 2025\u003c/li\u003e\n \u003cli\u003eMofjeld HO, Titov VV, Gonz\u0026aacute;lez FI, Newman JC (2001) Tsunami scattering provinces in the Pacific Ocean. Geophysical Research Letters 28:335\u0026ndash;337. https://doi.org/10.1029/2000GL011710\u003c/li\u003e\n \u003cli\u003eMonserrat S, Vilibić I, Rabinovich AB (2006) Meteotsunamis: atmospherically induced destructive ocean waves in the tsunami frequency band. Nat Hazards Earth Syst Sci 6:1035\u0026ndash;1051. https://doi.org/10.5194/nhess-6-1035-2006\u003c/li\u003e\n \u003cli\u003eMourre B, Santana A, Buils A, et al (2021) On the potential of ensemble forecasting for the prediction of meteotsunamis in the Balearic Islands: sensitivity to atmospheric model parameterizations. Nat Hazards 106:1315\u0026ndash;1336. https://doi.org/10.1007/s11069-020-03908-x\u003c/li\u003e\n \u003cli\u003eNational Weather Service When the National Weather Service is Watching Radar... In: National Weather Service. https://www.weather.gov/bmx/radar_aboutnwsradar. Accessed 25 Nov 2024\u003c/li\u003e\n \u003cli\u003eNesteckytė L, Kelp\u0026scaron;aitė-Rimkienė L, Rabinovich AB (2023) Hazardous meteotsunami-like sea-level oscillations in the Port of Klaipeda, the Baltic Sea. Nat Hazards. https://doi.org/10.1007/s11069-023-06311-4\u003c/li\u003e\n \u003cli\u003eNesteckytė L, Stankūnavičius G, Kelp\u0026scaron;aitė-Rimkienė L, \u0026Scaron;epić J (2024) 2023 November Storm in The Port of Klaipeda; Why was The Port Closed? Pure Appl Geophys. https://doi.org/10.1007/s00024-024-03596-9\u003c/li\u003e\n \u003cli\u003eOrlić M (1980) About a possible occurrence of the Proudman resonancein the Adriatic. Thalassia Jugosl\u003c/li\u003e\n \u003cli\u003eOrlić M (2015) The first attempt at cataloguing tsunami-like waves of meteorological origin in Croatian coastal waters\u003c/li\u003e\n \u003cli\u003eOrlić M, Belu\u0026scaron;ić D, Janeković I, Pasarić M (2010) Fresh evidence relating the great Adriatic surge of 21 June 1978 to mesoscale atmospheric forcing. J Geophys Res 115:2009JC005777. https://doi.org/10.1029/2009JC005777\u003c/li\u003e\n \u003cli\u003ePasquet S, Vilibić I (2013) Shelf edge reflection of atmospherically generated long ocean waves along the central U.S. East Coast. Continental Shelf Research 1\u0026ndash;8. https://doi.org/10.1016/j.csr.2013.06.007\u003c/li\u003e\n \u003cli\u003ePattiaratchi CB, Wijeratne EMS (2015) Are meteotsunamis an underrated hazard? Phil Trans R Soc A 373:20140377. https://doi.org/10.1098/rsta.2014.0377\u003c/li\u003e\n \u003cli\u003ePellikka H (2020) Dark-blue horizon: Sea level rise and meteotsunamis on the Finnish coast. Finnish Meteorological Institute\u003c/li\u003e\n \u003cli\u003ePellikka H, Laurila TK, Boman H, et al (2020) Meteotsunami occurrence in the Gulf of Finland over the past century. Nat Hazards Earth Syst Sci 20:2535\u0026ndash;2546. https://doi.org/10.5194/nhess-20-2535-2020\u003c/li\u003e\n \u003cli\u003ePellikka H, \u0026Scaron;epić J, Lehtonen I, Vilibić I (2022) Meteotsunamis in the northern Baltic Sea and their relation to synoptic patterns. Weather and Climate Extremes 38:100527. https://doi.org/10.1016/j.wace.2022.100527\u003c/li\u003e\n \u003cli\u003ePe\u0026scaron;ice P, Sulan J, Řez\u0026aacute;čov\u0026aacute; D (2003) Convection precursors in the Czech territory. Atmospheric Research 67\u0026ndash;68:523\u0026ndash;532. https://doi.org/10.1016/S0169-8095(03)00070-X\u003c/li\u003e\n \u003cli\u003ePiotrowski A, Szczuciński W, Sydor P, et al (2017) Sedimentary evidence of extreme storm surge or tsunami events in the southern Baltic Sea (Rogowo area, NW