By Paul Clark ·
Race Day vs the Calculator: Adjusting Your Pace for Heat, Humidity, Hills, Altitude and Wind
A predicted race time assumes a cool, flat, still day. How heat, dew point, hills, altitude and wind change the pace you can hold, with approximate rules of thumb and a worked half marathon.
A race predictor answers a narrow question: if race day is like the day you ran your input race, what time should you expect? It doesn’t know that your 10K PR came on a cold, still morning on a flat loop, or that the half marathon you’ve entered starts at 9 a.m. in late spring with 180 metres of climbing. The prediction is honest about the conditions it was given. The problem is that race day usually has different ones.
This post is about closing that gap. The formula behind the prediction is explained in running pace and race-time prediction. How to split a marathon once you have a goal is covered in how to pace a marathon, which also explains why to discount the prediction if your mileage is low. Here we look at the weather and the course: what each one costs, roughly, and how to turn that into a number you can run to.
A warning before any numbers. Everything below is an approximation drawn from widely used coaching rules of thumb and from general findings in exercise physiology. Individuals vary a lot, especially in heat. Use these figures to set a cautious starting pace, then run the race by feel.
What the predictor quietly assumes
The Running Pace Converter scales one result to other distances with Riegel’s formula and a fixed exponent of 1.06. Built into that one line are several assumptions the tool can’t check:
| Assumption | What breaks it |
|---|---|
| You’re trained for the target distance | Low mileage, short long runs (see the marathon post) |
| Race day matches the input race’s conditions | Heat, humidity, altitude, wind |
| The course is about as flat as the input course | Hills, especially late in the race |
| You run even splits | A fast start, congestion, aid-station stops |
| You start fresh and tapered | A hard week, poor sleep, illness |
The calculator does none of the adjusting in this post. It has no fields for temperature, elevation or wind, and its split table is strictly even. What it does well is the arithmetic after you’ve decided on a pace. Its second input mode accepts a pace plus a “Held for” distance, and returns your finish time and cumulative splits in km or miles. So the workflow is: predict, adjust in your head or on paper, then feed the adjusted pace back in.
One subtle point. Conditions affect the input race too. If your 10K PR was run in a heatwave, the prediction already carries a heat penalty, and applying another on a hot race day would count it twice. Use an input race run in reasonable conditions whenever you can.
Heat and humidity: the biggest cost and the hardest to judge
Running produces a lot of heat, and on a warm day it’s harder to get rid of. More blood goes to the skin for cooling and less is available to the working muscles, heart rate rises at any given pace, and the pace you can sustain drops. Studies of big-city marathon results, such as the work of Ely and colleagues published in 2007, found finishing times slow steadily as temperatures climb above cool conditions of about 10°C, and that slower runners lose proportionally more than elites. That last finding matters: a four-hour marathoner is out in the heat longer and usually has less heat acclimatisation to rely on.
Humidity matters because sweat cools you by evaporating, and moist air slows evaporation. Dew point measures this better than relative humidity. One rule of thumb often shared among US coaches adds the air temperature and dew point, both in °F, and reads off a pace adjustment:
| Temp °F + dew point °F | Approximate pace adjustment |
|---|---|
| 100 or less | None |
| 101–110 | 0–0.5% |
| 111–120 | 0.5–1% |
| 121–130 | 1–2% |
| 131–140 | 2–3% |
| 141–150 | 3–4.5% |
| 151–160 | 4.5–6% |
| 161–170 | 6–8% |
| 171–180 | 8–10% |
| Above 180 | Hard running not recommended |
Treat the table as a floor for slowing down, not a precise figure. If you’re not used to the heat, or you’re racing a marathon rather than a 10K, aim toward the top of each band or above it.
Let heart rate set the ceiling
On a hot day pace is a poor guide, because the pace you planned asks more of your body than it would on a cool day. Heart rate shows that extra strain, and that makes it useful as a limit early on. The Training Zones Calculator builds Friel heart-rate bands from your lactate threshold heart rate. With an LTHR of 170 bpm it gives Z2 as 145–151 bpm and Z3 as 153–160 bpm. If marathon effort for you normally sits at the top of Z2 and into low Z3, as the training zones guide describes, then on a hot day cap the first half at about 155 and let the pace be whatever it is.
Heart rate also drifts up over a long effort even when the effort is steady, so a fixed cap gets harder to hold in the second half. Use it to stop yourself starting too fast, not as a target for the whole race.
Heat safety. Heat illness is a real risk in warm races, including for fit runners. Dizziness, confusion, nausea, chills or goosebumps in the heat, or stopping sweating, are warning signs. Stop, find shade and tell the medical staff. Drink to thirst rather than forcing fluid down, because drinking far more than you sweat carries its own risks. This is general information, not medical advice. If you have a heart condition or other health concerns, talk to a doctor before racing in the heat.
