Wind and heat: fifteen years of measurements on arrow drift and scores

The question

Anyone who shoots outdoors knows that wind and heat cost points. Research has tried to say how many, and it has done so along two routes that for years did not meet. The first follows the arrow: calculations, wind tunnels and shooting machines measure how far a crosswind pushes it and which arrows drift less. The second starts from the scorecard, and sets competition scores beside the day's weather data.

Here I follow the studies in chronological order, from the first model in 2011 to the simulations of 2025. How to read the wind on the shooting line is covered in the article on wind and weather; this page says where the figures come from and how well they hold.

2011-2013: drift becomes a calculation

The first work is by James Park, an engineer at Monash University in Australia. In 2011 he split the drift of a compound arrow into two parts: the arrow turning towards the apparent wind as it lines up with the flow, and the sideways component of air drag, which at 70 m accounted for 84% of the shift. In the model drift grows linearly with wind speed and roughly with the square of distance, while three metres per second of head or tail wind changes the impact height at 70 m by only 24 mm. The calculation was put to the test twice: in a wind tunnel five Easton arrows deflected in the predicted order, and outdoors, at 70 m with gusts above 10 m/s, two models shot from a machine spread in a ratio close to the expected one.

The same paper reports the first observation from competition, which the author calls subjective: at the 2008 Australian National Championships the best compound archers, with little wind at 90 m, were just below the world record, and later, with gusts above 3 m/s, they scored far less than their 90 m scores predicted. Two years later Park laid the curves of equal drift over the limits of carbon fibre and found the other side of the coin. A stiffer material would allow thinner shafts, but also lighter ones, and drift would barely change, because what is gained in cross-section is lost in mass.

2014: the first estimate in points

In 2014 Park took the question to the score. His model divides the spread of the arrows into a part that grows with distance and a part that grows with its square, and places wind drift among the errors of the second kind. He applied it to the 2013 World Championships in Antalya, where men's recurve archers had shot 90 and 70 m in the morning with little wind and 50 and 30 m in the afternoon with about 3-5 m/s. At 70 m the scores were within half a point of those expected from the 90 m; at 50 m they were lower, and the difference attributed to the wind is worth about 5 points over 36 arrows. It is the first figure in points, but it comes from a single event, with the wind estimated, and morning and afternoon also differ in temperature.

2018-2020: the recurve in the computer

The recurve arrived in 2018 with a Japanese simulation built on coefficients measured in a wind tunnel and in free flight. At 70 m, with three metres per second, an A/C/E arrow ended 0.36 m from the centre in a crosswind, 0.10 in a headwind and 0.09 in a tailwind. The model also gave the mechanism: in the crosswind the arrow's angle of attack rose to 3.0 degrees, in the other two cases it stayed below one degree, and that angle decides whether the air around the shaft stays orderly or becomes turbulent.

In 2019 two papers ended up on opposite sides of the question of diameter. In a review of the atmosphere and target archery Park summarised his tests in the wind tunnel and at long range: all else being equal the thinner shaft drifts less, but the values are not published. In Osaka, meanwhile, a group building an archery robot made a simple calculation treating the arrow as a cylinder. Two things came out of it: even a wind nobody feels, 0.3 m/s, moves the competition arrow by about 2 cm at 70 m; and that arrow, thinner and a third lighter than a practice arrow, takes two thirds of the side force yet at 70 m drifts 3.8 times as far as the practice arrow at 18 m, because it stays in the air twice as long. The comparison does not isolate diameter, since weight and distance change too, but it shows that side force and drift are not the same thing. The review also counts for what it does not find: no study of arrow flight in rain, humidity or different temperatures, and none on the habit of adding mass to the bow against gusts.

In 2020 Park redid the calculation for the recurve with a model that lets the shaft flex. With a three metre per second crosswind an X10 600 drifts between 245 and 261 mm at 70 m depending on the point, more than four scoring rings of the 122 cm face. Diameter turns out to be the heaviest factor, from 253 mm for the X10 to 311 for the ACE and 441 for the aluminium X7; at equal diameter the heavier arrow drifts less, 245 mm with the 140 grain point against 261 with the 100 grain one, even though it leaves the bow more slowly.

