If the Savinja Overflows Its Banks Again
How safe is the Savinja Valley after the 2023 floods? A review of flood measures, state plans, local concerns and questions that remain unanswered.
AI-generated conceptual illustration of the Savinja Valley, not an authentic photograph, exact location, or documentary view of completed flood-protection measures.
How Safe Is the Savinja Valley Today, and Could Future Floods Be Managed Differently?
In August 2023, the Savinja demonstrated how quickly the boundary between a river and the land occupied by human settlements can change. Within hours, watercourses overflowed their banks, undermined roads, destroyed bridges and reshaped the river landscape in numerous locations. More than three years have passed since then. Across the Savinja Valley, damaged riverbanks are being restored, flood-protection works are under way and extensive national spatial plans are being prepared. The government promises a systematic reduction in flood risk. Residents of flooded settlements expect greater safety, while some farmers and landowners warn about the consequences of the proposed flood-retention basins.
Yet one question remains above all these efforts, and it cannot be answered by political promises or photographs of a reconstructed embankment.
What would happen if the Savinja Valley were struck again by a rainfall event as extreme as the one in August 2023?
Would the restored river channels and embankments withstand it? Would the planned retention basins reduce the flood peak? Could new river works cause water to threaten different areas? And are there additional ways to improve protection without causing disproportionate harm to nature, farmland and local communities? We cannot answer these questions simply by comparing photographs taken before and after restoration. We must understand what has actually been built, what remains at the planning stage and which proposed solutions have yet to be tested.
For this reason, the final part of our investigation compares three distinct situations: measures already implemented, the fully planned flood-protection system and additional opportunities to improve that system. We will rely exclusively on publicly available material and clearly distinguish between the results of official models, public statements about those results and our own research hypotheses.
What Did August 2023 Teach Us?
Orthophotographs of the Savinja and tributaries, acquired 7–8 August 2023 after the peak flood wave.
During the night of 3–4 August 2023, exceptionally intense rainfall affected several river basins in northern Slovenia. According to a report by the Slovenian Environment Agency (ARSO), some catchments already saturated by rainfall in July received between 150 and 200 millimetres of rain within six to twelve hours. Water rushed down the slopes. Mountain torrents carried gravel, trees and other debris. The Savinja and its tributaries reached exceptional discharges within a short period. [1]
ARSO reported the following peak discharges for the Savinja on 4 August 2023:
- Nazarje: 859 cubic metres per second.
- Letuš: 935 cubic metres per second.
- Celje – pedestrian bridge: 1,173 cubic metres per second.
- Laško: 1,450 cubic metres per second.
These figures refer to different gauging stations and to discharges determined through their respective measurement or hydrological assessment methods. They do not represent a single value applicable to the entire Savinja Valley. Of particular importance is the distinction between discharge determined at a specific gauging cross-section and later hydraulic reconstructions, which may also account for water flowing outside the ordinary river channel. At Nazarje, subsequent research also examined substantially larger reconstructed total discharges. Such figures must be presented alongside the methods used to derive them, rather than being compared uncritically with the 859 cubic metres per second reported by ARSO.
There were also differences in estimated return periods. ARSO estimated that the peak discharges at Nazarje and Letuš exceeded a 100-year return period, while those at Celje and Laško were associated with an estimated return period of approximately 50 years. This does not mean that the event was insignificant farther downstream. It means that the statistical magnitude of a flood is assessed in relation to the characteristics of each watercourse, gauging location and available dataset. But the severity of the event was not only a matter of discharge.
In many places, the geometry of river channels changed. Water eroded riverbanks, deposited large quantities of sediment, undermined building foundations and flowed through areas that remain dry under ordinary river conditions. A flood is therefore not simply a question of whether water can pass between two embankments. It is also a question of whether the channel will retain its shape, whether bridges will remain hydraulically unobstructed and whether accumulated gravel and woody debris will redirect the current. Under such conditions, a structure may behave very differently from how it performs during a more ordinary high-flow event.
Future flood-protection measures must therefore not be designed on the assumption that rivers, riverbanks and bridges will remain geometrically unchanged during the next major flood.
The most important warning from August 2023 was that, during an extreme event, water can transform the very landscape through which it flows.
What Has Actually Been Done Since the Floods?