Poland). GQ. https://doi.org/10.7306/gq.1385\u003c/li\u003e\n \u003cli\u003eProudman J (1929) The Effects on the Sea of Changes in Atmospheric Pressure. Geophysical Journal International 2:197\u0026ndash;209. https://doi.org/10.1111/j.1365-246X.1929.tb05408.x\u003c/li\u003e\n \u003cli\u003ePunkka A-J, Bister M (2015) Mesoscale Convective Systems and Their Synoptic-Scale Environment in Finland. Weather and Forecasting 30:182\u0026ndash;196. https://doi.org/10.1175/WAF-D-13-00146.1\u003c/li\u003e\n \u003cli\u003eRabinovich A (2020) Twenty-seven years of progress in the science of meteorological tsunamis. Pure Appl Geophys\u003c/li\u003e\n \u003cli\u003eRabinovich AB (2009) Seiches and Harbor Oscillations. In: Handbook of Coastal and Ocean Engineering. World Scientific, p 1192\u003c/li\u003e\n \u003cli\u003eRabinovich AB, \u0026Scaron;epić J, Thomson RE (2021) The meteorological tsunami of 1 November 2010 in the southern Strait of Georgia: a case study. Nat Hazards 106:1503\u0026ndash;1544. https://doi.org/10.1007/s11069-020-04203-5\u003c/li\u003e\n \u003cli\u003eRauhala J, Kahma KK, Stipa T, et al (2014) Recent observations of meteotsunamis on the Finnish coast. Nat Hazards 74:197\u0026ndash;215. https://doi.org/10.1007/s11069-014-1150-3\u003c/li\u003e\n \u003cli\u003eRenquist H (1926) Ein Seeb\u0026auml;r in Finnland. Zur Frage Nach der Entstehung der Seeb\u0026auml;ren, Geogr Ann 230\u0026ndash;236\u003c/li\u003e\n \u003cli\u003eRuić K, \u0026Scaron;epić J, Mlinar M, Međugorac I (2023) Contribution of high-frequency (T \u0026lt; 2 h) sea level oscillations to the Adriatic sea level maxima. Nat Hazards 116:3747\u0026ndash;3777. https://doi.org/10.1007/s11069-023-05834-0\u003c/li\u003e\n \u003cli\u003eSallenger AH, List JH, Gelfenbaum G, et al (1995) Large Wave at Daytona Beach, Florida, Explained as a Squall-line Surge. Journal of Coastal Research 11:1383\u0026ndash;1388. https://www.jstor.org/stable/4298439\u003c/li\u003e\n \u003cli\u003e\u0026Scaron;epić J, Metlicic N, Orlic M (2025) Synoptic conditions corresponding to the Adriatic meteorological tsunamis. https://meetingorganizer.copernicus.org/EGU24/EGU24-14855.html. Accessed 9 Apr 2025\u003c/li\u003e\n \u003cli\u003e\u0026Scaron;epić J, Vilibić I (2011) The development and implementation of a real-time meteotsunami warning network for the Adriatic Sea. Nat Hazards Earth Syst Sci 11:83\u0026ndash;91. https://doi.org/10.5194/nhess-11-83-2011\u003c/li\u003e\n \u003cli\u003e\u0026Scaron;epić J, Vilibić I, Belu\u0026scaron;ić D (2009) Source of the 2007 Ist meteotsunami (Adriatic Sea). J Geophys Res 114:C03016. https://doi.org/10.1029/2008JC005092\u003c/li\u003e\n \u003cli\u003e\u0026Scaron;epić J, Vilibić I, Monserrat S (2016) Quantifying the probability of meteotsunami occurrence from synoptic atmospheric patterns: Meteotsunami Index. Geophys Res Lett 43:10,377-10,384. https://doi.org/10.1002/2016GL070754\u003c/li\u003e\n \u003cli\u003e\u0026Scaron;epić J, Vilibić I, Rabinovich AB, Monserrat S (2015) Widespread tsunami-like waves of 23-27 June in the Mediterranean and Black Seas generated by high-altitude atmospheric forcing. Sci Rep 5:11682. https://doi.org/10.1038/srep11682\u003c/li\u003e\n \u003cli\u003eThe University Corporation for Atmospheric Research (2012) Weather Radar Fundamentals. https://research.atmos.ucla.edu/weather/C110/Documents/tmp/basic_wxradar/index.htm. Accessed 8 Jan 2025\u003c/li\u003e\n \u003cli\u003eTitov V, Gonz\u0026aacute;lez F (2000) Numerical study of the source of the 17 July 1998 PNG tsunami. https://doi.org/10.1007/978-94-017-3618-3_14\u003c/li\u003e\n \u003cli\u003eTitov V, K\u0026acirc;noğlu