Recommended Free Tools
iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Only a few subtle signs were identified, and only in hindsight. A review of two satellite images noted possible movement near the slope’s crown and apparent rockfall on its midslope. A separate Sentinel-1 analysis found no robust progressive radar signal. The available evidence does not show that the collapse was predictably or operationally warnable.
What changed in satellite images before the landslide?
In a 1 October 2026 review, Dave Petley compared Planet Labs optical images acquired on 12 and 25 August—the latter one day before the failure. He reported little obvious overall change. Clouds, shadows, and other image effects complicate comparisons, and the 25 August image shows no clear dust plume or fresh deposits.
Possible movement near the crown
Petley pointed to a possible shift in ice near the eventual crown. Shadows make that interpretation uncertain, so it should be treated as a candidate change in appearance, not a confirmed measurement of displacement.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Apparent rockfall on the midslope
The 25 August scene also appears to show more rockfall or detached material on the midslope, with boulders accumulating below. Petley suggested this could have reduced support for the upper slope—a possible part of the failure process—but emphasized that the changes were minor and difficult to separate from normal background activity. The review describes retrospective image interpretation, not a validated alarm signal. Read Petley’s image review in Eos.
Did Sentinel-1 radar detect a precursor?
No robust progressive trend was reported in the pre-event Sentinel-1 amplitude observations. An EarthArXiv preprint summary describes five Copernicus Sentinel-1 IW GRD passes from 29 June through 16 August 2026, compared with 80 terrain-matched control patches.
The summary reports a nominal decline in VV amplitude of −0.20 ± 0.03 dB per cycle (p ≈ 0.02), but says that trend was driven entirely by the earliest scene. After excluding it, the remaining four dates were essentially flat: +0.01 dB per cycle (p = 0.99). The preprint gives an approximate sensitivity limit of 1 dB of progressive change over two months. In other words, this analysis did not detect a progressive change above its stated sensitivity; it cannot rule out smaller or otherwise undetected physical movement.
Rank #2
These results come from the preprint’s indexed summary, not an independently inspected full manuscript, and should not be described as peer-reviewed findings. The EarthArXiv record is the source for the reported analysis.
Does this mean the landslide was predictable?
No. Finding possible changes after a disaster is different from demonstrating that they could have been recognized in real time and used to issue a reliable warning. The cited accounts establish neither an operational threshold nor a usable warning lead time. The optical interpretation is constrained by moderate image resolution, shadows, and ordinary rockfall; the radar record has only five pre-event observations and a stated detection limit.
Petley’s summary captures the tension: “Examination of Planet Labs satellite imagery before the collapse of the slope suggests that there were surprisingly few signs of incipient failure.” That does not mean there was no movement. It means the reported candidate signs were subtle, uncertain, and not shown to be dependable indicators of imminent failure.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What happened on 26 August?
The Nepal Society for Earthquake Technology (NSET) describes the event as beginning with detachment from a north-facing slope near the Lhende River in the upper Bhote Koshi watershed, near the China–Nepal border and roughly 20 km upstream of Rasuwagadhi. Its preliminary account says an ice–rock avalanche entrained loose proglacial sediment and transformed into a mobile, water-laden debris flow and flood. NSET also says the seismic signal recorded near Langtang/Lirung was generated by the mass movement itself; it was not the trigger. See NSET’s event account.
Quick Recap
What the available observations cannot establish
- They do not quantify actual pre-failure slope displacement or establish how much movement occurred.
- They do not demonstrate that a particular observed change crossed a threshold for warning.
- They do not establish a validated failure mechanism or a reliable warning lead time.
- A lack of detectable change in these images and radar observations is not proof that no physical movement took place.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