Extensive recovery and restoration work began across the Savinja Valley following the disaster. Damaged riverbanks were repaired, hazardous deposits were removed, transport connections were restored and new flood-protection projects were prepared. However, there is an important distinction between post-disaster recovery and completed flood protection. Restoration may involve returning a damaged structure or river section to its previous condition, locally improving it or constructing an entirely new facility. It does not necessarily mean that an individual intervention has reduced the overall flood risk faced by a settlement.
When assessing the valley's present level of protection, we must therefore distinguish between completed and functional measures, works still under construction and projects for which documentation exists but implementation has not yet taken place.
River Works Between Ločica and Male Braslovče
On 16 April 2026, the Water Directorate of the Republic of Slovenia (DRSV) reported that extensive works were under way along the Savinja between Male Braslovče and Ločica ob Savinji. The project envisages interventions along more than seven kilometres of the river and approximately 1.8 kilometres of the Ložnica flood-diversion channel. [2] The planned measures include widening the river channel, reshaping riverbanks, constructing and raising embankments, managing gravel bars, stabilising the riverbed and planting vegetation along the banks.
They also include systems of cribwork groynes and six grade-control structures with fish passages. This detail is particularly interesting in light of the second article in our investigation. Directing a river's flow through engineered structures is not an idea found exclusively in Schauberger's patents. Comparable categories of river-training structures are part of contemporary hydraulic-engineering practice. However, their construction, purpose and operation are not necessarily identical to Schauberger's historical proposals.
Their performance must be evaluated on the basis of their actual geometry, design discharges, structural stability and observed effects on river behaviour. The official project register indicates an implementation period extending to 31 December 2028. [10] We must therefore not assume that the entire project has been completed merely because construction has begun.
Celje and Polule
In January 2026, the Ministry reported on flood-protection works in the Polule area of Celje. An important component is a new bridge approximately 110 metres long, designed without supports in the river channel. The public presentation specified a design discharge of approximately 1,600 cubic metres per second for the bridge, associated with a 100-year flood while accounting for climate change. [3] However, this figure refers to the particular structure and its design assumptions. It is not a general value for the 100-year flood throughout the Savinja River.
Nor does a stated bridge conveyance capacity mean that the entire city is protected against every possible future flood. The river channel, embankments, tributaries, drainage from surrounding land and conditions both upstream and downstream also matter. In January 2026, the project was still under way, with completion then anticipated in June of the same year. Without a separately verified completion record, that expected date must not be presented as proof that the structure was actually completed and accepted into service.
Laško and the Confluence with the Rečica
The EU-cohesion-funded project to reduce flood risk in the Savinja basin also includes planned works around the lower reaches and confluence of the Rečica near Laško. The plan includes a flood-protection wall, modifications to road connections, a new bridge, mobile flood barriers for certain underpasses and improved drainage of water arriving from surrounding land. [2] At the time of the April 2026 publication, the construction contractor was still being selected through a public procurement procedure. The project therefore cannot be treated as flood protection that had already been completed at that time.
The Laško example highlights another important point. Flooding is not always caused solely by high water levels in the main river. Water may also reach threatened areas from tributaries, surrounding land and drainage infrastructure that can no longer function normally during high water.
What Counts as the Actual Implemented State?
A reliable assessment of present flood protection requires more than a list of projects. It requires information about the works actually carried out, their completion dates, the operational status of structures and the geometry of watercourses after construction. Individual components of a project may be complete while the overall project remains under construction. Such intermediate conditions require a separate assessment of the protection already in place.
The distinction is therefore essential:
A project under construction, a completed structure and a hydraulically verified flood-protection system are not the same thing.
What Is the Government Planning for the Future?
Official project map of modelled Q100 inundation changes with proposed flood-protection measures.
An important part of long-term flood-risk management involves a system of interconnected national spatial plans. On 22 December 2025, a proposal for extensive flood-protection works in the Lower Savinja Valley was opened for public review. The plan addresses an approximately 23-kilometre section of the Savinja between Letuš and Celje, together with the wider areas of the Bolska, Gozdnica and Ložnica watercourses. The planned intervention area covers approximately 1,078 hectares across seven municipalities. [4] [5]
This figure describes the area covered by the planning proposal. It must not be confused with the storage capacity of the proposed retention basins or with the total area of land that would actually be flooded during a particular event. According to official estimates, the proposed measures are expected to substantially reduce flood exposure for approximately 3,730 buildings and improve safety for around 13,300 residents. These are significant figures. However, they represent projected benefits of a fully implemented system, not evidence that this level of protection has already been achieved.