U, Synolakis C (2016) Development of MOST for Real-Time Tsunami Forecasting. J Waterway, Port, Coastal, Ocean Eng 142:03116004. https://doi.org/10.1061/(ASCE)WW.1943-5460.0000357\u003c/li\u003e\n \u003cli\u003eTitov V, Moore C (2021) Meteotsunami model forecast: can coastal hazard be quantified in real time? Nat Hazards 106:1545\u0026ndash;1561. https://doi.org/10.1007/s11069-020-04450-6\u003c/li\u003e\n \u003cli\u003eTitov VV, Moore CW, Greenslade DJM, et al (2011) A New Tool for Inundation Modeling: Community Modeling Interface for Tsunamis (ComMIT). Pure Appl Geophys 168:2121\u0026ndash;2131. https://doi.org/10.1007/s00024-011-0292-4\u003c/li\u003e\n \u003cli\u003eUSA-Reisen.de (2025) Atlantic City Stadt, Strand \u0026amp; Spa\u0026szlig;! Accessed 25 Nov 2024\u003c/li\u003e\n \u003cli\u003eVilibić I (2008) Numerical simulations of the Proudman resonance. Continental Shelf Research 28:574\u0026ndash;581. https://doi.org/10.1016/j.csr.2007.11.005\u003c/li\u003e\n \u003cli\u003eVilibić I, Domijan N, Orlić M, et al (2004) Resonant coupling of a traveling air pressure disturbance with the east Adriatic coastal waters. J Geophys Res 109:2004JC002279. https://doi.org/10.1029/2004JC002279\u003c/li\u003e\n \u003cli\u003eVilibić I, Horvath K, Strelec Mahović N, et al (2014) Atmospheric processes responsible for generation of the 2008 Boothbay meteotsunami. Nat Hazards 74:25\u0026ndash;53. https://doi.org/10.1007/s11069-013-0811-y\u003c/li\u003e\n \u003cli\u003eVilibić I, \u0026Scaron;epić J, Rabinovich AB, Monserrat S (2016) Modern Approaches in Meteotsunami Research and Early Warning. Front Mar Sci 3:. https://doi.org/10.3389/fmars.2016.00057\u003c/li\u003e\n \u003cli\u003eVillalonga J, Monserrat S, Gomis D, Jord\u0026agrave; G (2024) Observational Characterization of Atmospheric Disturbances Generating Meteotsunamis in the Balearic Islands. JGR Oceans 129:e2024JC020910. https://doi.org/10.1029/2024JC020910\u003c/li\u003e\n \u003cli\u003eWertman CA, Yablonsky RM, Shen Y, et al (2014) Mesoscale convective system surface pressure anomalies responsible for meteotsunamis along the U.S. East Coast on June 13th, 2013. Sci Rep 4:7143. https://doi.org/10.1038/srep07143\u003c/li\u003e\n \u003cli\u003eWhitmore P, Knight B (2014) Meteotsunami forecasting: sensitivities demonstrated by the 2008 Boothbay, Maine, event. Nat Hazards 74:11\u0026ndash;23. https://doi.org/10.1007/s11069-014-1056-0\u003c/li\u003e\n \u003cli\u003eWijeratne EMS, Pattiaratchi CB (2024) Meteotsunamis Generated by Thunderstorms. JGR Oceans 129:e2023JC020662. https://doi.org/10.1029/2023JC020662\u003c/li\u003e\n \u003cli\u003eYano J-I, Ziemiański MZ, Cullen M, et al (2018) Scientific Challenges of Convective-Scale Numerical Weather Prediction. Bulletin of the American Meteorological Society 99:699\u0026ndash;710. https://doi.org/10.1175/BAMS-D-17-0125.1\u003c/li\u003e\n \u003cli\u003ePort of Klaipėda. https://www.portofklaipeda.lt/. Accessed 25 Sep 2024\u003c/li\u003e\n \u003cli\u003eERA5 Reanalysis. In: Copernicus Climate Change Service (C3S): ERA5: Fifth generation of ECMWF atmospheric reanalyses of the global climate. Copernicus Climate Change Service Climate Data Store (CDS). https://cds.climate.copernicus.eu/cdsapp#!