The research, published in the Bulletin of Earthquake Engineering, was led by Snježana Markušić of the University of Zagreb together with colleagues from the University of Zagreb’s Faculty of Geotechnical Engineering and the Croatian Geological Survey. The team used the OpenQuake Engine, developed by the Global Earthquake Model Foundation and now the de facto international standard for hazard computation, running it in classical mode to integrate earthquake occurrence models, ground motion prediction equations, and site characterisation into a single probabilistic framework. The result is a set of hazard maps covering peak ground acceleration and spectral accelerations between 0.1 and 2.0 seconds, computed on a 0.1-degree grid for seven return periods ranging from 95 to 2475 years.

The study’s principal scientific contribution lies not in the computational workflow but in the construction, justification, and validation of a Croatia-specific seismic source model. The model comprises 32 area source zones grouped into nine seismotectonic super-zones, spanning the country’s three major tectonic domains: the comparatively rigid Adriatic Foreland, the NW-SE trending Dinarides fold-thrust belt dominated by reverse and thrust faulting, and the Pannonian Basin in the north, a Miocene back-arc extensional basin where normal and strike-slip faulting prevail. Earthquake recurrence parameters for each zone were estimated with the Weichert maximum likelihood method, which allows completeness to vary by magnitude, and the entire model was validated against independent seismicity observations before being combined with ground motion models.

The earthquake catalogue underpinning the analysis merged three sources: the Croatian Earthquake Catalogue, the International Seismological Centre Bulletin, and the SHARE European Earthquake Catalogue. After careful deduplication, a single declustering pass using the Gardner-Knopoff algorithm with Grünthal space-time windows yielded 90,003 mainshocks, with 74.2 percent of all events identified as aftershocks and 10.9 percent as foreshocks. Crucially, the catalogue incorporates the 2020 Petrinja sequence, something the pan-European ESHM20 model, with its 2020 data freeze, could not capture. Magnitude of completeness was determined independently for each super-zone through a temporal rate-stability analysis, revealing marked differences between the well-monitored Dinarides, where completeness stabilises from the 1980s onward, and parts of the Pannonian-Bosnian domain where detection is still improving.

One of the study’s most distinctive methodological innovations concerns focal mechanisms. Rather than assigning a single dominant faulting style to each zone, as even continental-scale models typically do, the team derived zone-specific nodal plane distributions from the CroFMS 2024 catalogue of 410 focal mechanism solutions. Using Kagan-angle clustering with silhouette-optimised cluster selection and moment-tensor averaging, they represented the observed diversity of faulting within each zone as a weighted, multi-cluster distribution. In the Zagreb zone, for example, 51 mechanisms resolved into three clusters of reverse, strike-slip, and oblique-reverse character, none exceeding 40 percent weight. Because ground motion models typically shift median peak ground acceleration by 10 to 20 percent between reverse and strike-slip ruptures, collapsing such a mixture into a single mechanism would systematically bias hazard estimates.

The ground motion characterisation was equally rigorous. From an initial candidate set of 19 models evaluated against a regional Croatian strong-motion dataset, five were selected and weighted using a combined ranking based on the Deviance Information Criterion and the Euclidean Distance-Based Ranking method, with meta-weights of 0.80 and 0.20 respectively. The top-ranked model, Kowsari et al. 2020 Model Y5, calibrated on a pan-European and Mediterranean dataset that closely matches Croatian strong-motion characteristics, was not available in OpenQuake’s library, so the team implemented it as a custom Python module validated to within 1 percent of published median predictions. Hazard was computed on a spatially varying, real ground-surface site model built from 733,761 site locations, capturing local soil amplification directly rather than defaulting to Eurocode 8’s uniform reference-rock condition.

The resulting hazard maps reveal a striking spatial pattern. Hazard peaks along the Adriatic coast, with the Dubrovnik-Neretva coast identified as the highest-hazard zone in the country at every intensity measure and return period. At Dubrovnik, 475-year peak ground acceleration reaches 0.40 g, with spectral acceleration at 0.2 seconds reaching 0.94 g, and 2475-year values climbing to 0.85 g and 2.00 g respectively. Split follows at 0.32 g and Rijeka at 0.27 g for the 475-year return period. Perhaps counterintuitively, the Sisak-Moslavina area, despite hosting the destructive 2020 Petrinja sequence, falls only in the moderate tier at approximately 0.22 g, while Osijek in the low-seismicity Pannonian domain records the country’s lowest values at just 0.08 g, roughly one-fifth of the coastal maximum.

