Training in the Haze: Which Number Actually Matters

I get the same message every haze season, usually around 6am, usually from someone already in her kit.

"Is it stupid to run today?"

And the honest answer is that it depends on a number most people are not looking at, because the number everyone is looking at is the wrong one for that decision.

So here is the whole thing in one place. What the numbers mean, which one to check and when, what actually changes when you exercise in it, and what to do about the kids.

The short version

Check the 1-hour PM2.5, not the headline PSI. Under 55, train as planned. 56 to 150, cut the long stuff. 151 to 250, intensity moves indoors. Above 250, everything moves indoors. Children, one band earlier than all of that.

There are two numbers, and they do different jobs

In Singapore, NEA publishes two things that matter to you.

The 24-hour PSI is a rolling average of the last 24 hours. It is the headline number, the one in the news, and it is the one that gets quoted in group chats. It is genuinely useful for one thing: planning tomorrow.

The 1-hour PM2.5 reading is the fine particle concentration right now, in micrograms per cubic metre, per region. This is the one that tells you whether to go out in the next hour.

NEA's own guidance is explicit about the split: refer to the 1-hour PM2.5 for immediate activities, and the 24-hour PSI forecast to plan ahead.

This distinction is the single most useful thing in this article. Because a 24-hour average lags badly. Haze blows in at 2pm, your 1-hour PM2.5 jumps to 180, and the 24-hour PSI still reads 85 because it is being dragged down by the clean air you had at 4am. You check PSI, see "Moderate", and head out into air that is three times worse than the headline suggests.

It also works the other way. The haze clears overnight, the 1-hour reading is 30, and the 24-hour PSI is still showing 140 from yesterday's muck. You cancel a session you could have done.

The bands

Singapore, 1-hour PM2.5. The one for right now, in micrograms per cubic metre:

Band 1, Normal 0 to 55

Band 2, Elevated 56 to 150

Band 3, High 151 to 250

Band 4, Very High 251 and above

Singapore, 24-hour PSI. The one for planning:

Good 0 to 50

Moderate 51 to 100

Unhealthy 101 to 200

Very Unhealthy 201 to 300

Hazardous above 300

Malaysia uses a different index entirely. The API is not the PSI. The bands happen to look similar, but it is a different calculation and you cannot assume a reading of 120 in Johor means the same thing as 120 in Singapore.

Good 0 to 50

Moderate 51 to 100

Unhealthy 101 to 200

Very Unhealthy 201 to 300

Hazardous 301 to 500

Emergency above 500

If you train in both countries, or you are racing across the Causeway, check the index that belongs to where your feet are.

Where to actually look, and why your weather app disagrees

In Singapore, NEA publishes current readings on the myENV app, on the haze microsite at haze.gov.sg, on the NEA website, and on its social channels. All of them give you readings by region, which matters, because you are training in one region and not in the national average.

In Malaysia, the Department of Environment publishes station-by-station API readings on its air quality portal. Find your nearest station rather than taking a state-level figure.

Now the bit that causes most of the confusion in group chats.

Your phone's weather app is almost certainly not showing you the PSI. Most international air quality apps and sites report a US-style AQI, which is a different calculation on different breakpoints. NEA says this directly: index values from other websites use different calculation methods and are not comparable with the 24-hour PSI.

So when your app says 95 and NEA says 140, nobody is lying to you. They are two different scales measuring overlapping things. Pick one source and stay with it, rather than collecting numbers until you find the one that gives you permission to do the session you wanted to do anyway.

One more thing worth knowing, because it explains a lot. The PSI is not a PM2.5 reading. It is computed from 24-hour averages of six pollutants: PM10, PM2.5, sulphur dioxide, nitrogen dioxide, ozone and carbon monoxide. Each gets its own sub-index, and the PSI is simply the highest of the six.

During transboundary haze, PM2.5 is almost always the one driving it. That is exactly why the 1-hour PM2.5 is the right number for haze decisions specifically, and why on a non-haze day a slightly raised PSI might be ozone on a hot still afternoon and have nothing to do with smoke at all.

What Sport Singapore actually says about exercising in it

This is the bit most haze articles skip. There are official guidelines written specifically for sport, not just general public health advice, and they are more useful for us because they talk about exertion rather than just going outside.

Two definitions matter, and they are theirs, not mine. Prolonged means continuous exposure for several hours. Strenuous means an activity involving a lot of energy or effort.

Set against the 24-hour PSI:

• 0 to 100: no restriction for anyone.
• 101 to 200, Unhealthy: healthy people should reduce prolonged or strenuous outdoor exertion. Children, elderly and pregnant women should minimise it. Anyone with heart or chronic lung disease should avoid it.
• 201 to 300, Very Unhealthy: healthy people should avoid prolonged or strenuous outdoor exertion. Children, elderly and pregnant women should minimise outdoor activity altogether.
• Above 300, Hazardous: healthy people should minimise outdoor activity. Children, elderly and pregnant women should avoid it.

Notice the pattern. The advice does not jump from fine to stay inside. It steps down in stages, and the first thing to go is not your run, it is the long run and the hard session.

Why exercise changes the maths

Here is the part that explains why you cannot just use the general public advisory and carry on.

When you exercise, two things happen at once. Your minute ventilation climbs to meet demand, so you are simply moving far more air through your lungs per minute than someone walking to the MRT. And you switch from breathing through your nose to breathing through your mouth, which bypasses the filtration your nose was doing for you.

The combined effect is not small. The research puts the deposition of ultrafine particles in your airways at roughly four times higher during light exercise, with a further doubling at high intensity.

4x

more ultrafine particles deposited in your airways during light exercise, and double that again at high intensity.