/home\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"natural-hazards","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"nhaz","sideBox":"Learn more about [Natural Hazards](https://www.springer.com/journal/11069)","snPcode":"11069","submissionUrl":"https://submission.nature.com/new-submission/11069/3","title":"Natural Hazards","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Meteotsunami, Numerical model, Forecast, Real-time data","lastPublishedDoi":"10.21203/rs.3.rs-6861009/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6861009/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eMeteotsunamis, often overlooked in global disaster discussions, pose significant threats to coastal communities and infrastructure, particularly in shallow coastlines, narrow bays, and harbors. Unlike seismic tsunamis, where source parameters can typically be determined quickly following an earthquake, forecasting meteotsunamis remains challenging due to difficulties in obtaining real-time high-resolution atmospheric pressure data that contain meteotsunami source parameters. Intense short-period atmospheric pressure disturbances are typical triggers of meteotsunamis, but these small-scale pressure anomalies are often missed in observations and numerical weather forecasts. This study explores the feasibility of using weather radar reflectivity as a proxy for atmospheric pressure anomalies to enhance meteotsunami forecasting capabilities. The study employs the Method of Splitting Tsunamis (MOST) model, commonly used for seismic tsunami forecasting, across three very different regions: the East Coast of the United States, the southeastern Baltic Sea, and the Adriatic Sea (the Mediterranean). The results demonstrate that radar-based modeling successfully captures meteotsunami generation and propagation dynamics, particularly for summer events driven by mesoscale convective systems. The findings suggest that real-time meteotsunami forecasting is achievable using weather radar inputs, providing a promising approach for coastal hazard mitigation and early warning systems. However, winter meteotsunamis which are often related to extratropical cyclones and wind-driven forces, and which are superimposed on ongoing storm surges require additional modeling refinements.\u003c/p\u003e","manuscriptTitle":"Advancing meteotsunami prediction: testing high-resolution modeling in harbor environments","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-13 10:25:52","doi":"10.21203/rs.3.rs-6861009/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"","date":"2025-06-13T07:12:53+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-06-11T15:42:39+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-06-10T09:04:40+00:00","index":"","fulltext":""},{"type":"submitted","content":"Natural Hazards","date":"2025-06-10T04:34:29+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"natural-hazards","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"nhaz","sideBox":"Learn more about [Natural Hazards](https://www.springer.com/journal/11069)","snPcode":"11069","submissionUrl":"https://submission.nature.com/new-submission/11069/3","title":"Natural Hazards","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"5b0f3675-6e6e-462c-971b-6f8632f2bef9","owner":[],"postedDate":"June 13th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2026-02-09T16:05:40+00:00","versionOfRecord":{"articleIdentity":"rs-6861009","link":"https://doi.org/10.1007/s11069-025-07894-w","journal":{"identity":"natural-hazards","isVorOnly":false,"title":"Natural Hazards"},"publishedOn":"2026-02-04 15:57:22","publishedOnDateReadable":"February 4th, 2026"},"versionCreatedAt":"2025-06-13 10:25:52","video":"","vorDoi":"10.1007/s11069-025-07894-w","vorDoiUrl":"https://doi.org/10.1007/s11069-025-07894-w","workflowStages":[]},"version":"v1","identity":"rs-6861009","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6861009","identity":"rs-6861009","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}