Skip to content

Ensemble values

Any numeric profile position can be a deterministic number or an ensemble object:

ensemble scalar
{
"members": 21,
"p10": 1.2,
"p25": 1.8,
"p50": 2.4,
"p75": 3.1,
"p90": 3.8
}

run.members declares total run membership. A position’s own members counts contributors at that position and can be lower when null or censored member values were excluded. Do not rank an absent value at zero.

A full-dropout position has exactly this shape:

full ensemble dropout
{
"members": 0,
"p10": null,
"p25": null,
"p50": null,
"p75": null,
"p90": null
}

Full dropout means no member contributed at that position. It differs from an absent optional field and from a scalar null forecast position.

  • p50 is the median, not the preferred or most likely member.
  • p25–p75 and p10–p90 describe member spread at one position; they are not calibrated event-probability intervals.
  • Percentiles remain ordered even when several collapse to the same value.
  • ceiledMembers, when present on a supported height position, counts members capped at the published column ceiling rather than resolving a physical top.

Wind direction is aggregated circularly. A cluster around 350° and 10° belongs near north, not near the arithmetic mean of 180°. Consumers use the published percentile object and package scene rather than recomputing it linearly.

Count contributors before reading a percentile

A percentile block says how many members contributed and how many reached the published column ceiling.

9 ensemble members contribute to the selected boundary-layer value; 9 are column-ceiled and 0 are absent. The published p10, p50, and p90 are 4,200, 4,200, and 4,200 metres.

01

Declared members

9

Run membership from the profile and the package analysis envelope.

02

Position accounting

Resolved
0
Ceiled
9
Absent
0

Ceiled values are lower bounds. Absent values never enter the ranking as zero.

03

Conditional distribution

p10
4,200 m
p50
4,200 m
p90
4,200 m

These magnitudes describe the contributing members, not a calibrated probability.

Counts and percentile values come directly from the validated ensemble scenario. windgram/analyze supplies the declared membership context; absent member values stay absent.Units member counts · altitude m ASL

Spread and censoring are different signals

Switch among controlled ensembles that are tightly grouped, widely grouped, or capped by the published column.

An interactive ensemble windgram comparing narrow and wide percentile bands with a separately counted column-censored case.

Declared members 9Largest usable-top IQR 149 mBL-top members at column ceiling 0
900m 2953ft 1599m 5245ft 2298m 7538ft 2996m 9831ft 3695m 12123ft 4394m 14416ft 12 13 14 15 16 17

Conclusion. Percentile-band width describes the retained positions, while column censoring is separate member-accounting metadata; a narrow line at the column ceiling is not evidence of a resolved top.

The reference scene draws the p25–p75 bands and p50 lines stored in each validated ensemble profile.Units height m MSL · percentile positions p25, p50, p75 · members count

Use isEnsembleValue when the band matters, isEnsembleDropout for the zero-member shape, and p50 from windgram/derive when a numeric derivation explicitly chooses the median. p50 returns number | null. Use isDeterministicProfile only when an integration genuinely rejects ensemble documents.

The scene graph builds median lines and percentile band geometry wherever the document supplies populated values. Dropout positions render as gaps; a level or hour whose required positions all dropped out contributes no scene element.

The column-ceiling study records how a 700 hPa source ceiling concealed spread that appeared when the same member derivation extended to 500 hPa.