The most consequential finding, however, emerges from a systematic comparison with the Eurocode 8 design spectra in Croatia’s National Annex. The comparison revealed a previously undocumented, period-dependent discrepancy: probabilistic hazard spectra exceed code values at short periods but fall below them at longer periods. At the peak ground acceleration level, the new hazard estimates exceed the National Annex anchor values by 31 to 46 percent at four of the six representative cities, Petrinja, Rijeka, Split, and Dubrovnik, consistent with real-surface site amplification exceeding the implicit reference-rock basis of the current zonation. Yet at a spectral period of 1.0 second, the ratio drops below unity at every city without exception, meaning the current Eurocode 8 Type 1 spectrum decays more slowly beyond its 0.4-second corner period than the hazard analysis suggests is warranted, making the code comparatively more conservative for longer-period structures such as tall buildings and bridges.

The authors are careful to position the work as a scientific foundation rather than a statutory document. It does not set design values or follow the formal drafting procedure a National Annex requires, but it offers exactly the kind of transparent, independently reproducible basis that national standardisation bodies can draw upon, much as earlier studies informed the current first-generation Annex. The findings arrive at a pivotal moment, as Croatia prepares to implement the second generation of Eurocode 8, published in 2024, which introduces revised spectral shape parameterisation allowing country-specific corner periods, a unified European reference hazard model, and new ground-type classifications. The study’s results on the period-dependent scaling of hazard between the 475-year and 2475-year return periods, which ranges from roughly 2.0 at short periods to 2.87 at 2.0 seconds, also suggest that a single uniform importance factor for essential structures may not capture the full site and period dependence of rare-event hazard.

Limitations remain, and the authors flag them candidly. The source model relies solely on area sources, with explicit fault sources for major structures such as the Petrinja and Pokupsko faults deferred to the next model generation, since a defensible hybrid model would require slip-rate data that do not yet exist for the faults responsible for Croatia’s largest earthquakes. The site model depends partly on inferred values, and the strong-motion database used for ground motion ranking is still relatively small. A sensitivity check on maximum magnitude assignment revealed that this single parameter can matter more than the ground motion or b-value branches that typically receive the most logic-tree attention, particularly at the 2475-year hazard level used for essential-facility design. With its open, reproducible input package and validated source model, the study provides both a warning and a toolkit: Croatia’s design spectra need period-dependent recalibration, and the scientific machinery to accomplish it now exists.

Subject of Research: Probabilistic seismic hazard analysis and earthquake hazard mapping for Croatia with implications for Eurocode 8 seismic design

Article Title: Probabilistic seismic hazard analysis for Croatia: hazard mapping and Eurocode 8 implications

Article References: Markušić, S., Stanko, D., Žilić, I., Fiket, T., Majurec, A., & Korbar, T. (2026). Probabilistic seismic hazard analysis for Croatia: hazard mapping and Eurocode 8 implications. Bulletin of Earthquake Engineering. https://doi.org/10.1007/s10518-026-02691-7

Image Credits: AI Generated

DOI: 10.1007/s10518-026-02691-7

Keywords: probabilistic seismic hazard analysis, Croatia, Eurocode 8, OpenQuake Engine, seismic hazard maps, Dinarides, Adriatic coast, Petrinja earthquake, ground motion models, uniform hazard spectra, seismic source model, site amplification

The project, titled Earthquake Risk Assessment of the City of Zagreb, was conceived in 2014 but only approved and funded two days after the March 2020 earthquake. Researchers from the University of Zagreb Faculty of Civil Engineering, working with the Croatian Centre for Earthquake Engineering, city departments, seismologists, geologists and geotechnical experts, carried out the work between 2021 and 2023 as part of a broader EU-financed aerial survey programme for disaster risk reduction. The team describes the initiative as a pilot intended to serve as a template for earthquake risk assessments across Croatia and as a foundation for changes to national regulations expected in 2026.

