A compass rose of directional frequencies, where each sampled direction owns its wedge, stacks its magnitude classes from the hub outward, and is read against labelled percent-of-total rings with calm counted in the hub.
Preview in your theme
Loading preview…
This component is part of Pro
The install command, prompt and concept map are included with a Pro license — one purchase unlocks every Pro component, forever.
status="ready" · 16 compass points, 5 speed classes, 8,760 hourly observations. WSW prevails at 17.1% of the year against SE at 1.0%, a 16.8:1 spread; the 465 calm hours (5.3%) sit in the hub because they have no bearing to be drawn at, but they still count in the denominator.
Surface wind · Shannon, 2024
Hourly METAR observations, direction the wind blew from
16 directions · 5 classes · rim 17.1%
8,760 hourly observations
Wind rose of Surface wind · Shannon, 2024: 16 sampled directions around the compass, 8,760 hourly observations in total, each ring a share of that total, north at twelve o'clock. 5 classes of wind speed are stacked from the hub outward, so distance from the centre is a class as well as a frequency. The prevailing direction is WSW at 17.1% of all hourly observations, mostly 7–11 kt. 5.3% of hourly observations were calm (under 1 kt) and have no direction, so they are counted in the total and printed in the hub rather than drawn on the compass. The mean resultant points from 250.2°, nearest WSW, with a concentration of 36%, where 0% means the directions cancel out and 100% means every reading came from one bearing. Every direction is listed in the table that follows.
11–6 kt22.9%
27–11 kt31.1%
312–17 kt24.4%
418–24 kt11.9%
525+ kt4.5%
Each ring is a share of all 8,760 hourly observations; the radius is the frequency itself, so a band's thickness is its share and the rings are evenly spaced. Classes stack from the hub outward, 1–6 kt nearest the centre and 25+ kt hatched at the rim. North is at twelve o'clock.
5.3% of hourly observations were calm (under 1 kt): they happened, so they are in the denominator, but they have no bearing to be drawn at and sit in the hub instead.
Mean resultant from 250.2° (nearest WSW), concentration 36% — 0% would mean the directions cancel out entirely.
Every sampled direction of Surface wind · Shannon, 2024, with its bearing, its share of all hourly observations, its rank, and the share held by each wind speed class.
Direction
Bearing
Hourly observation count
Share of all hourly observations
Rank
1–6 kt
7–11 kt
12–17 kt
18–24 kt
25+ kt
N
0°
155
1.8%
13 of 16
0.5%
0.6%
0.4%
0.2%
under 0.1%
NNE
22.5°
305
3.5%
9 of 16
1%
1.2%
0.8%
0.3%
0.1%
NE
45°
543
6.2%
7 of 16
1.9%
2.1%
1.5%
0.6%
0.2%
ENE
67.5°
603
6.9%
5 of 16
2.1%
2.3%
1.7%
0.6%
0.2%
E
90°
383
4.4%
8 of 16
1.3%
1.5%
1.1%
0.4%
0.1%
ESE
112.5°
172
2%
12 of 16
0.6%
0.7%
0.5%
0.2%
under 0.1%
SE
135°
89
1%
16 of 16
0.3%
0.3%
0.3%
under 0.1%
under 0.1%
SSE
157.5°
113
1.3%
15 of 16
0.4%
0.4%
0.3%
0.1%
under 0.1%
S
180°
274
3.1%
11 of 16
0.8%
1.1%
0.8%
0.4%
0.1%
SSW
202.5°
693
7.9%
4 of 16
1.8%
2.6%
2.1%
1%
0.4%
SW
225°
1,295
14.8%
2 of 16
3.1%
4.7%
3.9%
2.1%
0.8%
WSW
247.5°
1,498
17.1%
1 of 16
3.4%
5.5%
4.6%
2.6%
1%
W
270°
1,136
13%
3 of 16
2.7%
4.2%
3.5%
1.9%
0.7%
WNW
292.5°
598
6.8%
6 of 16
1.6%
2.2%
1.8%
0.9%
0.3%
NW
315°
283
3.2%
10 of 16
0.8%
1.1%
0.8%
0.4%
0.1%
NNW
337.5°
155
1.8%
13 of 16
0.5%
0.6%
0.4%
0.2%
under 0.1%
Calm (under 1 kt)
no direction
465
5.3%
not ranked
—
—
—
—
—
onSelect: click a direction, or Tab into the rose and walk it with ← → , then ↑ ↓ to step through the speed classes and Enter to pick one.
status="empty" · the payload is valid but nothing was sampled — a different sentence from “still loading” and from “the query failed”. A ready payload that arrives with no directions, or with no classes to stack them out of, lands on this branch too.
