Field notes
When the landscape moved
A hillside can move for years before collapsing, or it can fail in minutes. These six case studies connect satellite views, field observations and measured impacts to show why location, timing and the path of moving material matter as much as size.
FREE ILLUSTRATED REVIEW
When the Landscape Moved
6 selected cases, credited imagery or charts, original sources and an observation checklist. No sign-in or app required.
PDF · 10 pages · 0.9 MB · Sources checked November 11, 2026

The hillside that did not collapse matters too
A useful case collection includes more than dramatic failures. A comparison site can help reveal what a single disaster photograph leaves out.
This report pairs five landslide or debris-flow disasters with Paul’s Slide, which did not fail catastrophically during the 2015–2017 study window. Mud Creek and Paul’s Slide experienced similar weather but had different outcomes. That is a reason to examine movement histories and slope conditions, not use rainfall alone to predict collapse.
Follow the sequence, not just the scar
Keep prior movement, water conditions, acceleration, runout and consequences as distinct observations. A satellite image after a failure can show its extent without telling you when acceleration began.
The six cases range from the Annapurna–Seti cascade in Nepal to Oso, Mud Creek and Montecito. Rock avalanches, moving slopes and post-fire debris flows are different mechanisms. Preserve the source’s description and measurement method instead of forcing every case into one score.
A mass estimate in metric tons cannot be compared directly with a volume in cubic meters. Likewise, a preliminary mapped debris extent is not a measurement of speed. Unknown values remain unknown in the report.
What the report covers
This is a selected case-study collection, not a ranking or a complete inventory. Five records describe failures; Paul’s Slide is a comparison that did not fail catastrophically during the study period. Published estimates retain their units and dates. Satellite acquisition dates differ from event dates. Historical evidence is not a current hazard assessment.
The illustrated PDF keeps the original source links and image credits alongside each case. This historical collection does not replace current agency alerts or local safety guidance.
- Landslide and Deadly Flood in Nepal
- Large Landslide Detected in Southeastern Alaska
- USGS: One Year Later – The Oso Landslide in Washington
- Aerial Views of the Landslide in Oso, Washington
- Drought then Deluge Turned a Stable Landslide into Disaster
- Handwerger et al. (2019), Scientific Reports: stable sliding and catastrophic failure
- Deadly Debris Flows in Montecito
- USGS data: Debris-flow inundation and damage, January 9, 2018 Montecito event
Make your own observations comparable
We assembled the source-backed case records in Dattle. The PDF is open to read and download without an account. You can use the same record structure in your own collection, spreadsheet or notebook.
Start with one observation, a clear definition and its source. Add the next record using the same units and method where possible. AI can accelerate the setup by suggesting fields; it should not invent observations, citations or missing measurements.
This article and report were prepared with AI assistance, with linked sources and limitations retained. Historical imagery is credited source material; charts are drawn from the cited data.
Explore the other field notes
Follow another question through its measurements, images and source limitations.
KEEP IT SIMPLE
Your starting checklist
- Identify the mechanism and location.
- Keep earlier movement observations.
- Record rainfall amount and duration.
- Preserve units and measurement method.
- Include a clearly labeled comparison.
GIVE YOUR IDEA A HEAD START
Describe it. Make it yours.
Copy a prompt below. In Dattle, choose New Dattle → Describe with AI, paste it and select Generate. Review the suggested title and fields, choose Private or Shared, then save your Dattle.
AI drafts the setup; you add the records. Generation uses AI credits and is subject to your account limits. You can also start from a template or build manually.
Review the fields before saving. Add verified source records or your own observations, preserve image credits, and choose sharing permissions deliberately. Features and limits vary by plan.