60 min

the ceiling on planned outdoor sessions when the air is poor. Past 120 minutes is where the lung function findings sit.

So the same air is not the same dose. A woman walking her dog and a woman doing 6 x 1km at threshold in identical air are having completely different exposures.

Two more findings worth knowing:

• Sessions of 120 minutes or more in poor air are the ones most associated with reduced lung function and airway inflammation afterwards.
• The practical recommendation from that literature is to keep planned outdoor exercise under 60 minutes when air quality is poor.

That maps almost exactly onto the official advice, which is reassuring. Duration is the first lever. Intensity is the second.

What I actually do with my athletes

This is my own protocol, built on the above. It is not official guidance and your doctor's opinion beats mine.

Band 1, 0 to 55. Train normally. Nothing changes.

Band 2, 56 to 150. This is where most haze days sit, and where most of the bad decisions get made because it does not look dramatic. Normal sessions go ahead. What I pull is the long stuff. A 90-minute easy run becomes 50 minutes. The two-hour ride becomes the turbo. If you have a key session you have been building towards, keep it, but keep it short.

Band 3, 151 to 250. Intensity moves indoors. Treadmill, turbo, gym. If you must be outside, under 60 minutes and genuinely easy. This is a strength week, not a threshold week.

Band 4, 251 and above. Indoors. All of it. This is not a toughness question.

Two rules on top of all four bands

Check the reading for your region, not the national figure. Singapore publishes by region and the spread between north and west can be substantial. You are running in one of them, not the average.

Symptoms override the number. Scratchy throat, tight chest, a cough that was not there yesterday, a headache that arrives mid-session. Stop. The bands are population guidance. You are one person, and your airways get a vote.

What a haze week actually looks like

Here is a real example of how I rebuild a week, so this is not just theory.

Say you are eight weeks out from a half marathon. The planned week is an easy 8km, a threshold session, an easy 6km, a tempo run and a 90-minute long run. Haze arrives on the Tuesday and the 1-hour PM2.5 sits in Band 2 for four days, with one afternoon spiking into Band 3.

What I would do with that week:

• The easy runs stay outside but get shorter. 8km becomes 5km, 6km becomes 4km. Easy running is where you accumulate the most minutes for the least benefit, so it is the cheapest thing to trim.
• The threshold session moves to the treadmill. Same work, same paces, controlled air. If you hate the treadmill, this is the week to make peace with it.
• The tempo run gets swapped for strength. You were probably under-doing strength anyway, and a haze week is the best excuse you will get all season.
• The long run is the real decision. 90 minutes outdoors in Band 2 is exactly the kind of prolonged exposure the guidance is pointing at. I would either split it, 45 minutes outside early when readings are usually lowest plus 45 on the treadmill, or move the whole thing to the weekend if the forecast is better and accept one unbalanced week.

Net effect: you lose maybe 20 percent of your weekly volume and none of your quality. Eight weeks out, that is nothing. Your fitness does not fall off a cliff in seven days, and the body you bring to the start line will be in better shape than the one that ran every planned kilometre through a smoke haze and arrived with an inflamed airway.

The athletes who get hurt by haze season are not the ones who adapted. They are the ones who held the plan, got progressively more run down, and then could not work out why the last block felt so hard.

The kids

This is the part where I would be most conservative.

Children shift one band earlier

In every official band, children are grouped with the elderly and pregnant women, and the advice for that group tightens one full step earlier than for healthy adults. At Unhealthy, you reduce and they minimise. At Very Unhealthy, you avoid strenuous exertion and they minimise outdoor activity entirely. At Hazardous, you minimise and they avoid.

So the simplest rule I can give you, and the one I would use myself: treat your child's threshold as one band stricter than your own. If the reading has you shortening your run, their football training is cancelled. If it has you moving indoors, they are indoors already.

A few practical notes:

• Kids will not self-report. They will play until they are coughing and then keep playing. You are the monitor, not them.
• The intensity point applies to them more than you, because children's play is almost all intermittent high intensity. An hour in the playground is not an hour of walking.
• Schools in Singapore have their own haze protocols and will act on them. That does not cover the 4pm kickabout at the condo, which is on you.
• If your child has asthma, this stops being a judgement call and becomes a conversation with their doctor, before haze season rather than during it.

Masks, indoor air, and the things that do not work

Only masks rated N95 or equivalent are designed to filter particles in this size range. A surgical mask or a cloth mask is not doing the job you think it is doing.

And a mask is not a training solution. The breathing resistance that makes an N95 effective is the same resistance that makes a hard session in one unrealistic, and if you have any heart or lung condition you should speak to a doctor before using one at all. Masks are for getting to work. They are not for getting through intervals.

Indoors is only better if the air indoors is actually better. An open-plan condo gym with the doors propped open is outdoor air with a roof. A sealed, air-conditioned room with decent filtration is genuinely different. Worth knowing which one yours is before you plan a week around it.

The honest bit

The long-term picture is that exercise remains strongly protective. Being active in polluted air is still associated with a meaningful reduction in mortality risk compared with being inactive. Nobody is telling you that six weeks of haze should turn you into someone who does not train.

What the evidence does say is that the dose matters, and that duration and intensity are the two dials you control. You do not have to choose between training and your lungs. You have to choose between the session you had planned and a different session.

Haze season is annoying, it is unpredictable, and it arrives in the middle of whatever build you are in. But a week of shortened sessions and indoor strength work costs you very little. Three weeks of pretending it is not happening, in Band 3 air, running your long runs because they are in the plan, costs you more.

Check the 1-hour number before you go. Check the 24-hour forecast when you plan. Shorten before you harden. And be stricter with your kids than you are with yourself.

There is more on how this gets turned into a week-by-week plan on the training coaching page.