The same applies to statements by the project designers that, during a 100-year flood, the area of inundated agricultural land would decrease by approximately 1,000 hectares, or by around 60 percent of the agricultural land considered in the assessment. This is a published modelling estimate or project-level assessment that must be interpreted in the context of its hydrological assumptions and the specific version of the plan. [5]
What Does the Planned System Include?
The principal proposed measures include dry flood-retention basins at Levec, Dobriša vas, Roje, Šempeter, Dobrteša vas, Latkova vas and Letuš along the Savinja, as well as Kaplja vas along the Bolska. A multipurpose retention basin at Gozdnica is also planned, intended to provide both flood protection and opportunities for agricultural irrigation. The proposal further includes flood embankments, improvements to river channels and tributaries, and adaptations to bridges, roads and utility infrastructure. [5] Importantly, the official proposal is not limited to embankments and concrete walls.
It also incorporates the preservation and activation of floodplain areas, adaptation of river space and other approaches connected with natural processes. It would therefore be inaccurate to portray the government plan as fundamentally opposed to nature-based river management. The actual question is whether its particular combination of measures is sufficiently effective, socially and spatially acceptable, and sustainable in the long term.
How Much Water Would the Retention Basins Actually Store?
During the public discussion in January 2026, a combined planned storage capacity of approximately 10.99 million cubic metres was presented, potentially increasing to approximately 12.7 million cubic metres when additional retention areas were included. [6] We must distinguish between the originally presented storage capacity and a potentially expanded version. We must also identify the version of the project design to which these figures apply. Above all, storage capacity is not the same as a reduction in peak discharge.
Millions of cubic metres describe volume, while cubic metres per second describe discharge. These are two different physical quantities. The operation of a retention basin depends on the complete time-dependent pattern of inflow and outflow. If most of the water arrives within a short period, the basin may fill differently than it would during a prolonged flood wave containing the same overall volume. The timing of flood peaks along the Savinja and its tributaries also matters, as does the moment at which stored water is released.
The system must therefore be assessed as an interconnected whole. A larger total storage volume does not, by itself, guarantee a proportionally greater reduction in damage.
What Is a Dry Flood-Retention Basin?
A dry flood-retention basin is an area that normally contains no permanent reservoir and may permit certain forms of ordinary land use. During major floods, however, it is intended to store water temporarily. When properly designed, it can attenuate the flood wave and reduce peak discharge along critical downstream sections. But the water does not disappear. After being retained, it must leave the basin according to the intended operating regime.
Poorly timed operation or an unfavourable combination of flood waves can reduce the expected benefits. This is precisely why we need the results of coordinated hydrological and hydraulic modelling, rather than relying solely on the capacities of individual structures.
Planning in the Upper Catchment
The national approach also includes planning along the Savinja between Mozirje and Ljubno, within the Dreta basin and in other connected areas. In May 2026, Slovenia's Spatial Information System still listed the spatial plan for the Savinja between Mozirje and Ljubno as being in preparation. [7] In September 2026, the Ministry also announced preliminary archaeological investigations around Kaplja vas as part of the preparation of the national spatial plan for the Lower Savinja Valley. [12] This confirms that preparatory procedures were continuing.
It does not, by itself, establish that the entire plan had already been adopted or that all proposed retention-basin locations had received final approval. Each proposal must therefore be identified as an option under consideration, a publicly exhibited planning version or a legally adopted spatial plan.
Comprehensive flood protection requires an understanding of the entire river basin, but the existence of a comprehensive plan is not the same as its implementation.
How Much Protection Can Be Guaranteed?
Official flood-hazard maps for different exceedance probabilities.
Flood-protection projects frequently use terms such as the 100-year flood, flood safety and design discharge. These expressions can create the impression that once a planned system is complete, the flooding problem will have been solved for the next hundred years. That is not what they mean. A so-called 100-year flood, commonly represented by Q100, is a statistical estimate of a discharge with an approximate one-percent annual probability of exceedance, under the assumptions of the analysis used. It does not mean that such a flood can occur only once every hundred years.
Nor does Q100 have the same discharge value at every point along a river. The discharge at one gauging location differs from that at another location or tributary. Estimates also depend on the available observations, methods and hydrological conditions being considered. As the climate changes and new observations become available, some assumptions must be reassessed. Protection designed for Q100 is therefore not a promise that flood danger has disappeared.