Zagreb’s hazard profile is serious. The city lies along the southeastern edge of Medvednica mountain, and its seismic threat comes mainly from the Medvednica epicentral area, particularly two seismogenic sources: the Reverse Northern Medvednica boundary fault and the nearly perpendicular Kašina strike-slip fault. Historical records recall the Great Zagreb earthquake of 1880, assessed at magnitude 6.1, whose epicentre lay very near the 2020 event and which caused widespread damage and emigration. The official Croatian hazard map assigns the city area a reference peak ground acceleration on rock of 0.20 to 0.28 grams for a 475-year return period, and investigations so far indicate that Eurocode 8 ground types B and C predominate, although a full seismic microzonation of the city remains unfinished.

The greatest technical achievement of the project, according to its authors, was the creation of a comprehensive geographic information system building inventory database. Croatia has no national building registry, so the team overlapped three official data sources: the Digital Cadastral Plan, the Register of Spatial Units and the Existing Land Use 2020 dataset. Of more than 300,000 objects examined, roughly 230,000 proved to be buildings and auxiliary structures, while about 70,000 were stairways and terraces that were excluded. Engineers then collected 35 required attributes for each building, covering geographic position, site category, year of construction, material and type of the lateral load-resisting system, number of storeys, structural regularity, position within a block, interaction with adjacent buildings, floor and roof systems, foundations, occupancy, cultural heritage status and physical condition, among others. Additional attributes consistent with the Global Earthquake Model building taxonomy, including system ductility and seismic code level, were added.

Populating the database proved laborious. Post-earthquake damage assessment databases from Zagreb and Petrinja, cadastral records, aggregated census data, design documentation, and LiDAR recordings from the national aerial survey all fed into the model. Ten experts spent two years collecting, verifying and entering data, combining field surveys with remote desktop work. Public participation, tested through an online questionnaire, yielded fewer than 300 responses, with residents citing privacy concerns, lack of time and anxieties about illegally constructed or modified buildings. The resulting exposure model covers approximately 136,000 residential buildings. More than half of them, 52 percent, are unreinforced masonry structures, 39 percent are confined masonry, 8 percent reinforced concrete and about 1 percent timber. Among unreinforced masonry buildings, roughly three quarters have rigid floor diaphragms. The replacement value of the residential stock is estimated at about 86 billion euros.

The risk model itself rests on three components: hazard, exposure and vulnerability. Two hazard models were used, one developed within the project from eleven seismic area sources within 200 kilometres of the city’s main square, built on the Croatian Earthquake Catalogue, and the European model ESHM20. The project-specific model generated 100,000 stochastic event sets through Monte Carlo sampling, each representing one year of seismicity, and used six ground motion prediction equations in a logic tree. Notably, the local model proved considerably more conservative than the European one for weaker, more frequent earthquakes, producing higher average annual losses. In accordance with the project methodology, the reported results rely on the project-specific hazard model. On the vulnerability side, because locally derived fragility curves were lacking, the team used the Global Earthquake Model database of roughly 500 building classes, applying 98 vulnerability curves to Zagreb’s residential typologies, with consequence models for collapse, fatalities, injuries, displacement and financial loss.

The probabilistic results, computed with the OpenQuake Engine, are sobering. For an earthquake with a 500-year return period, the assessment predicts around 900 collapsed buildings, approximately 600 fatalities, about 2,000 injured people and roughly 79,000 residents requiring temporary shelter, with direct financial losses of about 9.6 billion euros. Around 9,000 buildings would reach complete damage, about 7 percent of the stock, while a further 5,100 would be heavily damaged. Average annual losses amount to roughly 215 million euros, an average annual loss ratio of 0.25 percent that the authors consider high. Financial losses concentrate in the historic centre, where cultural heritage protection raises replacement costs, while the largest numbers of collapsed buildings and fatalities are expected in the northeastern districts, driven mainly by the amount of exposed built-up area. Unreinforced masonry buildings account for about half of all expected financial losses.

Two deterministic scenarios sharpen the picture. A repeat of the 1880 earthquake, modelled at magnitude 6.3 on the Reverse Northern Medvednica boundary fault, would collapse around 1,460 buildings, kill about 1,050 people, injure roughly 3,400 and displace approximately 126,000, with losses near 14.6 billion euros. A repeat of the 2020 magnitude 5.4 event would produce about 300 collapsed buildings, 170 fatalities, 630 injuries, 25,000 displaced residents and losses of about 3.5 billion euros. When the modelled 2020 scenario was compared with actual observed impacts, agreement was good at both city-wide and district scales: the districts that suffered the heaviest losses in reality, led by Donji Grad and Gornji Grad-Medveščak, match the model’s predictions, although the post-earthquake government report estimated higher total financial losses, partly reflecting different unit prices and market changes.

