From forecast to the wind gauge
Weather forecasts give the wind 10 m above open ground. Athletes care about the reading on the sprint-straight gauge. That depends on two things about each track:
- Which way the sprints run. Only the part of the wind blowing along the straight counts. Wind from behind the athletes is a tailwind (green); wind in their faces is a headwind (red).
- How sheltered the track is. Stands, trees and buildings block much of the open-air wind. At calibrated tracks only 29–42% of the open-air wind along the straight reaches the gauge.
The expected reading is the forecast wind along the straight × the track's shelter factor. The "chance of a tailwind" allows for gusts: heats run within the same hour typically differ by about ±1 m/s. Forecasts further ahead are given extra uncertainty.
Which forecast: the site uses the ECMWF model (via Open-Meteo). Tested on 2026 meets it called tailwind or headwind right for 78% of heats, against 74% for Open-Meteo's default blend and 73–77% for the other models, and its "chance of a tailwind" figures match what happened (when it says 80%, about 80% of heats were tailwinds). Adding the spread between models, or the gust forecast, made no further difference, so neither is used.
The forecast is fetched fresh from Open-Meteo every time you open a track, and again every hour while the page stays open, so it always uses the newest weather model run. It goes 14 days ahead. Days 8–14 are marked outlook: hourly wind that far out is rough and often changes, so their chances get much wider uncertainty.
Calibrated tracks
These tracks were tested against past meet results. There are two levels:
- Calibrated (heats with start times, NSW): each meet was predicted hour by hour using only the other meets, then compared with the real gauge readings.
- Rough calibration (results by day, no heat times: Athletics Victoria and World Athletics): each meet day's average reading was compared with that day's wind along the straight. This reliably confirms which way the sprints run, and gives an approximate shelter factor. We tested the method on the hourly-calibrated tracks first: it got all their directions right.
| Track | Meets | Heats | Shelter | Typical error |
|---|
Calibration cut the error by up to half compared with using the raw forecast. In November 2026 the hourly-calibrated tracks were refitted for the ECMWF forecast on 801 heats from 2024–26 (each meet predicted using only the others): tailwind/headwind calls improved at every track, for example Blacktown 74% → 84% and Olympic Park 65% → 68%.
Upcoming meets
The meet list comes from the World Athletics calendar (every state) and Athletics Live (NSW, which publishes event start times). It's refreshed whenever the site is updated, so very new meets may be missing. Where start times are published, the site shows the expected wind for each event; heats usually run a little after the listed time.
Tailwind tracks and typical days
The "Tailwind tracks" lists and each track's "Typical season day" come from three seasons of past hourly weather (September–April, 2022–23 to 2024–25, ERA5 reanalysis), turned into an expected reading on the home straight with the same direction and shelter the forecasts use. "Useful tailwind" means +0.5 m/s or more. Tracks where sprints are sometimes run either way are left out of the rankings.
These are tendencies, not promises. At the 7 tracks with enough timed heats, the tracks the climate data ranks highest and lowest for tailwinds (Wollongong and the Hunter Sports Centre) were also highest and lowest in real gauge readings, but meets fall on too few days to rank every track reliably.
How accurate is it?
The site is much better at telling you whether you'll get a tailwind than how much. Gusts make heats in the same hour differ by about 1 m/s, so the exact reading is usually off by about 1 m/s and is within ±0.5 m/s only about a third of the time. Whether it's a tailwind or a headwind is far more reliable, especially when the expected wind is clearly away from zero.
| Expected on the straight | Tailwind or headwind right | Typical error in the reading |
|---|---|---|
| Over ±1.0 m/s | 9 in 10 heats | about 1.0 m/s |
| ±0.5 to ±1.0 m/s | 3 in 4 | about 1.0 m/s |
| ±0.25 to ±0.5 m/s | about 6 in 10 | about 0.8 m/s |
| Under ±0.25 m/s | a toss-up | about 1.0 m/s |
From 1,041 heats at Sydney Olympic Park, Campbelltown and The Crest, each meet predicted using only the other meets. The "chance of a tailwind" on each hour already allows for this.
Other tracks
Tracks without results data use the average shelter of the calibrated tracks, with wider uncertainty. We tested this by predicting each calibrated track using only the other tracks' shelter. It came out almost as accurate as the track's own factor.
The sprint direction comes from OpenStreetMap, usually from the separately mapped 100 m straight or the main grandstand. Where the map gives no clue, the direction was read from satellite images: the home straight is the one with the grandstand, pavilion or photo-finish building beside it. That method got all 19 tracks with result-checked directions right in a blind test. A few tracks still have no clear clue, so the site says so and doesn't call tailwind or headwind. Map directions can be wrong. Past results showed sprints run the opposite way at several tracks (including The Crest), and those were corrected. Grass ovals and fields wrongly mapped as running tracks are left out.
Things to know
- It's a guide, not an official reading. The gauge measures a few seconds of wind; gusts make heats in the same hour differ by about ±1 m/s.
- Some tracks run sprints either way. Officials may move the 100 m start to the other end to get a tailwind. At tracks where past results show this (Adelaide, Barden Ridge, Campbelltown) the site shows the standard direction, with buttons to switch to Reverse or With the wind (the start at whichever end gives a tailwind, hour by hour). That last option matters: in 2026, 3 of the 5 High Velocity meets at Barden Ridge and Campbelltown were run with the wind, and "with the wind" was right for 77% of heats at those tracks. Elsewhere the site shows the standard direction only.
- Your privacy. "Tracks near me" uses your location only in your browser to sort the list. It's never sent anywhere. Saved tracks are kept on your device. The only outside request is the weather forecast for the track you pick.
Sources
Forecasts: Open-Meteo. Tracks: © OpenStreetMap contributors; satellite imagery for checking directions: Esri World Imagery. Calibration: published results from Athletics NSW (HY-TEK archive and Athletics Live), the Athletics Victoria results hub and World Athletics competition results.