Hills cost more going up than they give back coming down
A hilly course is slower than a flat one even when it finishes at the same height it started. Climbing costs much more energy than running on the flat, and descending saves less than the climb cost, especially once the downhill is steep enough that you’re braking. One widely quoted coaching rule puts the cost at about 12–15 seconds per mile for each 1% of uphill grade, and the refund at about 8 seconds per mile for each 1% of downhill. In kilometres, that is roughly 7.5–9.5 s/km lost per 1% up and 5 s/km regained per 1% down.
Apply that to a 2 km climb at 3% on the way out of an out-and-back, with the same 2 km descent on the way home:
- Uphill: 3% × 7.5–9.5 s/km ≈ 22–28 s/km, × 2 km ≈ 45–56 s lost
- Downhill: 3% × 5 s/km ≈ 15 s/km, × 2 km ≈ 30 s back
- Net: about 15–25 seconds slower than flat
That’s small over a half marathon. The bigger risk is running the climb at flat-ground pace, which pushes you well above threshold. You’ll pay for that in the following kilometres, far more than the 20 seconds the hill itself costs. Run climbs by effort, let the pace fall away, and pick it back up on the descent without sprinting. Late climbs hurt more than early ones, so look at where the hills are on the course profile as well as how much climbing there is in total.
Altitude: noticeable from roughly 1,500 metres
At altitude each breath delivers less oxygen, so endurance performance drops. Sprint events don’t suffer, and the thinner air can even help them. In general, unacclimatised runners start to notice the effect from about 1,500 metres. The higher you go and the longer the race, the bigger it gets. The 1968 Olympics in Mexico City, at around 2,240 m, are the classic case: sprint and jump marks were excellent, and distance times were slow.
Published estimates of the slowdown vary widely with the athlete, how long they’ve had to acclimatise and the distance. That’s why a cautious range is more useful than a single figure. For a 10K or longer at 1,500–2,000 m, starting a few percent slower than your sea-level goal is a sensible opening, and more if you arrived the day before. A few days at altitude doesn’t fully acclimatise you. Full adaptation takes weeks. Use heart rate as your limit here too, since your usual heart rate at goal pace will come at a slower pace.
Wind: the headwind costs more than the tailwind gives back
On an out-and-back course or a loop, wind doesn’t cancel out. A headwind adds air resistance that you have to push against. A tailwind takes some away, but you spend less time in it because you’re going faster. Over a full loop you end up slower than on a still day. Air resistance takes only a few percent of a runner’s energy at normal race speeds, which is much less than for a cyclist, so a light breeze barely matters. A strong headwind on an exposed stretch can cost you several seconds per kilometre.
There are two practical responses. Run into the headwind by effort and accept the slower pace. And tuck in behind other runners when you can. Sitting a step or two behind someone cuts the wind’s effect noticeably, which is why elite pacers often run in a group in front of the leaders.
A worked half marathon
A runner’s 10K best is 46:00, run on a cool, flat course. The converter predicts a half marathon of about 1:41:30, a pace of 4:49/km. Race day forecasts:
- 22°C (72°F) with an 18°C (64°F) dew point. 72 + 64 = 136, which falls in the 2–3% band. They’ll take 2.5% because they haven’t trained much in the heat.
- One 2 km climb at 3%, descended on the way back: about 20 seconds net.
- Calm wind, sea level.
1:41:30 × 1.025 ≈ 1:44:02, plus about 20 s for the hill, gives roughly 1:44:20. That’s about 4:57/km over 21.0975 km. In pace mode, the runner enters 4:57 per km and sets “Held for” to 21.0975 km. The converter returns a finish of 1:44:26 and a split for every kilometre, plus a final row for the last 0.0975 km. Those splits are the plan for the flat sections. On the climb they run by effort, and they don’t try to win the lost time back on the descent.
Notice that the converter’s prediction table for other distances is now based on 4:57 pace in warm, hilly conditions. Ignore it for this race. Its purpose here is the split table.
If race morning turns out cooler than forecast, move back toward 4:50 after the first 5 km, once you have evidence. Never go the other way and plan for the cool-day pace in the hope that it won’t be hot.
Before you line up
Run your best recent race through the Running Pace Converter to get the cool, flat, still-day number. Then apply the heat, hills, altitude and wind adjustments above in writing, and put the adjusted pace back into the converter’s pace mode for a split table you can follow on the day. Set an early heart-rate limit from your bands in the Training Zones Calculator, and on a hot or high-altitude course, let that limit decide your pace more than the clock does.
Try the tools from this guide
-
Running Pace Converter & Race Time Predictor
Convert running pace between min/km and min/mi and predict your 5K, 10K, half and full marathon times.
-
Heart Rate, FTP & Pace Training Zone Calculator
Generate training zones from threshold heart rate, FTP or threshold pace for running, cycling and rowing.