The project’s results have already found practical application. In collaboration with the Zagreb City Office, researchers evaluated the city’s official evacuation routes for potential blockage by debris from collapsed buildings. Following Croatian regulations, which assume debris reaches a distance equal to half the building height, the team created GIS buffers around building footprints, intersected them with road surfaces, and ranked buildings by their probability of collapse in the 1880-type scenario. The analysis identified safer access zones for civil protection forces but showed that problems are expected in the central part of the city, prompting a subsequent detailed study of the Lower Town using discrete element simulations of masonry façade collapse. The authors caution that current fragility models do not capture non-structural elements such as chimneys, parapets and gables, which caused many road blockages in the 2020 earthquake.

The team is candid about the model’s limitations and future needs. There is no microzonation for the entire city, secondary hazards such as landslides and liquefaction were not considered, many building attributes relied on expert judgement, and locally specific fragility and vulnerability functions remain a critical gap, although Croatian research projects are now developing them. Indirect losses and retrofit scenarios are not yet included, and the building database has not been established as an officially maintained data source. Even so, the authors conclude that earthquake risk for Zagreb is high and that the model provides a robust foundation for prioritising retrofitting and preparedness. Masonry buildings, responsible for about 80 percent of expected damage and losses, emerge as the clear priority. With national regulatory changes expected in 2026, the pilot marks a decisive shift from general awareness to data-driven risk management, offering a methodological and data-rich framework that other seismic-prone cities in the region and beyond can adapt.

Subject of Research: Building-level earthquake risk assessment of Zagreb, Croatia, using GIS exposure data and OpenQuake probabilistic modelling

Article Title: Earthquake risk assessment of the city of Zagreb, Croatia: recent advances

Article References: Šavor Novak, M., Uroš, M., Baniček, M., Demšić, M., & Atalić, J. (2026). Earthquake risk assessment of the city of Zagreb, Croatia: recent advances. Bulletin of Earthquake Engineering. https://doi.org/10.1007/s10518-026-02662-y

Image Credits: AI Generated

DOI: 10.1007/s10518-026-02662-y

Keywords: earthquake risk, Zagreb, Croatia, seismic hazard, exposure model, OpenQuake, unreinforced masonry, building inventory, vulnerability, evacuation routes, GIS database, disaster risk reduction

The study’s central innovation lies in its sampling method. Rather than lowering bottles from a vessel or relying on remote sensors, trained divers descended to specific underwater structures, including a small harbor basin, limestone drop-offs, a wooden shipwreck resting at 22 to 26 meters, and a submerged pipe extending from the shoreline to a depth of 47 meters. The pipe, officially designated as a decommissioned wastewater overflow, was a particular target of suspicion. Divers could visually identify the exact point where an effluent plume emerged, position sampling bottles directly within it, and document the surrounding environment on underwater slates, a level of spatial precision that neither autonomous vehicles nor surface-based sampling can match.

In total, the team collected 30 samples across depths ranging from half a meter to 47 meters, supplemented by four nearby reference dive sites. Every sample was brought to a mobile field laboratory within 30 minutes, where pH, salinity, and redox potential were measured before the water chemistry could shift. Trace metal samples were filtered, acidified with nitric acid, and stored cool, while samples for surfactant analysis were preserved with a small amount of acetonitrile to halt microbial degradation. Back in Germany, the metals were measured by inductively coupled plasma optical emission spectroscopy after solid phase extraction, a pre-treatment step essential for stripping away the interfering salt matrix of Adriatic seawater and concentrating the analytes to detectable levels.

The baseline hydrochemistry told a story of stability. Surface waters reached 25 degrees Celsius, with a distinct thermocline between 20 and 25 meters separating a warm, well-mixed upper layer from a cooler, more stagnant deep zone. pH held steady at 8.06, absolute salinity averaged 39.3 grams per kilogram, and redox potential of roughly 441 millivolts confirmed that the entire water column was well oxygenated, conditions that keep iron and manganese oxides stable and prevent the release of adsorbed metals from sediments.