Surface wind · Shannon, 2024
Nothing on the compass yet
A rose needs one row per sampled direction — including the directions that recorded nothing — and at least one wind speed class to stack them out of.
status="loading" · a twelve-petal skeleton, aria-hidden, driven by no timer and not pulsing under reduced motion.
Surface wind · Shannon, 2024
Loading Surface wind · Shannon, 2024
status="error" · the query failed; “Try again” appears only when onRetry is passed. Pressing it switches to ready, and focus moves to the card rather than falling back to the document body with the button that just unmounted.
Surface wind · Shannon, 2024
Couldn't load this rose
The data source didn't respond.
radiusScale="sqrt" · the same year with the radius area-corrected, so a petal’s ink — not how far it reaches — is proportional to its frequency. The cost is on show: the percent rings crowd toward the rim, so a band’s thickness stops being its share.
Surface wind · Shannon, 2024
16 directions · 5 classes · rim 17.1%
8,760 hourly observations
Wind rose of Surface wind · Shannon, 2024: 16 sampled directions around the compass, 8,760 hourly observations in total, each ring a share of that total, north at twelve o'clock. 5 classes of wind speed are stacked from the hub outward, so distance from the centre is a class as well as a frequency. The prevailing direction is WSW at 17.1% of all hourly observations, mostly 7–11 kt. 5.3% of hourly observations were calm (under 1 kt) and have no direction, so they are counted in the total and printed in the hub rather than drawn on the compass. The mean resultant points from 250.2°, nearest WSW, with a concentration of 36%, where 0% means the directions cancel out and 100% means every reading came from one bearing. Every direction is listed in the table that follows.
11–6 kt22.9%
27–11 kt31.1%
312–17 kt24.4%
418–24 kt11.9%
525+ kt4.5%
Each ring is a share of all 8,760 hourly observations; the radius is area-corrected, so a petal's ink is proportional to its frequency and the rings crowd toward the rim. Classes stack from the hub outward, 1–6 kt nearest the centre and 25+ kt hatched at the rim. North is at twelve o'clock.
5.3% of hourly observations were calm (under 1 kt): they happened, so they are in the denominator, but they have no bearing to be drawn at and sit in the hub instead.
Mean resultant from 250.2° (nearest WSW), concentration 36% — 0% would mean the directions cancel out entirely.
Every sampled direction of Surface wind · Shannon, 2024, with its bearing, its share of all hourly observations, its rank, and the share held by each wind speed class.
Direction
Bearing
Hourly observation count
Share of all hourly observations
Rank
1–6 kt
7–11 kt
12–17 kt
18–24 kt
25+ kt
N
0°
155
1.8%
13 of 16
0.5%
0.6%
0.4%
0.2%
under 0.1%
NNE
22.5°
305
3.5%
9 of 16
1%
1.2%
0.8%
0.3%
0.1%
NE
45°
543
6.2%
7 of 16
1.9%
2.1%
1.5%
0.6%
0.2%
ENE
67.5°
603
6.9%
5 of 16
2.1%
2.3%
1.7%
0.6%
0.2%
E
90°
383
4.4%
8 of 16
1.3%
1.5%
1.1%
0.4%
0.1%
ESE
112.5°
172
2%
12 of 16
0.6%
0.7%
0.5%
0.2%
under 0.1%
SE
135°
89
1%
16 of 16
0.3%
0.3%
0.3%
under 0.1%
under 0.1%
SSE
157.5°
113
1.3%
15 of 16
0.4%
0.4%
0.3%
0.1%
under 0.1%
S
180°
274
3.1%
11 of 16
0.8%
1.1%
0.8%
0.4%
0.1%
SSW
202.5°
693
7.9%
4 of 16
1.8%
2.6%
2.1%
1%
0.4%
SW
225°
1,295
14.8%
2 of 16
3.1%
4.7%
3.9%
2.1%
0.8%
WSW
247.5°
1,498
17.1%
1 of 16
3.4%
5.5%
4.6%
2.6%
1%
W
270°
1,136
13%
3 of 16
2.7%
4.2%
3.5%
1.9%
0.7%
WNW
292.5°
598
6.8%
6 of 16
1.6%
2.2%
1.8%
0.9%
0.3%
NW
315°
283
3.2%
10 of 16
0.8%
1.1%
0.8%
0.4%
0.1%
NNW
337.5°
155
1.8%
13 of 16
0.5%
0.6%
0.4%
0.2%
under 0.1%
Calm (under 1 kt)
no direction
465
5.3%
not ranked
—
—
—
—
—
A sheltered marina a few miles inland, same year and same classes: 30.3% of hours are dead calm and the busiest direction is SW at 13.6%. Fitted to its own data the rim is that 13.6%, so on this scale the marina looks as windy as the airport above.