It represents a particular design level for reducing risk. It is also important to distinguish between the resilience of an individual structure and the flood safety of a wider area. A bridge may convey its design discharge while water still floods somewhere else. An embankment may protect one part of a floodplain while influencing water levels or flow paths in neighbouring areas. A retention basin may successfully attenuate the peak of one flood wave but have a different effect under another rainfall pattern.
During extreme events, sediment accumulation, woody debris, riverbank damage and other changes can affect the operation of infrastructure. For this reason, residual flood risk must be assessed alongside the intended design level of protection.
What happens if the expected flood conditions are exceeded?
What happens if an embankment is damaged or woody debris accumulates beneath a bridge?
Will evacuation routes remain usable?
Will residents receive warnings in time?
These questions do not represent distrust of hydraulic engineering. They are an essential part of responsible flood-risk management. Residents need to understand not only which events a protection system is designed to withstand, but also what they must do when its capacity is no longer sufficient.
Who Receives Protection, and Who Bears the Burden?
Flood-protection projects have social as well as technical consequences. Reducing risk for people living in densely populated areas is an important public objective. But water that is to be held back before reaching a settlement must be given space somewhere else. The proposed retention basins in the Lower Savinja Valley have therefore raised the question of how much of the burden of protecting settlements should fall on farmland and its owners. In January 2026, farmers and their representative organisations warned about the potential consequences of the proposed measures for food production, hop farming, perennial crops and the long-term survival of agricultural holdings.
At an event organised by the Chamber of Agriculture and Forestry of Slovenia, a farmers' representative stated that the plan could substantially affect 394 agricultural holdings, including 35 farms with young successors. [8] These figures represent a publicly expressed estimate or concern raised during the discussion. They do not mean that an equally severe loss or risk has been independently established for every farm. Concerns also involved temporary flooding of fertile land, sediment deposits, access routes, utility infrastructure, groundwater and the protection of drinking-water sources.
Such questions cannot be dismissed simply by pointing out that dry retention basins usually contain no standing water. If land is designated for temporary floodwater storage, the consequences for its continued use must be understood. It must also be clear who will bear the costs of possible losses, clean-up work and restrictions on land use. However, the opposite claim must not be accepted without scrutiny either. The mere fact that land lies within a proposed retention-basin area does not mean that it will flood every year or that agricultural production will necessarily become permanently impossible.
Such an assessment requires data about the expected frequency, depth and duration of flooding, sediment composition, remediation procedures and restrictions on land use.
The Perspective of Residents in Flood-Prone Settlements
On the other side are residents of settlements that were heavily flooded in August 2023. In February 2026, the Ločica-Breg citizens' initiative expressed support for continuing the planning process and warned about the danger of delaying flood-protection measures. At the same time, it criticised aspects of the public consultation process concerning the proposed national spatial plan. [9] This distinction matters. The initiative did not express general opposition to flood-protection planning. Its criticism concerned the circumstances of the public discussion and the conduct of certain participants.
In its view, moving the project forward is important for protecting homes and residents. That concern is understandable. People who lost their homes or faced immediate danger are entitled to expect known risks to be reduced. Yet a proposed solution cannot be assessed solely through the benefits received by one group or the costs borne by another. A meaningful comparison requires information about how many buildings and areas will be protected, how much land will experience more or less flooding, what the groundwater impacts will be and who will bear the associated costs.
It is especially important to distinguish between areas where the plan reduces flood exposure and land deliberately designated for temporary water storage. In December 2025, the government's project designers also presented an estimate that, during a 100-year flood, the area of agricultural land inundated within the study area would be substantially reduced. [5] This is an important counterargument to the assertion that the plan as a whole must necessarily be harmful to agriculture. However, the projected overall benefit does not resolve the consequences for individual farms and plots within the proposed retention areas.
The overall benefit to a region and the distribution of losses among individual landowners are two different questions. A fair decision must address both.
Giving space back to a river is not merely a hydraulic decision. It is also a decision about property, livelihoods and the future of local communities.
Three Scenarios for the Savinja's Future
Official schematic overview of measures planned across the wider Savinja catchment.
To compare the available options transparently, we must first define a reference condition. We can call it S0. This is the condition against which subsequent changes are assessed. For example, S0 could represent the reconstructed state of the river system immediately before the August 2023 floods, if the purpose is to determine how restoration and new interventions have changed flood behaviour across the basin. But S0 is not a fourth proposed flood-protection system.