One sample shattered that calm. At station SP05, drawn directly from the sewer pipe outlet at 47 meters, the water was turbid, yellowish-brown, and carried a distinct sewage odor. Its salinity of just 10.9 grams per kilogram marked it as freshwater-dominated effluent, its pH of 6.44 represented dramatic local acidification driven by carbon dioxide from microbial breakdown of organic matter, and its redox potential of nearly minus 99 millivolts revealed an anoxic micro-environment. That oxygen-starved plume matters chemically: the aerobic degradation of detergents requires molecular oxygen, so the plume’s own organic load effectively shields its pollutants from breakdown, allowing them to persist and spread into the stratified layers of the bay.

The surfactant results were the study’s most striking. Linear alkylbenzene sulfonates, or LAS, the world’s most widely used anionic detergents, are exclusively anthropogenic and therefore unambiguous tracers of domestic wastewater. At SP05, the divers measured a concentration of roughly 1469 parts per billion, nearly 300 times the background level. Every other sample in the bay, including the designated reference site, contained a consistent low-level signal of about 5 parts per billion. That ubiquitous background suggests either that the bay has reached a chemical steady state in which continuous inputs during the tourist season balance slow degradation, or that diffuse sources such as greywater runoff, contaminated karst groundwater, or discharges from recreational vessels are feeding the system. Because no comparable Mediterranean baseline datasets for marine surfactants exist, the authors say their values provide a critically needed reference for future monitoring along the Adriatic coast.

The trace metal analysis, by contrast, revealed contamination confined to specific spots rather than systemic pollution. Most elements were detected below 5 parts per billion, at or beneath concentrations reported for other Adriatic locations, confirming largely pristine ambient water. Zinc stood out with occasional values exceeding 20 parts per billion, consistent with antifouling paints, galvanized materials, and corrosion-protection coatings. Statistical analysis of the metal data revealed two clearly separated element clusters. An aluminum-cadmium-copper-zinc group points to maritime activity: copper and zinc leach from antifouling coatings, aluminum and zinc come from sacrificial anodes on hulls and harbor infrastructure, and cadmium travels as a characteristic impurity in zinc alloys. A second iron-manganese-nickel cluster reflects the region’s geology, where weathering of red-oxide-bearing limestone and terra rossa soils naturally releases these elements into the sea.

Principal component analysis sharpened this distinction, explaining 72 percent of the variance in two factors that separate total metal load from the contrast between the anthropogenic and geogenic clusters. The biplot flagged telling anomalies: samples taken beside the wooden shipwreck showed elevated iron, manganese, and nickel, consistent with the slow corrosion of the wreck’s iron fasteners, while a harbor sample displayed elevated copper, zinc, and cadmium, a signature of touristic boat traffic. One outlier station, where a parallel team had disturbed sediment shortly before sampling, illustrated how easily diving-based measurements in soft sediments can be skewed by resuspended pore water, a caution the authors fold into their methodological recommendations.

Crucially, the team emphasizes that none of these signals currently threatens the bay’s ecology. Outside the immediate outfall, metal concentrations and the pervasive 5 parts per billion surfactant background remain well below toxicological thresholds, and the Northern Adriatic’s high dilution capacity appears to buffer these localized pressures effectively. But the molecular warning is clear. The authors recommend modernizing the wastewater infrastructure at the submerged sewer pipe and expanding future work into multi-seasonal campaigns, mass-spectrometric separation of detergent homologs, and dive-computer-based three-dimensional mapping of sampling positions. Their broader message is that scientific diving, paired with rigorous multi-parameter chemistry, can detect subtle anthropogenic change in coastal waters long before it escalates into visible environmental damage, offering coastal managers an early-warning tool precisely where tourism pressure is growing fastest.

Subject of Research: Anthropogenic trace metal and surfactant contamination assessment of a Croatian coastal bay in the Northern Adriatic Sea using scientific diving

Article Title: Assessment of anthropogenic trace metal and surfactant influence on a coastal area in the Northern Adriatic Sea using scientific diving

Article References: Franke, E., Franke, N., Viehweger, C., & Vogt, C. (2026). Assessment of anthropogenic trace metal and surfactant influence on a coastal area in the Northern Adriatic Sea using scientific diving. Discover Oceans, 3(1), Article 47. https://doi.org/10.1007/s44289-026-00158-w

Image Credits: AI Generated

DOI: 10.1007/s44289-026-00158-w

Keywords: marine pollution, scientific diving, Northern Adriatic Sea, trace metals, linear alkylbenzene sulfonates, wastewater tracers, coastal monitoring, antifouling, principal component analysis, seawater chemistry, Croatia, tourism impact