Surface wind · Dooneen marina, 2024
8 directions · 5 classes · rim 13.6%
7,224 hourly observations
Wind rose of Surface wind · Dooneen marina, 2024: 8 sampled directions around the compass, 7,224 hourly observations in total, each ring a share of that total, north at twelve o'clock. 5 classes of wind speed are stacked from the hub outward, so distance from the centre is a class as well as a frequency. The prevailing direction is SW at 13.6% of all hourly observations, mostly 1–6 kt. 30.3% of hourly observations were calm (under 1 kt) and have no direction, so they are counted in the total and printed in the hub rather than drawn on the compass. The mean resultant points from 233.3°, nearest SW, with a concentration of 11.7%, where 0% means the directions cancel out and 100% means every reading came from one bearing. Every direction is listed in the table that follows.
11–6 kt31.1%
27–11 kt24%
312–17 kt11.2%
418–24 kt2.8%
525+ kt0.6%
Each ring is a share of all 7,224 hourly observations; the radius is the frequency itself, so a band's thickness is its share and the rings are evenly spaced. Classes stack from the hub outward, 1–6 kt nearest the centre and 25+ kt hatched at the rim. North is at twelve o'clock.
30.3% of hourly observations were calm (under 1 kt): they happened, so they are in the denominator, but they have no bearing to be drawn at and sit in the hub instead.
Mean resultant from 233.3° (nearest SW), concentration 11.7% — 0% would mean the directions cancel out entirely.
Every sampled direction of Surface wind · Dooneen marina, 2024, with its bearing, its share of all hourly observations, its rank, and the share held by each wind speed class.
Direction
Bearing
Hourly observation count
Share of all hourly observations
Rank
1–6 kt
7–11 kt
12–17 kt
18–24 kt
25+ kt
N
0°
385
5.3%
8 of 8
2.4%
1.8%
0.8%
0.2%
under 0.1%
NE
45°
481
6.7%
6 of 8
2.9%
2.3%
1.1%
0.3%
under 0.1%
E
90°
767
10.6%
3 of 8
4.8%
3.8%
1.7%
0.3%
under 0.1%
SE
135°
565
7.8%
5 of 8
3.4%
2.6%
1.4%
0.4%
under 0.1%
S
180°
398
5.5%
7 of 8
2.7%
1.6%
1%
0.1%
under 0.1%
SW
225°
980
13.6%
1 of 8
6%
4.7%
2.1%
0.6%
0.1%
W
270°
876
12.1%
2 of 8
5.4%
4.3%
1.9%
0.4%
0.2%
NW
315°
582
8.1%
4 of 8
3.5%
2.9%
1.2%
0.4%
under 0.1%
Calm (under 1 kt)
no direction
2,190
30.3%
not ranked
—
—
—
—
—
domainPercent={20} · the same marina against a rim both cards can share, since Shannon peaks at 17.1%. Nothing about the data changed; the rose is simply no longer allowed to fill its own frame, which is the whole difference between “busy” and “busy compared to what”.
