Skip to content

Source records stay separate from interpretation

RiverStat.us stores observations, forecasts, flood categories, and alerts as public source records. The UI explains availability and freshness without creating a safety rating.

1

Collect

Pull public source records from agency APIs and feeds.

2

Normalize

Store source identifiers, timestamps, units, and provenance in D1.

3

Refresh

Track source cadence, freshness windows, and unavailable states explicitly.

4

Display

Render current readings, forecasts, and unavailable states without inventing ratings.

Freshness

Freshness describes the age and parseability of source data. It is not a condition rating or a field instruction.

Observations

Observation panels show source-native parameter codes, values, units, observed timestamps, fetched timestamps, and qualifiers when available.

Forecasts and flood categories

Forecast and flood panels render only when persisted source data includes NWPS context for the same gauge or forecast point.

Unavailable states

Known routes can render unavailable states when a source does not provide a displayable value, mapping, forecast, or threshold.

Forecast display model

Observed and forecast values remain separate source records.

Observation chart unavailable

Run source ingestion to populate chartable station observations.

Status mapping

Fresh
Recent, parseable source record.
Delayed
Readable but older than expected cadence.
Unavailable
No displayable source-backed value.

Coverage limits

Some gauges do not publish forecasts, flood thresholds, or linked alert context. Those gaps remain visible instead of being filled with inferred values.

Official guidance

Emergency decisions should use official NWS, USGS, local government, and emergency management guidance first.