From an Article to Our Own Index – Simply Explained
On 20 August, CORRECTIV published the Fluss-Atlas (river atlas): an investigation into low flow at 356 German gauges. Low flow is the state where a river carries significantly less water than usual. Gauges are the measuring stations along rivers that record water level and discharge every day.
The result: at 214 of those gauges the annual low point fell in a statistically significant way between 1992 and 2025. Statistically significant means the decline is large enough that it is very likely not coincidence. At 240 gauges the lowest value of the entire measurement series falls in the window from 2018 to 2025. The article convinced me.
Whether the calculation behind it was any good, I could not judge myself. I put the article to Claude and asked. According to Claude, the newsroom had run a careful test: one that accounts for the fact that in such series each year hangs on the one before. That is more work than the obvious approach. The article also answers the question of why any of this concerns people: drinking water, shipping, a Rhine gauge at Kaub reading six centimeters.
What bothered me about the map
The article comes with an interactive map of Germany carrying 356 colored dots. I looked at it for a while and could not read it.
The color legend shows percentage values, but they are not percentages in the usual sense. The label “per year” is missing. Anyone looking at the map without prior knowledge does not learn what the numbers refer to.
The dot size grows with the size of the catchment. A catchment is the total area from which rainwater flows into a particular stretch of river. To the eye, a large red circle first of all means badly affected. But the size only says how much area the gauge drains.
What bothered me most was the Rhine chain. Maxau, Worms, Kaub, Andernach, Cologne, Duisburg and Rees all sit on the same river. Each gauge further downstream largely measures the same water as the one above it, plus the tributaries in between. On the map that turns into half a dozen large, overlapping red circles. The west looks more broadly affected than an independent sample would show it.
The alarm-red look was not even necessary here. The finding holds up when you recalculate it.
What we built from it
I asked Claude what we could do with the same data ourselves. The basis lies open: the Federal Institute of Hydrology publishes daily discharge values through a public interface.
Over the following working days an analysis of our own took shape: 361 gauging stations, around 4.7 million daily values, water years 1992 to 2025. A water year runs not from January to December but from November to October, so the summer low-flow period is not split by the calendar year.
Two decisions set our analysis apart from the Fluss-Atlas.
The first concerns the Rhine chain. We chained the gauges along the direction of flow and credited each one only with the area it drains in addition to the gauges above it. The same water counts only once. A gauge with a catchment of 3 square kilometers weighs less than one with 159,000.
The second concerns what we measure. The lowest discharge of a year captures only the depth of the low point. We additionally count how many days a gauge spends below its own long-term low-flow threshold, how long the longest continuous spell ran, and how much water is missing across the year. The difference is considerable: the median number of low-flow days per gauge and year rises from five or six days in the period 1992 to 2013 to 45 days in the period 2014 to 2025.
For the statistical test we deliberately followed CORRECTIV in order to stay comparable. 224 of 359 gauges show a significant increase in low-flow days. Because that many parallel tests also produce chance hits, we additionally corrected. That leaves 211.
The shares land close to those of the Fluss-Atlas. In one place almost exactly: in our analysis the lowest value of the series falls in the 2018 to 2025 window at 69.6 percent of gauges. CORRECTIV gives around 68 percent.
What the page shows
The analysis is now in this site’s menu, in German and English.
The image at the top of this post is its centerpiece. Each row is a water year, each column a day, and the brightness stands for the share of gauges that were below their threshold on that day. You need no statistics to see that toward the lower rows the bright bands start earlier in the year, last longer and grow wider.
Below it sit time series with toggles, distributions across decades, a station map, and a methodology section that lays every number open for checking. 2018 is the driest year of the series, followed by 2019 and 2022. Seven of the ten driest years fall in the period 2015 to 2025.
The code of the analysis is public: AeternaLabsHQ/niedrigwasser-trendindex. Anyone who wants to can run the whole calculation again themselves.
What the analysis does not deliver is stated on the same page. It assigns the trend to no cause. About the future it says nothing. And the measuring network has gaps: from the state networks of Lower Saxony, Bremen and Hamburg, not a single station is included.
There is one point where CORRECTIV is clearly ahead, and I do not want to pass over it. The article tells you what the finding means for drinking water, shipping and fish. Our page explains carefully what was measured, and nowhere answers why it matters.
While proofreading the finished page, I came across a formulation that gave me pause. The methodology said the year 1991 had been excluded because it distorted the trend. That is the wrong reason, since it would let you strike any inconvenient year. The right one lies in the water year: because it starts on 1 November, water year 1991 begins on 1 November 1990. The measurement series, however, only begins on 1 January 1991. The first year is therefore missing its first two months; it has 304 days instead of 365. We swapped the justification.