Surface wind · Dooneen marina, 2024
8 directions · 5 classes · rim 20%
7,224 hourly observations
Wind rose of Surface wind · Dooneen marina, 2024: 8 sampled directions around the compass, 7,224 hourly observations in total, each ring a share of that total, north at twelve o'clock. 5 classes of wind speed are stacked from the hub outward, so distance from the centre is a class as well as a frequency. The prevailing direction is SW at 13.6% of all hourly observations, mostly 1–6 kt. 30.3% of hourly observations were calm (under 1 kt) and have no direction, so they are counted in the total and printed in the hub rather than drawn on the compass. The mean resultant points from 233.3°, nearest SW, with a concentration of 11.7%, where 0% means the directions cancel out and 100% means every reading came from one bearing. Every direction is listed in the table that follows.
11–6 kt31.1%
27–11 kt24%
312–17 kt11.2%
418–24 kt2.8%
525+ kt0.6%
Each ring is a share of all 7,224 hourly observations; the radius is the frequency itself, so a band's thickness is its share and the rings are evenly spaced. Classes stack from the hub outward, 1–6 kt nearest the centre and 25+ kt hatched at the rim. North is at twelve o'clock.
30.3% of hourly observations were calm (under 1 kt): they happened, so they are in the denominator, but they have no bearing to be drawn at and sit in the hub instead.
Mean resultant from 233.3° (nearest SW), concentration 11.7% — 0% would mean the directions cancel out entirely.
Every sampled direction of Surface wind · Dooneen marina, 2024, with its bearing, its share of all hourly observations, its rank, and the share held by each wind speed class.
Direction
Bearing
Hourly observation count
Share of all hourly observations
Rank
1–6 kt
7–11 kt
12–17 kt
18–24 kt
25+ kt
N
0°
385
5.3%
8 of 8
2.4%
1.8%
0.8%
0.2%
under 0.1%
NE
45°
481
6.7%
6 of 8
2.9%
2.3%
1.1%
0.3%
under 0.1%
E
90°
767
10.6%
3 of 8
4.8%
3.8%
1.7%
0.3%
under 0.1%
SE
135°
565
7.8%
5 of 8
3.4%
2.6%
1.4%
0.4%
under 0.1%
S
180°
398
5.5%
7 of 8
2.7%
1.6%
1%
0.1%
under 0.1%
SW
225°
980
13.6%
1 of 8
6%
4.7%
2.1%
0.6%
0.1%
W
270°
876
12.1%
2 of 8
5.4%
4.3%
1.9%
0.4%
0.2%
NW
315°
582
8.1%
4 of 8
3.5%
2.9%
1.2%
0.4%
under 0.1%
Calm (under 1 kt)
no direction
2,190
30.3%
not ranked
—
—
—
—
—
Degenerate input · one direction owns the whole turn. SVG defines an elliptical arc whose start and end points coincide as drawing nothing, so this has to be split into two half turns or the card comes out blank.
Ridge mast · first week on site
1 direction · 3 classes · rim 93.9%
49 hourly observations
Wind rose of Ridge mast · first week on site: 1 sampled direction around the compass, 49 hourly observations in total, each ring a share of that total, north at twelve o'clock. 3 classes of wind speed are stacked from the hub outward, so distance from the centre is a class as well as a frequency. The prevailing direction is SW at 93.9% of all hourly observations, mostly 7–11 kt. 6.1% of hourly observations were calm (under 1 kt) and have no direction, so they are counted in the total and printed in the hub rather than drawn on the compass. The mean resultant points from 225°, nearest SW, with a concentration of 100%, where 0% means the directions cancel out and 100% means every reading came from one bearing. Every direction is listed in the table that follows.
11–6 kt28.6%
27–11 kt44.9%
312–17 kt20.4%
Each ring is a share of all 49 hourly observations; the radius is the frequency itself, so a band's thickness is its share and the rings are evenly spaced. Classes stack from the hub outward, 1–6 kt nearest the centre and 12–17 kt hatched at the rim. North is at twelve o'clock.
6.1% of hourly observations were calm (under 1 kt): they happened, so they are in the denominator, but they have no bearing to be drawn at and sit in the hub instead.
