Missouri & Watersheds · Water in a Changing Landscape

Dependency design · not executable yetSurface water, hydrography, cropland, roads, and general change/overlay support are not public CubeDynamics nouns and verbs.

The decision

Where has mapped surface water changed along a bounded Missouri River, reservoir, or tributary AOI, and which human or agricultural systems are near those changes?

A specific South Dakota water body will be selected only after source QA shows a genuine, interpretable temporal signal. The AOI will not be cherry-picked from a result to imply causation.

The missing information

An early water map and a recent water map are hard to compare mentally. A change layer is more useful when the contributing dates remain visible and nearby systems can be inspected without treating proximity as a cause.

The nouns

Noun	Public status	Publication requirement
Surface water	Missing	Stable water classification, dates, cloud/ice and sensor QA
Hydrography	Missing	Flowline/waterbody identity and topology provenance
Cropland	Missing	Year-specific classes and accuracy limitations
Roads or buildings	Missing	Completeness, date, geometry, and scale limits
Elevation	Missing	Datum, resolution, and resampling rules

The analytical sentence

This is a design target, not executable code:

EARLY WATER → COMPARE WITH RECENT WATER → KEEP CHANGE → OVERLAP NEARBY SYSTEMS

The general change verb must define categorical transitions rather than subtracting class codes. The overlay must preserve distance or intersection semantics explicitly.

What happened?

No water-change result has been produced. Publishing a polished map before validating clouds, ice, classification stability, hydrologic identity, and temporal comparability would put documentation ahead of the package.

QA publication gate

The future page must visibly show EARLY WATER → RECENT WATER → CHANGE before any decision overlay, plus observation counts, missingness, classification categories, spatial alignment, and a numerical change-area check.

Decision view

Unavailable. The eventual final map may show where measured water-class transitions occur near cropland or infrastructure. It must not infer why water changed.

What this does and does not tell us

Spatial and temporal co-occurrence could guide field review or data requests. It would not establish hydrologic causation, water availability, damage, or regulatory consequence.

Fork this question

• Compare two defensible seasons rather than arbitrary annual snapshots.
• Replace cropland with roads after both sources pass QA.
• Ask how results change with the proximity distance made explicit.