Landslide Science Naming conventions

Naming conventions

Many Alaska placenames recur across distinct landslides (there are several "Moose Creek"s), and many failures sit near a named feature without being on it. To keep every landslide's unique_name both human-readable and unambiguous, names are built from a base placename followed by compact disambiguators, read left→right from most to least significant. Every unique_name is unique across the inventory.

1. A single landslide at a uniquely-named feature

Use the placename alone for a slow landslide; add a time qualifier for a catastrophic one (the year if known, otherwise the epoch):

2. A slow and a catastrophic landslide at the same place

They share the base name — the time qualifier on the catastrophic one is enough to tell them apart:

If they're in somewhat different spots on the same feature, add a distinguishing letter:

3. Several of the same type along one feature

Distinguish the slopes with letters:

4. One slope failing catastrophically more than once

Distinguish the events by year (or epoch). When more than one event falls in the same time-bin — the same year, or both Holocene/Modern — append a .N index: .1 for the first, .2 for the second, and so on.

5. A genuinely different feature that shares a name

When two unrelated landslides would otherwise carry the same placename (the "Moose Creek" case), append a number to the second feature, then letters for slopes on it. The number distinguishes the feature; the letters distinguish slopes:

6. Weakly-associated placenames (the ~ distinctor)

When a landslide is only near a named feature — not on it — insert a ~ right after the base name to flag that the association is loose. The tilde reads as “approximately / near,” which is exactly the meaning. (Names containing "trib" or "neighbor" are the old, ad-hoc form of this and should be migrated to the ~ form.)

Note: the distinctor used to be an X, but that collided with the slope-letter sequence once it reached the 24th slope (also X) — so it is now ~. Records still carrying a lone X are flagged for review.

ComponentMeans
Eureka Glacier the nearby named feature the name borrows from
~ this landslide is near, not on, Eureka Glacier
2 the 2nd distinct feature borrowing the Eureka Glacier name
C the 3rd failing slope on that feature
2016.2 the year of the catastrophic failure, and the .2 index marking the 2nd failure of that slope in 2016

… so Eureka Glacier ~ 2 C 2016.2 reads as the full, ordered description above.

The side of the ~ matters. Everything before the ~ is the base feature the landslide is loosely tied to; a number after the ~ counts which near-feature it is. So a number before vs. after the tilde means different things:

Order of components

Read left to right, components run from most to least significant: base placename (which may already carry a case 5 feature number, e.g. Toyota Creek 2) → optional ~ → optional near-feature number (which one near the base) → slope letter → year or epoch (catastrophic) → per-time-bin .N index. Omit any component that isn't needed.

Names are applied by editors. The viewer flags likely non-conforming names (e.g. those containing "trib"/"neighbor", or an un-disambiguated base shared with another record) for review, and refuses a rename or a drawn polygon that would collide with an existing name — nudging toward a disambiguator rather than a silent duplicate.

← Back to Methods.