It is a reference baseline. From this starting point, we can define three principal scenarios.
S1 — Measures Already Implemented
The first scenario represents conditions with the structures and interventions that have actually been completed and are functional as of a clearly specified cut-off date. It must include documented changes to river channels since 2023, completed restoration work and finished flood-protection measures. It does not include elements that remain unbuilt. If a structure is only partly constructed or has not been shown to be operational, its status must be addressed separately. S1 is therefore not the state that existed before the 2023 floods.
It is the latest verifiable condition on the ground. Modelling it would require verified information about the geometry of watercourses, embankments, bridges, retention areas and other relevant structures. Only on that basis could we assess how selected flood events would develop under the conditions that have actually been established.
S2 — Full Implementation of Planned Government Measures
The second scenario represents a hypothetical condition in which all measures contained in a precisely identified version of the national plans have been implemented. It includes retention basins, river-channel works, embankments, bridges, floodplain areas and other interconnected measures. For S2, the exact planning document or design version must be specified. The proposal published in December 2025 must not automatically be treated as an unchangeable final solution if it was subsequently amended during the spatial-planning process.
Under the same selected hydrological event, modelling would assess how water levels, flow velocities, the extent of flooding in settlements and other important consequences differ from S1. Special attention would be required to evaluate retention-basin operation under different rainfall patterns and flood-wave timings. The consequences of basin filling, conditions beyond the design assumptions and partial operational failure should also be examined. S2 must not be an idealised assumption that every component will always operate perfectly.
It is also important that S2 includes all nature-based and spatially adapted measures already incorporated into the selected official planning version. Only then can a subsequent comparison be fair.
S3 — Government Measures with Additional, Testable Improvements
The third scenario represents S2 supplemented by additional solutions that merit separate technical testing. Possible measures include further restoration of connections between rivers and their floodplains, carefully selected spatial adjustments, improved sediment management and particular forms of local flow-guiding structures. Among the latter, some contemporary engineering approaches related to Schauberger's historical ideas may also deserve investigation. Their inclusion in S3, however, does not mean that they are more effective.
Scenario S3 is not inherently safer, cheaper or more environmentally appropriate than S2.
An additional intervention may improve conditions along one riverbank while increasing erosion elsewhere. It may improve fish passage without affecting the flood peak. It may reduce flooding in a settlement while requiring additional land for temporary water storage. Every proposed modification must therefore be clearly defined and tested. Avoiding double-counting is particularly important.
If S2 already includes a floodplain or a nature-based intervention along a particular river section, that same measure must not be presented again as a new improvement under S3. S3 must contain genuinely additional or modified interventions. The comparison must reveal their added value, not merely different names for similar solutions.
What Should Hydraulic Models Demonstrate?
To compare S1, S2 and S3 scientifically, the scenarios would need to use the same underlying hydrological conditions. This means comparable rainfall events, consistent assumptions about initial catchment saturation, appropriate inflows and clearly documented calculation methods. The models should first reproduce known flood events with sufficient accuracy. An important test would be the reconstruction of August 2023. It would need to account for the fact that river channels changed during the event and that part of the water flowed outside ordinary gauging cross-sections.
Model reliability should therefore also be evaluated against observed high-water marks, documented inundation areas and recorded damage. Several different hydrological conditions would be required to compare future scenarios. These might include an event similar to August 2023, design floods with different return periods and relevant climate scenarios. However, the comparison must not be based solely on water levels in Celje. It should also examine flooding in other settlements, flow velocities, erosion loads, sediment deposition, effects on tributaries and the operation of retention basins.
Two-dimensional hydraulic models play an important role in such analyses. For certain complex local situations, three-dimensional simulations or physical laboratory experiments may also be useful. The essential requirement is comparability. If S3 is tested with different inflow conditions from S2, any difference in results cannot be attributed directly to the additional measure. The same problem arises if sediment transport and debris accumulation are considered in one scenario but excluded from another. The investigation should address six principal outcomes.
First: the actual effect of the measures on flood hazard in individual settlements. Second: any increase in hazard elsewhere, both upstream and downstream. Third: sensitivity to sediment, woody debris and changes in river-channel geometry. Fourth: the resilience of the system when design conditions are exceeded or individual components fail to function as expected. Fifth: construction and maintenance costs, together with consequences for agriculture, ecosystems and infrastructure.
