No detectable precursor in Sentinel-1 backscatter before the 26 August 2026 Langtang Lirung rock–ice avalanche, Nepal
Abstract
On 26 August 2026 a rock–ice avalanche on the west flank of Langtang Lirung, Nepal, initiated a debris flow and flood that ran roughly 100 km down the Lhende Khola, Bhote Koshi and Trishuli. Several contemporaneous reports asserted that satellite radar had tracked a month of accelerating slope movement before failure. This note tests that claim in the amplitude record. Using Copernicus Sentinel-1 IW GRD amplitude on a single relative orbit, this note (i) locates the failure independently and (ii) tests directly for a precursory signal. Co- and cross-polarised change between 16 and 28 August is concentrated on the reported detachment coordinate: within 500 m, 30.0% of VV pixels and 25.4% of VH pixels changed by more than 6 dB, against 0.7% and 0.4% respectively across the remainder of a 308 km² scene. Across five pre-event passes from 29 June to 16 August, measured against 80 terrain-matched control patches, VV shows a nominally significant negative trend (−0.20 ± 0.03 dB cycle⁻¹ across independent control draws, p ≈ 0.02) that is produced entirely by the first scene of the series; excluding it, the remaining four dates are flat (+0.01 dB cycle⁻¹, p = 0.99). VH shows a monotonic rise (+0.62 ± 0.03 dB cycle⁻¹, p below 0.01, robust to exclusion of either endpoint) which is rejected by a same-orbit seasonal baseline: in 2025 the equivalent rise occurred earlier and had completed by late June, whereas in 2026 the face stood approximately 3 dB below the 2025 plateau at the same date and subsequently rose by about the same amount. The apparent precursor is the completion of a seasonal transition: a ten-year climatology on the same relative orbit places 2026 within 0.3 dB of the 2017–2025 median across the analysis window, and the July–August rise is what the median year does over that interval. No precursory signal is present in the freely available amplitude record at 12-day sampling; the detection limit is approximately 1 dB of progressive change over two months. Interferometric measurements reported subsequently indicate downslope motion of order 10 mm per month, an order of magnitude below what amplitude can resolve; the two results are complementary, and the amplitude limit quantified here is what separates them.
Related
- Calibrated detection limits in routine Sentinel-1 products 50.6 hours before the Langtang Lirung rock–ice avalanche, Nepal
- A steadily flowing glacier and no detectable precursor before the 11 October 2026 Budhi Gandaki rock–ice avalanche and flood, Nepal
- Very-high-resolution optical imagery reveals key clues to the Langtang Lirung ice–rock avalanche
- The 26 August 2026 Langtang Lirung ice–rock avalanche and its transboundary cascade: a multi-source observational reconstruction in the Bhote Koshi basin
- Glacier changes in Langtang Catchment, Central Nepalese Himalaya from the Little Ice Age (∼1815 CE) to 2023 CE
BibTeX
@misc{sah2026detectable,
title = {{No detectable precursor in Sentinel-1 backscatter before the 26 August 2026 Langtang Lirung rock–ice avalanche, Nepal}},
author = {Ajay Kumar Sah},
howpublished = {EarthArXiv},
year = {2026},
month = sep,
doi = {10.31223/x59n5n},
url = {https://eartharxiv.org/repository/view/14847/},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - No detectable precursor in Sentinel-1 backscatter before the 26 August 2026 Langtang Lirung rock–ice avalanche, Nepal
AU - Ajay Kumar Sah
PY - 2026
DA - 2026/09/06
PB - EarthArXiv
DO - 10.31223/x59n5n
UR - https://eartharxiv.org/repository/view/14847/
AB - On 26 August 2026 a rock–ice avalanche on the west flank of Langtang Lirung, Nepal, initiated a debris flow and flood that ran roughly 100 km down the Lhende Khola, Bhote Koshi and Trishuli. Several contemporaneous reports asserted that satellite radar had tracked a month of accelerating slope movement before failure. This note tests that claim in the amplitude record. Using Copernicus Sentinel-1 IW GRD amplitude on a single relative orbit, this note (i) locates the failure independently and (ii) tests directly for a precursory signal. Co- and cross-polarised change between 16 and 28 August is concentrated on the reported detachment coordinate: within 500 m, 30.0% of VV pixels and 25.4% of VH pixels changed by more than 6 dB, against 0.7% and 0.4% respectively across the remainder of a 308 km² scene. Across five pre-event passes from 29 June to 16 August, measured against 80 terrain-matched control patches, VV shows a nominally significant negative trend (−0.20 ± 0.03 dB cycle⁻¹ across independent control draws, p ≈ 0.02) that is produced entirely by the first scene of the series; excluding it, the remaining four dates are flat (+0.01 dB cycle⁻¹, p = 0.99). VH shows a monotonic rise (+0.62 ± 0.03 dB cycle⁻¹, p below 0.01, robust to exclusion of either endpoint) which is rejected by a same-orbit seasonal baseline: in 2025 the equivalent rise occurred earlier and had completed by late June, whereas in 2026 the face stood approximately 3 dB below the 2025 plateau at the same date and subsequently rose by about the same amount. The apparent precursor is the completion of a seasonal transition: a ten-year climatology on the same relative orbit places 2026 within 0.3 dB of the 2017–2025 median across the analysis window, and the July–August rise is what the median year does over that interval. No precursory signal is present in the freely available amplitude record at 12-day sampling; the detection limit is approximately 1 dB of progressive change over two months. Interferometric measurements reported subsequently indicate downslope motion of order 10 mm per month, an order of magnitude below what amplitude can resolve; the two results are complementary, and the amplitude limit quantified here is what separates them.
ER -