2021: heat enters the data

The second route opens with two economists. Qiu and Zhao set the day's weather beside the qualification scores of outdoor competitions held in China between 2010 and 2016: 35,954 performances by 3,196 archers, almost all recurve, with each archer compared with themselves. The heat that matters is the heat index, which combines temperature and humidity, and the curve is an inverted U: against the band between 18 and 22 degrees, with no wind, 0.25 points per arrow are already lost between 22 and 26 degrees and 1.07 above 34, 13.5% of a mean of 7.91. At the sample's mean wind, 2.35 m/s, the losses from moderate heat fall to 0.19-0.29 points, but above 34 degrees they stay at 0.82. Extreme heat also widens the gap, because the top tenth loses significantly less; between men and women there is no difference.

The same sample answers other questions. With temperature in the model, wind speed lowers the mean score and the share of tens, up to the observed maximum of 5.38 m/s, because with more wind competitions are postponed. Rain on the day and humidity on its own have no significant effect, and cool days in the main model do not differ from the reference band. Experience protects: moving out of the least experienced quarter of the sample takes 0.33 points off the loss above 34 degrees. Years of habit protect too, since archers who train in southern China, where very hot days are far more frequent, lose 0.296 points less than those who train in the north, while heat met in the months before the competition changes almost nothing. It is an association on a single circuit, with weather from stations and not from the shooting line. Before it there was only one Malaysian archer followed in nine conditions, with no statistical test between conditions and in a journal with doubtful editorial standards.

2025: confirmation in the computer and the archers' voice

In 2025 a second simulation, independent and built with computational fluid dynamics, repeated the comparison on an X10 launched at 57 m/s: at 70 m, with three metres per second, 25.8 cm of drift in a crosswind, 8.5 in a headwind and 7.8 in a tailwind. The crosswind estimate is more than a quarter lower than the 2018 one, and the difference shows how much the model's assumptions weigh. The same work finds less drift with more mass and with the centre of mass further forward and, in another set of calculations, a strong effect of the arrow's attitude at launch: with three degrees of initial angle of attack the maximum drift reaches 0.27 m, with the arrow aligned it stops at 0.125.

The same year brought the archers' voice. Sixty-five Korean archers, averaging 327 points out of 360 at 70 m, described in writing one of their shots that landed in the X ring. In aiming and release the most cited theme is wind management, 76 mentions out of 157, and 58 refer to adjusting the strength of the shot against 10 that refer to calculating the wind. In the draw and anchor, though, the external condition named most often is silence.

What holds today

On the physics the routes agree. Three independent models say that at 70 m a crosswind moves the arrow far more than a head or tail wind, and Park's calculations together with the 2025 simulation say that at equal diameter the heavier arrow drifts less. They are calculations, and in the registry they remain transferred until someone measures them outdoors. On diameter the 2019 readings remain opposed and the wind measurements change weight and diameter together, so the claim about the thinner shaft is contested. The attitude with which the arrow leaves the bow also divides the models: for Park one degree is worth about ten millimetres out of 245, in the 2025 simulation three degrees double the drift.

On the score there is only one large dataset, the Chinese one, with the Antalya event beside it. Wind lowers the score and extreme heat lowers it by about a tenth, but a single large study, with one event beside it, only reaches reported. The same goes for experience and years of heat reducing the loss and for moderate cold, which costs nothing in that sample but with a fragile result. Rain, humidity and temperature split, because the first two carry no measurable weight and the third does. That wind is the main problem outdoors and that it is handled by adjusting the strength of the shot comes only from the Korean archers' recollections.

What remains open

The measurement that would join the two routes is missing: wind recorded by an anemometer beside the shooting line, end by end, set beside the score of the same end. Today the wind in competition data comes from weather stations, and drift is calculated or measured with machines. There is no test that isolates weight at equal diameter and spine, and no study at all of the wet arrow or of different temperatures. Adding mass to the bow against gusts has been measured neither on a machine nor on archers. Heat, finally, has been studied on a single circuit: nobody has tested an acclimatisation protocol on archers, and cold near freezing has never entered a sample.

@misc{campagna2026_enresearchwindheat,
  author = {Campagna, Matteo},
  title = {Wind and heat in archery: research since 2011},
  year = {2026},
  howpublished = {Research, matteocampagna.com},
  url = {https://matteocampagna.com/en/research/wind-heat},
  note = {accessed 2026-10-16}
}