Sixth: the additional benefits of proposed improvements compared with the system already planned. Uncertainties should also be presented alongside the results. A flood map is not a prophecy. It is the output of a model based on specific data and assumptions. Our investigation into Viktor Schauberger established that some forms of flow-guiding structures are technically testable.
But discussion of his legacy cannot substitute for modelling. Flow-retarding and guiding elements described in historical patents, together with later meandering ramps and groynes, can influence local hydraulic conditions. For the Savinja, their stability, erosion effects, impacts on aquatic organisms and behaviour during extreme discharges would have to be assessed separately. Without such evidence, we cannot claim that a Schauberger-inspired structure would reduce the flood peak throughout the entire valley. Nor can we rule out in advance that some of these ideas might be useful for a particular local problem.
A scientifically sound answer depends on the results of a specific test.
What Do We Know About Existing Models, and What Remains Unknown?
Publicly available government documentation confirms that planning has been supported by hydrological and hydraulic studies, an assessment of alternatives and an environmental report. In December 2025, documentation was published for public consultation that included project-level estimates of expected reductions in flood exposure. [5] [11] It would therefore be wrong to claim that no technical analyses have been carried out. However, the publicly available summaries do not demonstrate that an assessment directly equivalent to the S1, S2 and S3 comparison proposed in this investigation has been completed.
Some technical documents have been made publicly available. For others, the public record primarily establishes their existence or use within the planning process. Where we have not examined a document's complete contents, we will not present it as though we had. Likewise, without appropriate modelling results, we will not invent percentages for reductions in discharge, inundated areas or financial losses. The S1, S2 and S3 scenarios are therefore, at this stage, a proposed framework for transparent comparison, not the findings of a new hydraulic simulation.
How Can We Live with a River That Will Flood Again?
No flood-protection system can guarantee complete safety against every possible future event. That does not make flood-protection measures pointless. It means that their purpose must be to reduce risk and consequences while acknowledging the limitations of even the best engineering solutions. For the Savinja Valley, much more will be required than the completion of individual embankments and retention basins. There will need to be maintenance, monitoring of riverbed changes, inspections of structures, updates to flood-hazard maps and preparation for events that exceed design assumptions.
Clear public warnings, safe evacuation routes and responsible land-use planning will also be essential, so that future development does not unnecessarily increase exposure. Where feasible, we must consider how to preserve or restore the space rivers need for their natural dynamics. But this question must remain grounded in actual conditions. Not every river-channel widening project is feasible. Not every floodplain is suitable for additional water retention.
Not every nature-based structure is sufficiently stable for every watercourse. And not every objection to a particular intervention reflects a failure to understand flood risk. In a complex valley where settlements, agriculture, transport, industry and aquatic ecosystems coexist, there is no universal solution. There are, however, opportunities for better decision-making. Public discussion can become more transparent if measures are compared using the same data, if their consequences are assessed across their full area of influence and if the limitations of models are explained openly.
It is equally important that residents understand the difference between measures already implemented, promised protection and plans that have not yet been formally adopted or completed. Otherwise, a dangerous sense of security may develop that is not justified by conditions on the ground.
What Remains Unresolved?
Based on the public sources examined, we can establish that extensive recovery work began after 2023 and that major interconnected flood-protection measures have been proposed. However, without further verifiable information, we cannot claim that the Savinja Valley is already protected against a recurrence of the August 2023 event. Nor can we claim that the proposed government system is the only technically possible combination of measures, or that alternatives would necessarily perform better. A responsible conclusion requires comparison.
That comparison must account for residents' safety, agriculture, ecosystems, costs and long-term maintenance.
Conclusion — We Cannot Command a River, but We Can Understand It Better
In the first article, we examined how the Savinja's landscape has changed throughout history and why everything built along the river becomes important during extreme floods. In the second, we investigated Viktor Schauberger, his documented patents and attempts to direct water flow, and modern hydraulic research capable of testing related ideas. The third article brings us to a decision more important than any individual historical theory.
How will we act before the next major flood?
We cannot protect the Savinja Valley merely by hoping that August 2023 will never happen again. Nor can we protect it through the romantic belief that a single forgotten invention could solve every problem. We need a comprehensive understanding of the catchment, effective and tested interventions, and a willingness to improve existing solutions whenever the evidence supports doing so. What matters is not whether an idea was first proposed a century ago or only recently. What matters is whether it works, where it works, under what conditions and what consequences it produces.