Mean resultant from 225° (nearest SW), concentration 100% — 0% would mean the directions cancel out entirely.
Every sampled direction of Ridge mast · first week on site, with its bearing, its share of all hourly observations, its rank, and the share held by each wind speed class.
Direction
Bearing
Hourly observation count
Share of all hourly observations
Rank
1–6 kt
7–11 kt
12–17 kt
SW
225°
46
93.9%
1 of 1
28.6%
44.9%
20.4%
Calm (under 1 kt)
no direction
3
6.1%
not ranked
—
—
—
Degenerate input · eight identical histograms. Every petal reaches the rim and the directions cancel out exactly, so the component reports the rose as level — no peak is named, and the resultant arrow is withheld rather than pointing wherever the arithmetic noise landed.
Wind-tunnel reference run · eight headings
8 directions · 5 classes · rim 11.9%
840 hourly observations
Wind rose of Wind-tunnel reference run · eight headings: 8 sampled directions around the compass, 840 hourly observations in total, each ring a share of that total, north at twelve o'clock. 5 classes of wind speed are stacked from the hub outward, so distance from the centre is a class as well as a frequency. Every direction recorded 100 hourly observations, so the rose is an even disc and nothing prevails. 4.8% of hourly observations were calm (under 1 kt) and have no direction, so they are counted in the total and printed in the hub rather than drawn on the compass. Every direction is listed in the table that follows.
11–6 kt28.6%
27–11 kt42.9%
312–17 kt19%
418–24 kt3.8%
525+ kt1%
Each ring is a share of all 840 hourly observations; the radius is the frequency itself, so a band's thickness is its share and the rings are evenly spaced. Classes stack from the hub outward, 1–6 kt nearest the centre and 25+ kt hatched at the rim. North is at twelve o'clock.
4.8% of hourly observations were calm (under 1 kt): they happened, so they are in the denominator, but they have no bearing to be drawn at and sit in the hub instead.
Every direction recorded the same number, so there is no prevailing direction to name and none is named.
Every sampled direction of Wind-tunnel reference run · eight headings, with its bearing, its share of all hourly observations, its rank, and the share held by each wind speed class.
Direction
Bearing
Hourly observation count
Share of all hourly observations
Rank
1–6 kt
7–11 kt
12–17 kt
18–24 kt
25+ kt
N
0°
100
11.9%
level
3.6%
5.4%
2.4%
0.5%
0.1%
NE
45°
100
11.9%
level
3.6%
5.4%
2.4%
0.5%
0.1%
E
90°
100
11.9%
level
3.6%
5.4%
2.4%
0.5%
0.1%
SE
135°
100
11.9%
level
3.6%
5.4%
2.4%
0.5%
0.1%
S
180°
100
11.9%
level
3.6%
5.4%
2.4%
0.5%
0.1%
SW
225°
100
11.9%
level
3.6%
5.4%
2.4%
0.5%
0.1%
W
270°
100
11.9%
level
3.6%
5.4%
2.4%
0.5%
0.1%
NW
315°
100
11.9%
level
3.6%
5.4%
2.4%
0.5%
0.1%
Calm (under 1 kt)
no direction
40
4.8%
not ranked
—
—
—
—
—
Degenerate input · bearings sampled unevenly (128° owns a 72.5° wedge behind the hole in the compass, 035° only 25°), one heading that recorded nothing, a NaN, a negative count, a row two classes short and a row one class long, plus a label nobody trimmed. Three entries are dropped and said out loud in the footnote — none of it is quietly clamped.
Ridge mast · raw export, unvalidated
8 directions · 4 classes · rim 32.8%
299 ten-minute records
Wind rose of Ridge mast · raw export, unvalidated: 8 sampled directions around the compass, 299 ten-minute records in total, each ring a share of that total, north at twelve o'clock. 4 classes of gust speed are stacked from the hub outward, so distance from the centre is a class as well as a frequency. The prevailing direction is 250° at 32.8% of all ten-minute records, mostly 3–6 m/s. 8.7% of ten-minute records were calm (below the 0.4 m/s start-up speed) and have no direction, so they are counted in the total and printed in the hub rather than drawn on the compass. The mean resultant points from 256.5°, nearest 250°, with a concentration of 53.4%, where 0% means the directions cancel out and 100% means every reading came from one bearing. Every direction is listed in the table that follows.
