Flood-risk data sources in Hungary: OVF maps, levees and river forecasts

Who produces flood-risk data in Hungary

Hungarian flood-risk data is produced by one national body and its regional arms. The Országos Vízügyi Főigazgatóság (OVF), the national water directorate, is the central authority for flood defense, ice defense, inland excess water and water-resource management. It works through 12 regional water directorates (vízügyi igazgatóságok): Észak-dunántúli in Győr, Közép-Duna-völgyi in Budapest, Alsó-Duna-völgyi in Baja, Közép-dunántúli in Székesfehérvár, Dél-dunántúli in Pécs, Nyugat-dunántúli in Szombathely, Felső-Tisza-vidéki in Nyíregyháza, Észak-magyarországi in Miskolc, Tiszántúli in Debrecen, Közép-Tisza-vidéki in Szolnok, Alsó-Tisza-vidéki in Szeged and Körös-vidéki in Gyula. Each directorate operates the levees, pumping stations and gauges in its territory and declares flood readiness on its own defense sections.

Inside OVF sits the Országos Vízjelző Szolgálat, the national water forecasting service, which collects, processes and publishes the hydrological data that characterise the rivers every day: about 40,000 data points from roughly 1,700 meteorological stations and nearly 800 hydrological observation points, turned into forecasts for 98 cross-sections on the Danube, Tisza and Dráva-Mura systems. The operational data, the readiness maps and the Floods Directive products are published on vizugy.hu, the forecasts on hydroinfo.hu, and the planning documents on vizeink.hu.

The legal spine is Directive 2007/60/EC, transposed by Government Decree 178/2010, which obliges the authorities to prepare a preliminary flood risk assessment, flood hazard and flood risk maps, and a flood risk management plan on the six-year cycle the directive prescribes. Hungary is unusual among its neighbors in that the plan is national rather than per basin, reflecting a country whose lowlands drain to two rivers, the Danube and the Tisza, and whose flood defense has been a single state undertaking since the Tisza regulation began in 1846.

That history is the key to reading any Hungarian flood map. The regulation of the Tisza, organized from 27 August 1846 and substantially complete by 1880, cut the river's Hungarian length to 966 km, with 589 km of dead channels and 136 km of new bed, and enclosed the former floodplain behind levees. The lowland is therefore not exposed to floods in the ordinary sense: it is protected by embankments whose height and condition decide everything. A hazard map that shows a village dry is showing that the levee holds; a map that ignores the levee shows the same village under meters of water. Both statements are true, and the sections below explain which sources say which.

The Floods Directive cycle in Hungary

OVF summarizes the implementation as four phases. The preliminary assessment identified the flood-prone areas needing detailed study, the first territorial hazard and risk maps followed, the first national plan was approved, and the first six-year review produced a revised plan adopted by government resolution. The maps, in OVF's description, show the probability with which an area is inundated and the water depths that can develop.

OVF, Árvízi kockázatkezelés (Floods Directive implementation page); Directive 2007/60/EC for the third-cycle deadlines.

Source inventory

Every public dataset a flood-risk analysis of a Hungarian address can draw on. The national products are published for viewing; where a publisher does not state a reuse license on the pages checked, the license column says so rather than guessing.

Compiled from the publishers named in the sources list; cadence and license as stated by each publisher on the pages checked.

Levees, readiness grades and why defense dependence matters

The defining feature of Hungarian flood risk is that most of the exposed lowland is protected rather than merely low. The Tisza regulation and the Danube works turned a floodplain into a defended polder landscape, and the water directorates have been maintaining, raising and reinforcing those embankments ever since. OVF's flood-defense remit covers not only river floods but ice defense on the Danube and Tisza and inland excess water (belvíz), the ponding of rain and snowmelt on flat land that cannot drain while the rivers run high. All three appear on vizugy.hu as separate readiness map modules, and all three can hit the same property.

Readiness is the operational vocabulary. When a river rises past the thresholds set for a defense section, the responsible directorate declares a readiness grade for that section and the vizugy.hu map turns color. The grades escalate with the stage and with the works required, from patrols to continuous defense, and the situation reports behind them are the fastest public account of what a specific levee is doing during an event. For a property assessment they also carry history: a section that has been in the highest grade repeatedly is a section whose levee has been tested.

Defense dependence changes how the other sources must be read. The official Hungarian maps model the identified areas with the defenses in place, so a village behind a sound levee shows a low probability of inundation. The JRC European maps are explicitly undefended and show the same village under water at the 1-in-100-year discharge, because they ignore the embankment. Neither is wrong: one answers "what happens if the defenses hold", the other "what is at stake if they do not". A serious assessment carries both, and adds the levee crest height, which can be read from any terrain model good enough to resolve it. That is the main reason terrain data matters more in Hungary than the flat landscape suggests.