A river does not follow our expectations, plans or promises. It responds to rainfall, terrain, discharge, sediment and the space available to it.
If we understand these conditions better and take them seriously in our decisions, the next great flood may still cause the Savinja to overflow its banks. But its consequences may be less severe. That is a goal worth pursuing.
Sources and further reading
- Slovenian Environment Agency (ARSO). Izjemne poplave v Sloveniji med 4. in 8. avgustom 2023 [Exceptional Floods in Slovenia, 4–8 August 2023]. Official hydrological report. Source 1 Source 2
- Water Directorate of the Republic of Slovenia (2026). Operacija »Zmanjšanje poplavne ogroženosti na porečju Savinje – 1. sklop« v izvajanju [Flood-Risk Reduction in the Savinja Basin — Phase 1 Under Way]. 16 April 2026. Source
- Ministry of Natural Resources and Spatial Planning and Water Directorate of the Republic of Slovenia (2026). Enormna vlaganja v poplavno varnost Savinjske regije [Major Investments in Flood Protection in the Savinja Region]. 22 January 2026. Source
- Water Directorate of the Republic of Slovenia and Ministry of Natural Resources and Spatial Planning (2026). DPN Spodnja Savinjska dolina: celovite rešitve za zmanjšanje poplavne ogroženosti [Lower Savinja Valley National Spatial Plan: Comprehensive Flood-Risk Reduction Measures]. 16 January 2026. Source
- Ministry of Natural Resources and Spatial Planning (2025). DPN za zmanjšanje poplavne ogroženosti v Spodnji Savinjski dolini – zaščita ljudi, naselij in kmetij [National Spatial Plan for Flood-Risk Reduction in the Lower Savinja Valley — Protecting People, Settlements and Farms]. 22 December 2025. Source
- Slovenian Press Agency (STA), published by N1 (2026). Kmetje zahtevajo takojšen umik državnega prostorskega načrta za Savinjsko dolino [Farmers Demand Immediate Withdrawal of the National Spatial Plan for the Savinja Valley]. 19 January 2026. Source
- Republic of Slovenia, Spatial Information System. Državni prostorski načrt za zmanjšanje poplavne ogroženosti na porečju Savinje od Mozirja do Ljubnega ob Savinji [National Spatial Plan for Flood-Risk Reduction Along the Savinja from Mozirje to Ljubno]. Spatial plan under preparation, ID 7479. Source
- Chamber of Agriculture and Forestry of Slovenia (2026). Gradnja suhih zadrževalnikov ogroža obstoj kmetij v Spodnji Savinjski dolini [Construction of Dry Retention Basins Threatens the Survival of Farms in the Lower Savinja Valley]. 7 January 2026. Source
- Slovenian Press Agency (STA), published by Velenjčan (2026). Civilna iniciativa kritična do razgrnitve prostorskega načrta za poplavno varnost Savinjske doline [Citizens' Initiative Criticises Public Consultation on the Savinja Valley Flood-Protection Plan]. 7 February 2026. Source
- Water Directorate of the Republic of Slovenia. Zmanjšanje poplavne ogroženosti na porečju Savinje – 1. sklop [Flood-Risk Reduction in the Savinja Basin — Phase 1]. Official project register; scheduled completion 31 December 2028. Source
- Ministry of Natural Resources and Spatial Planning and Water Directorate of the Republic of Slovenia (2026). Javna razprava o DPN Spodnja Savinjska dolina: dialog za večjo poplavno varnost [Public Discussion of the Lower Savinja Valley National Spatial Plan: Dialogue for Greater Flood Safety]. 20 January 2026. Source
- Ministry of Natural Resources and Spatial Planning (2026). Državni prostorski načrt za zmanjšanje poplavne ogroženosti v Spodnji Savinjski dolini [National Spatial Plan for Flood-Risk Reduction in the Lower Savinja Valley]. Announcement concerning preliminary archaeological investigations during preparation of the plan, 2 September 2026. Source
- Ministry of Natural Resources and Spatial Planning. Državni prostorski načrt za zmanjšanje poplavne ogroženosti v Spodnji Savinjski dolini [National Spatial Plan for Flood-Risk Reduction in the Lower Savinja Valley]. Publicly available alternatives assessment, environmental report and draft spatial plan; public consultation from 22 December 2025 to 6 February 2026. Source