As of September 14, 2026, the Danube is not recovering uniformly along its Romanian stretch. Operational data from the Lower Danube River Administration (AFDJ) show stabilisation or slight rises at some downstream points, but at Baziaș—the reference station for where the river enters Romania—the level remained at 540 cm, the same as on September 13. In the south-western section, from Calafat to Corabia, levels continued to fall.
For more information on Cernavodă, see Romania left without nuclear power: what happens after both Cernavodă reactors were shut down, which provides complementary details on this subject.
The claim that the Danube is “starting to recover” must therefore be read with an important qualification: it refers to localised improvements in water levels in some places, not confirmation that discharge has recovered along the entire Romanian stretch. For river transport, the distinction matters directly, as vessels depend on the depth and continuity of the navigable channel, not merely on the day-to-day change at a single water gauge.
What the September 14 data show: the river is changing differently from one section to another
The AFDJ bulletin of September 14, 2026 allows a straightforward comparison of levels across four consecutive days. The picture is mixed: Baziaș was unchanged, Cernavodă rose by three centimetres compared with September 13, while at many points between Calafat and Corabia the level lost a few more centimetres.
| Station | September 11 | September 13 | September 14 | Change over the past day |
|---|---|---|---|---|
| Baziaș | 548 cm | 540 cm | 540 cm | unchanged |
| Calafat | -184 cm | -190 cm | -191 cm | -1 cm |
| Corabia | -149 cm | -151 cm | -153 cm | -2 cm |
| Giurgiu | -191 cm | -192 cm | -190 cm | +2 cm |
| Cernavodă | -240 cm | -245 cm | -242 cm | +3 cm |
| Brăila | -15 cm | -16 cm | -15 cm | +1 cm |
| Tulcea | 7 cm | 10 cm | 11 cm | +1 cm |
Note: Levels are measured against the local “zero” datum at each station. They should therefore not be compared between localities as though they represented the same physical elevation of the water. The relevant comparison is the change at the same station from one day to the next.
At Baziaș, the level of 540 cm on September 14 followed 548 cm on September 11, 544 cm on September 12 and 540 cm on September 13. In other words, the decline stopped temporarily, but this data set contains no indication of a rise where the Danube enters the country.
Further downstream, the situation is not the same everywhere. Cernavodă rose from -245 cm to -242 cm, Giurgiu from -192 cm to -190 cm, and Brăila from -16 cm to -15 cm. By contrast, Calafat fell to -191 cm, Rast to -182 cm, Bechet to -124 cm and Corabia to -153 cm. This is why the phrase “the Danube is recovering” oversimplifies a fragmented hydrological reality.
Water level, discharge and depth: three indicators that do not mean the same thing
Reports about the river often use the terms “level”, “gauge height” and “discharge” interchangeably. They describe different things.
- Water level is the level measured at a station, relative to its local datum. It is the indicator updated frequently in AFDJ bulletins.
- Discharge is the volume of water passing through a cross-section in one second, expressed in cubic metres per second. At Baziaș, this is the indicator hydrologists use to describe the quantity of water with which the Danube enters Romania.
- Navigable-channel depth indicates how much water is actually available along the shipping route. For vessels, this is the decisive practical indicator.
A water-level rise of one or three centimetres can be useful, especially during a period of exceptionally low water. But on its own, it does not mean that discharge has recovered, nor that a barge can load to normal capacity. Riverbed shape, sandbanks, maintenance works and the distribution of water between branches can produce different effects from one section to another.
The same distinction matters when comparing these figures with information from late August. On August 29, rises of up to nine centimetres were reported at some stations in the Bulgarian section, but not at Ruse or Silistra, where the water level was still falling. Navigation restrictions remained in place at that time. Data published on August 30 on the Bulgarian section illustrate the same rule: local rises are not equivalent to normalisation across the entire river.
Why navigation does not return to normal after a rise of a few centimetres
In its September 14 bulletin, AFDJ maintains a long list of points where minimum depths are below four metres. On certain sounding routes, the reported minimum values are as little as 1.1–1.5 metres in the Călărași–Cernavodă area and 1.2–1.8 metres at several points in the south-western section.
These figures explain why the issue cannot be reduced to the question, “Has the Danube risen or fallen?” In low-water conditions, carriers must reduce cargo loads, adjust draught, alter routes or wait for better conditions. A change of a few centimetres may improve conditions locally without removing the channel’s bottlenecks.
A note of caution is also needed in interpreting the figures. The AFDJ table on depths below four metres is updated using the reference level of September 13, but the sounding data for the listed points carry different dates. The document is highly useful for identifying vulnerable sections, but should not be read as a real-time, minute-by-minute map of depth at every point.
What stabilisation at Baziaș means for Romania
On September 10, information from the Calafat area indicated a discharge of 1,400 m³/s at Baziaș, compared with the September multiannual average of 3,800 m³/s. A possible fall towards 1,300 m³/s was forecast for September 14. That forecast described an exceptionally difficult period, not a rapid return to normal values.
The AFDJ bulletin of September 14 does not publish a new discharge value at Baziaș in the same format; instead, it provides station levels and local trends. It is therefore accurate to say that the level at Baziaș stabilised from September 13 to 14, not that discharge rose. Confirming a recovery in discharge requires an updated hydrological measurement for the same cross-section.
For Romania, the positive signal is limited but still relevant: Cernavodă, Brăila, Galați and Tulcea show slight increases or stabilisation. The signal for caution is stronger: Baziaș did not rise, while the Calafat–Corabia section remains in decline. In addition, AFDJ estimated for the following two days a fall of 1–3 cm at Giurgiu, no change at Cernavodă and a rise of 1–3 cm at Brăila—further evidence that there is no uniform trend.
How to monitor future developments correctly
Readers interested in transport, fishing, port activity or conditions in riverside communities should follow three categories of data separately:
- Discharge at Baziaș, to understand the quantity of water entering Romania.
- Daily AFDJ water levels, to track local changes at Romanian stations.
- Navigation notices and channel depths, to assess the actual effect on navigation.
The AFDJ page for Danube water levels remains the operational reference for Romanian stations. For context on the crisis of recent weeks and the very low values forecast at Baziaș, readers can also consult InfoBuz’s analysis of the Danube’s low discharge in August.
The conclusion from the September 14 data is a nuanced one: there are places where water levels have begun to regain a few centimetres, but these are not enough to declare that the Danube has moved beyond its low-water period. Until discharge at Baziaș and depths at critical points show sustained improvement, restrictions and navigation difficulties will remain a practical problem.
Frequently asked questions
Did Danube discharge rise at Romania’s entry point on September 14, 2026?
AFDJ operational data from September 14 show that the level at Baziaș remained at 540 cm, unchanged from September 13. The bulletin does not provide a new discharge value in m³/s, so it does not confirm an increase in discharge at the country’s entry point.
Why can the Danube level rise at Cernavodă if it is not rising at Baziaș?
Water propagates differently through different sections, and the local level is influenced by riverbed shape, tributaries, branches, hydraulic works and downstream conditions. A change at one station is not automatically or identically transferred to all others.
Does a rise of a few centimetres mean vessels can operate normally?
No. Navigation depends on usable depth and the continuity of the navigable channel. AFDJ continued to report critical points with minimum depths below four metres, so a small local rise does not automatically remove restrictions.