Gauges and forecasts from the Országos Vízjelző Szolgálat

The forecasting service is the most data-rich of the Hungarian sources and the one with the longest continuity. OVF describes a daily cycle in which roughly 40,000 data points from about 1,700 meteorological stations and nearly 800 hydrological observation points are collected, processed and published, feeding forecasts for 98 cross-sections across the Danube, Tisza and Dráva-Mura systems. The daily water conditions map has a legacy of more than a century, and every 1 March the service issues a spring runoff report forecasting the season on the major rivers.

Hungary sits downstream of every neighbor on the Danube and the Tisza, so the service exchanges hydrological data daily with the other Danube basin countries and maintains the OPADAT operative hydrometeorological database under the Danube Commission's recommendations. In practice this means a flood wave on the Hungarian Danube is forecast days ahead from Austrian, Slovak and German gauges, which is why the 2013 and 2024 events were met with preparations rather than surprise. The forecast map on hydroinfo.hu is the public face of that chain; the vizugy.hu hydrological data pages carry the observations.

For a property assessment the forecasting service answers two questions the maps cannot: how high the nearest gauge has actually been, and how quickly a wave arrives. Record stages at Budapest, Szeged or Szolnok are the calibration points for any depth model, and the forecast lead time is what turns a hazard into an insurable or uninsurable one.

Terrain: what Lechner publishes and what it does not

Elevation data in Hungary is held by the Lechner Tudásközpont, the state knowledge centre that absorbed the former Institute of Geodesy, Cartography and Remote Sensing (FÖMI) and now runs the national geoportals. Its products are cataloged in the INSPIRE geonetwork and sold through the Geoshop geoportal, renovated under the VIZEK project, where an order begins by selecting an area and the system then shows which products are available for it. Viewing services are free with attribution; the downloadable products are priced. The related fentrol.hu archive holds the national aerial photographs and orthophotos back to the 1980s, with 8,000 photographs added in the first half of 2026, and requires registration.

What Hungary does not publish is an open, countrywide LiDAR terrain model of the kind its northern neighbors release at 1 m under open licenses. The elevation theme available through the national INSPIRE services is a coarser model, and airborne laser scanning exists for projects and cities rather than as a national open product. For flood modeling this has two consequences: the open fallback is the Copernicus DEM GLO-30, a 30 m surface model that cannot resolve a levee crest or a village street, and any depth estimate at a property that does not come from the official maps must either buy Lechner elevation products for the area or accept the coarser layer with its wider error.

The practical advice is therefore different from the Czech or Polish case. Where an address falls inside an identified area, the official OVF map is the terrain-aware answer and should be used as such. Outside those areas, a defensible estimate needs purchased elevation data or a documented margin of error, and the levee position must be taken from the water directorates rather than inferred from a 30 m grid.

European and satellite layers

• JRC river flood hazard maps. Issued on 8 March 2024 under CC BY 4.0, they give water depth for nine return periods from 1-in-10 to 1-in-500 years on basins above 150 km² at about 100 m, produced with the LISFLOOD and LISFLOOD-FP models that also power EFAS. The JRC states that they are not an official flood hazard map. In Hungary their most important property is that they are undefended: they show the full lowland floodplain of the Tisza and the Danube as if no levee existed, which is the worst case, not the expected one.
• European Flood Awareness System (EFAS). The Copernicus forecasting system produces river flood probabilities up to 15 days ahead for the whole Danube basin, which makes it the natural cross-check on the national forecasts for a downstream country. It is an early-warning product for authorities and registered users rather than a hazard map.
• Copernicus Emergency Management Service mapping. Rapid Mapping delineates flooded areas for activated events and Global Flood Monitoring turns every Sentinel-1 pass into a systematic flood archive. For a defended landscape the observed extents matter twice: they show where water actually went, and where it did not go because a levee held.
• EEA Flood Risk Areas Viewer. The European Environment Agency publishes the areas of potential significant flood risk that Hungary has reported, in the same frame as its Slovak, Romanian, Serbian, Croatian and Austrian neighbors, which is the only place a Danube or Tisza reach can be read across a border in one view.

Reference floods in the public record

Four events anchor the modern Hungarian record: the Danube flood of spring 2006, the flash floods in the north-east in 2010, the Danube record of June 2013 and the September 2024 wave from Storm Boris. The figures below come from encyclopedic summaries that cite Hungarian press and government statements; the OVF and water directorate reports are the authoritative versions.

Wikipedia summaries citing Hungarian press and government sources.

How a model combines these sources

A property-level flood model for a defended landscape carries two answers side by side: the official OVF hazard map where an address falls inside an identified area, which assumes the levees hold, and an undefended estimate built from the JRC layers and the available terrain, which shows what is at stake if they do not. The gauge record and the readiness history of the nearest defense section calibrate the first; satellite-observed floods validate both before any figure is published. Availability by market is listed on the coverage page.

Questions