Skip to content
Watchberg
Articles

8 October 2026/5 min read

What the Birch Glacier told satellites nine months before Blatten

L-band radar saw the Kleines Nesthorn accelerating from 2016. The signal was there; the system to read it across every slope was not.

By Julian Walder

On 28 May 2025, about 10 million cubic metres of rock and ice broke away from the Kleines Nesthorn and the Birch Glacier above Blatten in the Lötschental, Valais. The mass ran into the valley, buried most of the village and dammed the Lonza river. Roughly 90 percent of the buildings were destroyed or severely damaged. One person was reported missing. The Swiss Insurance Association put the total damage at about CHF 320 million, of which about CHF 255 million was insured.

The village itself was empty. About 300 residents had left their homes on 19 May, nine days earlier, after a local natural-hazard observer raised the alarm. The Canton of Valais runs a network of geologists, about 90 local observers, monitoring instruments and contingency plans for exactly this purpose, and on that day the system worked.

Nine days of ground radar

Once the observer's alert came in, the response was fast. A ground-based interferometric radar was installed on 19 May, the same day residents left. A camera followed on 27 May. Those instruments tracked the mountain through its final acceleration: in the last hours before the collapse the moving mass was travelling at more than 10 metres per day.

Ground radar is the right tool for that phase. It measures the slope every few minutes, it works at night and through cloud, and it gives the hours of warning that let a canton decide to close a road or keep people out. What it cannot do is watch a slope nobody has asked it to watch. It arrived at Blatten nine days before the end, because that was when someone knew to ask.

What the satellites had recorded

The more uncomfortable part of the story came afterwards. In August 2025, ESA published the results of a retrospective analysis by its MODULATE project, a FutureEO research study by Swiss and Italian research groups whose findings appeared in Natural Hazards and Earth System Sciences in June 2026. Using archived L-band radar from the ALOS-2 and SAOCOM satellites, the team found that the Kleines Nesthorn had been moving since 2016 or 2017. By 2023 the displacement was about 50 centimetres per year. By August 2024 it exceeded 150 centimetres per year: three times the rate of the year before, nine months before the collapse.

A separate study, published in 2026, looked at the same slope with C-band data from the Sentinel-1 satellites and a small-baseline (SBAS) time-series method. It found more than 60 millimetres of cumulative line-of-sight displacement, with a marked acceleration after April 2025 and exponential growth by the last usable acquisition on 21 May 2025.

Two independent satellite systems, at two wavelengths, had recorded the same slope changing its behaviour. Neither record was being read as a screening signal at the time.

Why nobody was looking

This is not a story about negligence. The Birch Glacier and the Kleines Nesthorn had been observed since the 1990s. The problem is structural. ESA's own write-up of the result states plainly that instrumenting every alpine slope with ground sensors is logistically and financially out of reach. Satellite radar can see millimetre-to-metre motion across whole regions, but in practice it has been used retrospectively, or manually and project by project, by specialists who already know which slope to look at.

The result is a kind of selection bias. The slopes that get watched closely are the ones someone already worries about. A slope that is accelerating quietly, outside anyone's project, generates data that nobody opens.

The researchers behind the L-band analysis said as much in their own terms: the methods work, but they need refining, and the data must become more accessible so that decision-makers can use it in time.

The signal was there for nine months. What was missing was a system that screens every slope and tells an expert where to look first.

What a screening layer is for

Watchberg is built around that gap. It screens whole alpine regions on a weekly to monthly cadence with Sentinel-1 C-band radar, NISAR L-band radar and Sentinel-2 optical imagery, measures the motion of thousands of slopes, and flags the moment a slope's rate of movement departs from its own history. Each flagged slope is ranked by what lies below it and handed to the responsible geologist or operator as a short, sourced briefing: what changed, the evidence with dates and datasets, the confidence and the data gaps, and a recommended next step such as an expert review, a drone survey or ground radar.

Applied to the Kleines Nesthorn, the published record suggests that a change from about 50 to more than 150 centimetres per year on a slope directly above a village would have ranked near the top of any regional list by late summer 2024. Whether an automated pipeline, using only the data available at the time, would have flagged it that early is the first thing Watchberg has to prove. That back-test, with no look-ahead, is part of the pilot programme planned for 2027, and every performance claim remains projected until it is done.

The honest limit

Satellites do not predict the hour of a collapse. The final acceleration at Blatten, from metres per day to more than 10 metres per day, was measured by ground radar, and that is where it should be measured. What satellites can offer is earlier notice, measured in years to weeks, that a slope deserves attention. The decision about what to do with that notice belongs to the cantonal geologist, the commune and the operator, as it did on 19 May 2025.

Sources

  • Blatten
  • InSAR
  • L-band

Related

Join the 2027 pilot programme.

Watchberg is preparing pilots with hazard offices and operators in three Swiss cantons. If you carry responsibility for alpine slopes, let us show you what a weekly briefing would look like for your region.

Request a pilot briefingRead the articles

Pilot programme 2027. In discussion with hazard offices in three cantons.