10–3 m/s32.8%
23–6 m/s36.5%
36–10 m/s14.7%
410+ m/s7.4%
Each ring is a share of all 299 ten-minute records; the radius is the frequency itself, so a band's thickness is its share and the rings are evenly spaced. Classes stack from the hub outward, 0–3 m/s nearest the centre and 10+ m/s hatched at the rim. North is at twelve o'clock.
8.7% of ten-minute records were calm (below the 0.4 m/s start-up speed): they happened, so they are in the denominator, but they have no bearing to be drawn at and sit in the hub instead.
Mean resultant from 256.5° (nearest 250°), concentration 53.4% — 0% would mean the directions cancel out entirely.
3 count entries were unusable (not a finite number, or below zero) and were left out of every total on this card. 1 direction reports fewer classes than the legend declares; the missing ones are read as zero, never shifted onto the next class.
Every sampled direction of Ridge mast · raw export, unvalidated, with its bearing, its share of all ten-minute records, its rank, and the share held by each gust speed class.
Direction
Bearing
Ten-minute record count
Share of all ten-minute records
Rank
0–3 m/s
3–6 m/s
6–10 m/s
10+ m/s
010°
10°
24
8%
5 of 8
4%
2.7%
1%
0.3%
035°
35°
31
10.4%
4 of 8
3%
4.7%
2%
0.7%
060° (anemometer swapped mid-season)
60°
7
2.3%
7 of 8
1.3%
1%
0%
0%
128°
128°
0
0%
8 of 8
0%
0%
0%
0%
205°
205°
33
11%
3 of 8
6%
0%
3.7%
1.3%
232°
232°
58
19.4%
2 of 8
7.4%
10.4%
0%
1.7%
250°
250°
98
32.8%
1 of 8
8.7%
14.7%
6.4%
3%
305°
305°
22
7.4%
6 of 8
2.3%
3%
1.7%
0.3%
Calm (below the 0.4 m/s start-up speed)
no direction
26
8.7%
not ranked
—
—
—
—
Dense input · 36 headings at 10° each, 1,496 ten-minute records, 17.4% of them calm. The rim cannot carry 36 labels, so the stride rule prints every second one and all 36 stay in the screen-reader table; the three headings that recorded nothing are ticks on the floor ring rather than gaps.
Glen mast · ten-minute records, 36 headings
36 directions · 4 classes · rim 8.9%
1,496 ten-minute records
Wind rose of Glen mast · ten-minute records, 36 headings: 36 sampled directions around the compass, 1,496 ten-minute records in total, each ring a share of that total, north at twelve o'clock. 4 classes of gust speed are stacked from the hub outward, so distance from the centre is a class as well as a frequency. The prevailing direction is 310° at 8.9% of all ten-minute records, mostly 3–6 m/s. 17.4% of ten-minute records were calm (below the 0.4 m/s start-up speed) and have no direction, so they are counted in the total and printed in the hub rather than drawn on the compass. The mean resultant points from 310.2°, nearest 310°, with a concentration of 24.1%, where 0% means the directions cancel out and 100% means every reading came from one bearing. Every direction is listed in the table that follows.
10–3 m/s27.9%
23–6 m/s29.8%
36–10 m/s18.2%
410+ m/s6.7%
Each ring is a share of all 1,496 ten-minute records; the radius is the frequency itself, so a band's thickness is its share and the rings are evenly spaced. Classes stack from the hub outward, 0–3 m/s nearest the centre and 10+ m/s hatched at the rim. North is at twelve o'clock. Compass labels are printed every second direction; all 36 are in the table.
17.4% of ten-minute records were calm (below the 0.4 m/s start-up speed): they happened, so they are in the denominator, but they have no bearing to be drawn at and sit in the hub instead.
Mean resultant from 310.2° (nearest 310°), concentration 24.1% — 0% would mean the directions cancel out entirely.
Every sampled direction of Glen mast · ten-minute records, 36 headings, with its bearing, its share of all ten-minute records, its rank, and the share held by each gust speed class.
Zyeon UI